Thermodynamics of the Phase Equilibriums of Some Organic Compounds

A comprehensive investigation of the phase equilibriums and determination of thermodynamic properties of pure substances is a significant object of the chemical thermodynamics. Data on the phase transitions, heat capacities, and saturation vapor pressure over the solid and liquid phases are used in many fields of science and technology, including calculations on the basis of the third law of thermodynamics. Theoretical and practical applications of thermodynamic data require verification of their reliability. The Clapeyron equation combines different properties of coexisting phases: temperature, vapor pressure, volume, enthalpy of the phase transitions, and caloric values Cp and Cv . Using this equation allows one to verify numerical data for thermodynamic concordance, to reveal unreliable quantities, and to predict failing thermodynamic properties. Mutual concordance and reliability of the calorimetric data on the heat capacity, the saturated vapor pressures, and the properties of phase transition can be verified by comparison of the absolute entropies determined from the experimental data by the third thermodynamic law, ( )( ) o S g m expt with those ones calculated by statistical thermodynamics, ( ) o S stat m . A congruence of these values within errors limits justifies their reliability. Critical analyses of the recent data on thermodynamic properties of some organic compounds are published by the National Institute of the Standards and Technology [NIST], USA. Literature data on the vapor pressures and the enthalpies of vaporization for n-alkanes C5 – C20 were reviewed and critically analyzed in the reference (Ruzicka & Majer, 1994). Thermodynamic properties of many classes of organic compounds were considered in monograph (Domalski & Hearing, 1993; Poling et al., 2001) that favoured the development of the Benson’s calculation method. This chapter deals with reviewing and summarizing the data on the phase equilibriums carried out for some functional organic compounds by the low temperature adiabatic calorimetry, comparative ebulliometry, and vaporization calorimetry in the Luginin’s Laboratory of Thermochemistry [LLT] of the Moscow State University [MSU] and other research centres. The numerous data on the heat capacity, the vapor pressure, enthalpies of the phase transitions, and derived thermodynamic functions were obtained for series of freons, cyclic hydrocarbons and fluorocarbons, and derivatives of ferrocene. A sufficient attention was given to the critical analyses of the thermodynamic data, their reliability, and to interconnections between the properties and some structural parameters of the


Introduction
A comprehensive investigation of the phase equilibriums and determination of thermodynamic properties of pure substances is a significant object of the chemical thermodynamics. Data on the phase transitions, heat capacities, and saturation vapor pressure over the solid and liquid phases are used in many fields of science and technology, including calculations on the basis of the third law of thermodynamics. Theoretical and practical applications of thermodynamic data require verification of their reliability. The Clapeyron equation combines different properties of coexisting phases: temperature, vapor pressure, volume, enthalpy of the phase transitions, and caloric values C p and C v . Using this equation allows one to verify numerical data for thermodynamic concordance, to reveal unreliable quantities, and to predict failing thermodynamic properties. Mutual concordance and reliability of the calorimetric data on the heat capacity, the saturated vapor pressures, and the properties of phase transition can be verified by comparison of the absolute entropies determined from the experimental data by the third thermodynamic law, () ( ) o Sg m e x p t with those ones calculated by statistical thermodynamics, () o Ss t a t m . A congruence of these values within errors limits justifies their reliability. Critical analyses of the recent data on thermodynamic properties of some organic compounds are published by the National Institute of the Standards and Technology [NIST], USA. Literature data on the vapor pressures and the enthalpies of vaporization for n-alkanes C 5 -C 20 were reviewed and critically analyzed in the reference (Ruzicka & Majer, 1994). Thermodynamic properties of many classes of organic compounds were considered in monograph (Domalski & Hearing, 1993;Poling et al., 2001) that favoured the development of the Benson's calculation method. This chapter deals with reviewing and summarizing the data on the phase equilibriums carried out for some functional organic compounds by the low temperature adiabatic calorimetry, comparative ebulliometry, and vaporization calorimetry in the Luginin's Laboratory of Thermochemistry [LLT] of the Moscow State University [MSU] and other research centres. The numerous data on the heat capacity, the vapor pressure, enthalpies of the phase transitions, and derived thermodynamic functions were obtained for series of freons, cyclic hydrocarbons and fluorocarbons, and derivatives of ferrocene. A sufficient attention was given to the critical analyses of the thermodynamic data, their reliability, and to interconnections between the properties and some structural parameters of the

www.intechopen.com
Thermodynamics of the Phase Equilibriums of Some Organic Compounds 599 The setup consists of a differential ebulliometer used for measuring the boiling and condensation temperatures and manometer system operating in the manostat mode. The m a i n p a r t o f M S s y s t e m i s a m e r c u r y -c o n t act (tungsten) manometer that serves for automatic control and determination of the pressure inside the ebulliometer. Argon was introduced into the system to maintain the constant pressure equal to that of the saturation vapors of the substance under study. The temperature of the (liquid + vapor) equilibrium was measured at 20 fixed pressures controlled by manometer system. A schematic view of modified Swietoslawski -type ebulliometer is given in Fig. 2. The differential ebulliometer was used for determination of the temperature dependence of the vapor pressure by measuring the boiling, T boil , or (rarely) condensation, T cond , temperatures and for estimation of an ebulliometric degree of purity for the samples by the difference (  TT boil cond ). The latter is below 0.005 K for the pure substances. The boiling and condensation sections and other parts of differential ebulliometer were made of "Pyrex" glass and were sealed together. The modification of the ebulliometer was directed for solving three basic problems: 1) increasing the thermometric sensitivity of a system used for temperature measurements of the (liquid -vapor) equilibrium; 2) decreasing a heat exchange of the temperature sensors with the surrounding, and 3) reducing the superheating of the boiling liquid, that leads to increasing the accuracy of the temperature measurements. For increasing the sensitivity of the thermometers, their protecting tubes were soldered in the boiling and condensation sections of the ebulliometer. Sensing elements of vibrationresistant thermometers (   100 R ) consisted of a few platinum spirals wound around glass capillaries. The latter had coefficients of linear expiation close to that of platinum. Connecting wires (current and potential) of the thermometer were vacuum -tight sealed through the glass-molybdenum part of a passage (12, Fig. 2) of the protecting tube. The thermometers were graduated in Mendeleev's Institute of Metrology (S. Petersburg) at the triple -point temperature of water (273.16 K) and melting temperatures of tin (505.118 K) and gallium (302.920 K). The summary error of graduation is   3 31 0 K . A special system of heat insulation of the thermometers was employed. It consisted of: glass screens washed by boiling liquid or by condensate (2 1 or 2 1 '), respectively (Fig. 2); silver radiation screens; vacuum jackets (   3 1.3 10 p Pa ); heat -insulating layers (asbestos), and electrically heated screens (2 5 and 2 5 ') for upper parts of thermometers overhanging from ebulliometer. Application of such heat -insulating system made it possible to conduct precision temperature measurements without heating the main part of the ebulliometer. The error of temperature measurements caused by heat exchange of the thermometers with the surroundings, were estimated on the basis of heat -exchange laws to be   3 11 0 K . The superheating of the liquid was reduced by using several internal and two external boiler heaters which promoted to smooth boiling of the liquid. Performed modification of the ebulliometer allowed cutting down substantially the amount of liquid which was spent for heating the inner surfaces of instrument up to working temperature.
Thus, the necessary volume of liquid was reduced several times: down to  3 5cm when measuring only the boiling temperature and to  reduce the pT parameters determination to the precision temperature measurements. The temperature was automatically measured by potentiometer method and the results were displayed on a personal computer [PC] screen with the aid of the AK-6.25 computermeasurement system designed at All-Russia Research Institute of Physico-technical and Radio-technical Measurements [VNIIFTRI]. An automatic maintenance of the constant pressure was attained by a mercury-contact manometer which was controlled by vacuum pump via an electromagnetic valve (Fig. 1). The pressure of argon fluctuated in the limits from ( 20 to 40 ) Pa. The boiling temperature was measured at the highest pressure in the cycle at the moment of mercury-to-tungsten contact. The manometer was thermostated at the temperature (300.00±0.02) K. The measurements of the boiling and condensation temperatures were conducted after attaining thermodynamic equilibrium in the ebulliometer. To be assured that the liquid under study had not decomposed, the boiling temperature at one of initial points of the pT curve was measured several times during the ebulliometric experiments. Errors of temperature S T and vapor pressure S p measurements were calculated as: where  1 tS T and  3 tS T denote the instrumental errors of the temperature measurements (   3 51 0 K ) at the substance research and at graduation of the mercury-contact manometer; t is Student's criterion;   3 31 0 2 SK T denotes the error of graduation of the thermometer; and ( / ) 1 dp dT and ( / ) 3 dp dT are temperature coefficients of the pressure for standard and studied substances, respectively. The total uncertainty of temperature measurement was   3 61 0 SK T . The error of graduation of the mercury-contact manometer by means of water and n-decane and the error of determination of the vapor pressure of the substance under study were equal to (. ) S grad p = (13 to 20) Pa and S p = (20 to 26) Pa, respectively. The accuracy of ebulliometric measurements was checked by determinations of the saturation vapor pressures of substances having significantly different boiling temperatures, namely benzene and undecane. The normal boiling temperatures of the standard substances obtained in this work agree within errors limits   0.01 K with precise values of reference (Boublik et al., 1984). Comparative ebulliometry was employed for determination a series of saturation vapor pressures in dependence on temperature for some freons; halogen -ethanes and -propanes; alkyladamantanes; cis-and trans-hydrindanes, cis-and trans-decalines, and their fluoridated counterparts. The mathematical processing of the observed boiling temperatures and vapor pressures were conducted by the semi -empirical equation: www.intechopen.com  H  T  T  TTT  TT  m   TT  T T  T T   (1) where T denotes the mean temperature and    () ,, 12 HT m , and  3 are parameters.
Equation (1) was derived by integration of the Clapeyron equation: with the approximation for   / HZ vap m : where Z denotes the difference of compression factors of gas and liquid. Equation (3) , as a linear function of the temperature. The treatment of the pT parameters was carried out by the least-squares method [LSM] using orthogonal functions (Kornilov & Vidavski, 1969). Mathematical processing of the saturation vapor pressures is given in Appendix. A system of normal equations of LSM is a diagonal matrix relative to the orthogonal functions. The latter are mutually independent that allows to evaluate their uncertainties and those ones for the ln{ ( ) p T /Pa} and  () HT vap m functions and, as a result, to choice of an adequate number of terms of relations (1) and (3) by curtailing or expanding terms to suit the accuracy of the parameters of these relations without a new treatment of pT data. Final equations for these functions are set out for compactness, as: where A, B, C, and D are constants related to the parameters of equation (1) Table 1. Thermodynamic parameters of comparative ebulliometry for compounds studied: freons; halogen -ethanes and -propanes; 1,3-dimethyladamahtane [1,3-DMA], 1,3,5trimethyladamahtane [1,3,5-TMA] and 1-ethyladamahtane [1-EA]; perfluorobicyclo(4,3,0)nonanes [cis-and trans-C 9 F 16 ], bicyclo(4,3,0)nonanes, [cis-and trans-C 9 H 16 ], perfluorobicyclo(4,4,0)decane, [cis-and trans-C 10 F 18 ], bicyclo(4,4,0)decanes [cis-and trans-C 10 H 18 ]; perfluoro-N-(4-methyl-cyclohexyl)piperidine [C 5 F 10 N-C 6 F 10 -CF 3 ] Boublik et al., 1984) www.intechopen.com Thermodynamics of the Phase Equilibriums of Some Organic Compounds 603 is satisfied, the parameter  3 (D) may be accepted as a reliable one. Here F and (1, ) 0.05 Ff denote evaluated and tabulated values of the F -criterion, and f is a number of degrees of freedom. Comparing the criteria F and (1, ) 0.05 Ff according to (6) showed an adequate fit of the pT parameters. Table 1 summarizes the purity of the compounds determined by gas -liquid chromatography [g.l.c.] and adiabatic calorimetry, the temperature interval, T ( pT ), and number, n, of pT -parameters, the coefficients of equations (4) and (5) and mean-square deviation [MSD] of calculated p calc -values from experimental ones, p ,

The enthalpy of vaporization
Experimental determinations of the enthalpies of vaporization were carried out by direct calorimetric methods and by indirect ones, on the basis of the temperature dependences of saturation vapor pressures. The first method is more precise but the second one is more often used because of it's applicability for wider series of the substances. The enthalpies of vaporization of some compounds under study were determined at T = 298.15 K by calorimetric method using a carrier gas (nitrogen) (Wadsö, 1966). The method is based on measuring the energy dissipated in calorimeter for compensation of the endothermic vaporization effect. The carrier gas was employed for hastening an evaporation process and, thus, for increasing an accuracy. A modified LKB 8721-3 setup consists of some commercial parts, namely calorimetric vessel with an air brass jacket and a carrier gas system and three missing parts designed in (Varushchenko et. al., 1977): precise water thermostat, electrical scheme, and an air thermostat. The latter replaced a thermostated room that was provided for operating by this method. The calorimeter is intended for the substances with vapor pressures from 0.066 kPa to 26.6 kPa at 298 K (or normal boiling temperatures from (335 to 470) K). A mass (0.5 to 1.0) g of substance was required for a series from 6 to 8 experiments. The calorimetric experiment was conducted at an adiabatic and, at the same time, at isothermal conditions. The temperature of the calorimetric vessel measured by a thermistor was maintained constant and equal to that of the thermostat (298.15±0.02) K. Electrical energy used for compensation of the energy of vaporization (20 to 40) J was measured by a potentiometer method with accuracy 0.01 per cent. The mass, m, of a substance evaporated (0.07 to 0.3) g was determined to   4 1 10 g as the difference between masses of calorimetric vessel before and after an experiment. As the calorimeter was non-hermetic, the main error in mass determination arose from a loss of substance in weighing the vessel due to connecting and disconnecting it with the calorimetric system. All preliminary procedures such as filling the vessel with liquid, weighing it, and placing into its air jacket were made inside an air thermostat at  298 T K. In so doing, we reduced a loss of the substance from the vessel and the temperature over fall of the latter. The value of  H vap was corrected for a small quantity of energy absorbed during the passage of nitrogen through the calorimeter under low pressure. The calorimeter was tested by measuring the enthalpies of vaporization of n-alkanes from C 6 to C 10 . Obtained values of  H vap at  298.15 T K agree with well established literature values (Majer & Svoboda, 1985) within (0.2 to 0.5) per cent. A main method of determination of the enthalpies of vaporization is until now an indirect one based on the temperature dependence of the vapor pressure. This is caused by a less complicated technique for precise vapor pressure determinations than direct calorimetric measurements of  H vap m . The best-accuracy estimations of  H vap m values are attained for a moderate range of vapor pressure (5 to 150) kPa. The literature data on the enthalpies of vaporization obtained by indirect method are usually published without uncertainties, that can be explained by fitting the pT -parameters with  ln( ) ( ) p fT equations, coefficients of which were correlated. An accuracy determination of the enthalpies of vaporization in indirect method is given in Appendix. The  H vap m values obtained by indirect method were computed by equation (5) using the Z difference which took into account the vapor deviation from ideality and volume changes of both phases. The Z values were calculated from formula: The molar volume () V liq m of liquid was evaluated on the basis of density; an adequate value for the volume of vapor, () Vg m , was calculated from the volume-explicit virial expansion truncated after the second virial coefficient B v . The values of B v were evaluated on the basis of critical quantities (part 4.2) by the Tsonopolous extension of Pitzer and Curl's method (Poling et al., 2001). Comparing two series of Z values estimated from experimental and calculated values of () Vg m of hydrocarbons enable us to accept the errors of Z evaluation ≤ 1 per cent. Freons and halogenalkanes.  (Boublik et al., 1984). It has been shown that these equations allowed us to estimate the  H vap m values with uncertainties  2 per cent in extrapolation intervals  50 T K. Mutual congruence of some thermodynamic properties in set of related compounds (Table  2) can be drawn from comparison of these properties in dependence on some physicochemical characteristics having influence upon intermolecular interactions in liquid state. Fig. 3  coefficients K m were calculated by analogy with . In spite of the large atomic weight of fluorine in comparison with hydrogen, thermodynamic values of compounds decrease when hydrogen is substituted for fluorine that can be explained by decreasing of the  () liq and K m parameters. Minimum T c , ..
T nb , and  (298.15 ) HK vap m values are inherent to completely halogenated 1,1,1-trifluoro-2,2-dichloroethane, which has the lowest values of dipole moment and K m coefficient. Maximum values of corresponding properties are observed for the most polar compounds, 1,1,2,-trichloroetane, the gauche conformer of which is stabilized by the dipole interaction in the liquid phase. Analysis of the data shown in Fig. 3 allows to conclude that the values of critical and normal boiling temperatures and enthalpy of vaporization vary in a series of compounds according to the combined action of the parameters responsible for intermolecular interactions and short range order of the liquid phase, thus proving the mutual consistency of the thermodynamic data in the series of halogenated ethane and propane. Cyclic perfluorocarbons and hydrocarbons. A thermodynamic study of perfluorated cyclic organic compounds has scientific and practical importance. Perfluorocarbons [PFC] have high chemical and thermal stability, absolute biological inertness, and weak intermolecular interactions [IMI]. The combination of these properties can be assigned to high C-F bond strength and the shielding effect of fluorine atoms towards the carbon framework. The weakness of IMI is responsible for the ability of PFC to dissolve and transfer considerable amounts of gases, in particular, oxygen and carbon dioxide. On account of these properties, PFC have found wide application in biology and medicine as efficient gas-transfer media (blood substitutes).  (Boublik et. al., 1984). Table 3. Normal boiling temperatures, ..   values were evaluated on the basis of the enthalpies of vaporization by empirical method developed within a theory of regular solutions (Lawson et al., 1978). Table 3 presents derived thermodynamic values of cyclic compounds. The values of the normal boiling temperatures and the enthalpies of vaporization of cis-isomers are more than those of trans-isomers in the series of perfluorobicyclo-nonanes and -decanes and their hydrocarbons analogues. Despite the more molecular mass, the normal boiling temperatures .. T nb and the  H vap m values of the perfluorocarbons are less than those of the hydrocarbons. On the contrary, the oxygen capacities are two times more in the series of perfluorocarbons which can be explanted by more poor intermolecular interactions of PFC. Fig. 4 presents the critical temperatures, enthalpies of vaporization, and oxygen capacities, () 2   , for cis-and trans-perfluorobicyclo(4,3,0)nonanes (1 and 2), for components of Ftorosan blood substitute (Ries, 1991), namely perfluorobicyclo(4,4,0)-decanes (3 and 4), perfluoro-N-(4-methylcyclohexyl)piperidine (5), and for some of their hydrocarbon analogues (6-9), respectively. Due to smaller energies of intermolecular interactions, the critical temperatures and enthalpies of vaporization of perfluorocarbons are less, but oxygen capacities are more, than appropriate properties of appropriate hydrocarbons. Despite the more molecular mass, the normal boiling temperatures and enthalpies of vaporization of perfluorocarbons are less than those of appropriate hydrocarbon. This can be explained less coefficients of molecular packing, K m , and therefore by more intermolecular distances, and as a consequence less intermolecular interactions of perfluorocarbons in comparison with their hydrogen -containing counterparts.

The vapor pressure and enthalpies of vaporization of the hard-volatile compounds
The saturation vapor pressures of the solid and liquid substances having p < 1 kPa were determined by a dynamic method of evaporating the sample in a stream of the carrier inert gas. In calculation of the enthalpy of vaporization, the volume of vapor is well described by the ideal gas law and the volume of liquid can be easily neglected without introducing essential error into the  H vap m value. But the / dP dT or ln( ) / dpd T derivatives are determined not enough reliably because the saturation vapor pressure is a weak function of the temperature. Thus, an accuracy of determination of the enthalpy of vaporization is restricted for the compounds with low vapor pressures at about 298 K temperature. The temperature dependences of the vapor pressures for the ferrocene derivatives [FD] were determined by a transpiration method elaborated and fully described by Verevkin S.P. and coathers (Emel'yanenko et al., 2007). Here, only the main features of the method are given. The determination of the vapor pressure is based on the measurements of the mass of substance transpired in the stream of carrier gas (nitrogen) and the volume of the gas flowing. The vapor pressure of the substances was obtained by Dalton law for partial vapor pressures of the ideal gas mixture. A sample of the substance (~0.5 g) was placed into the Utube, temperature of which was controlled with accuracy ±0.1 K. A nitrogen flow, controlled by a precision Hoke valve and measured with a bubble gauge, was passed through the tube. The transferred substance was condensed in a cooled trap and was analyzed chromatographically using the external standard (hydrocarbons). The rate of the nitrogen flow was adjusted to ensure that the condensed and vapor phases were in stable equilibrium. The saturation vapor pressure p sat was calculated by the formula: ; m and M are the mass of the sample under study and molecular weight of FD, respectively; V(N 2 ) and V(FD) are the nitrogen and FD volumes, respectively, V(N 2 ) > V(FD); and T is the U-tube temperature. The V(N 2 ) value was determined from the flow rate and the measurement time.
The pT parameters of the solid FD were measured in the pressure and temperature intervals from (0.01/0.11 to 0.44/4.9) Pa and from (311/342 to 341/379) K, respectively. Appropriate pressure and temperature intervals for the liquid FD were from (0.3/1.87 to 7.88/130) Pa and from (298/384 to 358/430) K, respectively. The vapor pressures of FD were approximated by equation: where a and b are coefficients,   () ( / ) ,, , CC g Cc r l i q pm pm pm is the difference between the heat capacities of the vapor and condensed phases, and T st = 298.15 K is the standard temperature (arbitrarily chosen). Equation (9) was deduced by integration of the correlation (Kulikov et al., 2001). The latter was obtained on the basis of Clausius-Clapeyron equation value is independent on the temperature in the pT interval under study. The enthalpy of vaporization was calculated by the formula: obtained by differentiation of equation (9) with respect to 1/T. The ideal gas heat capacities of the ferrocene derivatives [FD] were obtained by additive Chickos and Acree method (Chikos & Acree Jr., 2003) that is defined as "an atom together with all of its ligands". Table 4 lists the purity of ferrocene derivatives determined by adiabatic calorimetry (part 3.2), coefficients a and b of equations (9) and (10), and enthalpies and entropies of vaporization and sublimation of FD at T = 298.15 K.
Comp ounds  values ranged from (70.3±1.0 to 76.78±0.85) kJ·mol -1 was obtained, the most part of the data being focused in the range between (72 and 74) kJ·mol -1 . Uncertainties of these quantities were probably the random errors. Taking into account the uncertainties of the initial vapor pressure data making up from (1.5 to 2) per cent, a total value of the random and systematic errors could be  2 %. Therefore, the errors of the enthalpies of vaporization and sublimation as derivative values of the vapor pressure in the transpiration method were evaluated as ±  2 %.

The heat capacity and thermodynamic properties of the phase transitions
A heat capacity is a capability of the substance for absorbing some quantity of the energy that increases its temperature by 1 degree K. A measurement of the heat capacity is performed by adiabatic and isothermal methods. The first one allows attaining the most complete thermodynamic equilibrium or, in any case, the thermal balance in the calorimetric system. The adiabatic method is used for exploring the thermal processes with different times of relaxation and the metastable phases which can exist in wide temperature ranges. The heat capacities and thermodynamic properties of the phase transitions were investigated in this work by low-temperature adiabatic calorimetry (Varushchenko et al., 1997a).

Experimental
The measurements of the heat capacities were conducted in a fully automated setup, consisted of a vacuum adiabatic calorimeter, a data acquisition and control system, AK-9.02, and a personal computer, PC (Fig. 5). The setup was produced in the National Scientific and Research Institute of Physical Technical and Radio-Technical Measurements (Mendeleevo, Moscow Region). The main principles of its construction were published in (Pavese & Malyshev, 1994). The calorimetric cell consists of a container, 1, a copper sleeve, 2, in which the container is tightly held, and an adiabatic shield, 3. A bronze brass lid, 4, serves for vacuum-tight sealing the container by means of indium gasket and a simple manifold. To decrease the heat capacity of the empty calorimeter, the miniature rhodium-iron resistance thermometer, 5,  (5 0 ) 0 R was mounted on the inner surface of the adiabatic shield. The thermometer, which was calibrated on ITS-90, is destined for temperature measurements from (0.5 to 373) К with accuracy   3 3 10 K. The temperature difference between the calorimeter and the adiabatic shield is measured by a four-junction thermocouple, 6, (Cu + 0.1 per cent Fe alloy against Chromel), one end of which was mounted on the copper sleeve 2 and the other one was placed on the inner surface of the adiabatic shield, 3. A manganin calorimeter heater (R = 300 Ω) was wound non-inductively on the sleeve, 2. A well-known three-lead circuit diagram was employed for wiring the current and potential leads of the heater. Since the resistances of the current leads are equal, this diagram enables us to account for the heat generated in the leads between the calorimeter and the shield. To reduce the level of heat radiation, the shield was wrapped with several layers of aluminium-coated Lavsan film, 7, (ACLF, an analog of Mylar). The container sleeve, 2, is suspended inside the adiabatic shield on three nylon threads, 8, which are stretched by a spring, 9 (Fig. 5). The calorimeter cell has been fixed on an epoxy/fibre-glass tube, 10, of the cryostat, CR. The tube, 10, is fastened to a copper plug, 11, by means of a bayonet joint. The only removable part of the calorimeter cell is the container for the specimen. The vacuum jacket, 12, is made from oxygen-free copper. The vacuum seal of the cryostat is provided by a KPT-8 silicon/boron nitride paste, which has high thermal conductivity value and gives a stable vacuum junction after freezing. The paste is put between the upper part of the jacket, 12, and the plug, 11, in its grooves, 13. The top part of the cryostat (CR) has a valve, 14, detachable vacuum, 15, and cable, 16, joints; the latter connects the electrical leads of the calorimeter cell to AK-9.02 and PC. Both parts of the cryostat are jointed by the stainless steel tubes, 17. Due to small size (l = 120 mm, d = 22.5 mm), the cryostat is immersed directly into a commercial transportation Dewar vessels. This allows us to exclude an intermediate Dewar vessel and, thus, to reserve the coolants. A coupling nut, 18, with a Teflon shell and a rubber ring is used to fasten the cryostat airtight inside the neck of the Dewar vessel. A T-connection, fitted on the neck of the nitrogen Dewar vessel, enables us to pump out nitrogen vapors to lower the bath temperature if necessary. The calorimeter cell is cooled down by thermal conductivity via electrical leads and by radiation heat transfer. The leads of the thermometer, heaters, and differential thermocouple form a heat shunt with the preset thermal resistance and they provide cooling of the calorimeter from room temperature to approximately T = 78 K, and from T = 78 К down to T = 5 К for about 7 h in each Dewar vessel. The helium heat-exchange gas is not used for this purpose in order to avoid problems, connected with it desorption. To reduce the heat losses by radiation, the additional radiation screens, 20, are used (Fig. 5). The data acquisition system AK-9.02 is a single unit, connected with a personal computer [PC]. The system AK-9.02 and the PC perform the measurements of all values that are necessary for the determination of the heat capacity, as well as the control of the measurement process and data processing. The thermometer resistance and the calorimeter power heating are measured by a potentiometer method with cyclic inversion of the direction of thermometer current for excluding the thermal electromotive forces. All the procedures that control the measurement process are carried out by the PC, which has a simple and user-friendly interface. The results of the measurements are printed and displayed on the screen for visual monitoring. An adiabatic condition in calorimeter is maintained by the AK-9.02 system, which allows keeping the temperature drop between the container and shield on the average within    3 (1 3) 10 K. Owing to modification of the calorimeter , the drop of temperature was reduced to ~ 0.5 mK at the expense of using an eleven -junctions thermocouple instead of four -junction one and employing an additional heater (R ~ 133 Ω) mounted in the upper part of the shield, to which electrical wires of the thermometer and the main heater were connected. Additional heater allows making up a lack of the second adiabatic shield that is usually employed in the adiabatic calorimeters, but cannot be place in our miniature device. Due to small size, the cryostat with the calorimeter was placed in the transport Dewar vessels with refrigerants (liquid helium or nitrogen), that allows us to exclude an intermediate Dewar vessel and, thus, to keep the coolants. There is no constant pumping of the cryostat during the operation, since high vacuum inside the cryostat was kept by means of cry-sorption provided with an efficient charcoal getter. The degree of vacuum in cryostat is controlled by the value of the heater current in the adiabatic shield. This value was determined in a process of the calorimeter production using nitrogen and helium baths. The automatic procedure of the heat capacity measurements is performed by AK -9.02 system running under PC control (Pvese & Malishev, 1994). The program realizes a method of the discrete input of the energy in two modes: constant increments of temperature, T (from 1 to 2) K during measurement of the heat capacity and constant impulses of energy in studying the phase transitions. The calorimetric experiment consists of six periods (Fig. 6). In the first period the calorimeter is heated to a desired temperature. A steady temperature equilibration is attained in the second period. In the third period the temperature of the calorimeter is monitored over a chosen time interval to acquire information about the temperature drift rate, V i and to obtain the linear relation between the values V i and the time by the least-squares method. During the fourth (heating) period the electrical energy is supplied to the calorimeter, and the heating-up time is observed. The fifth period is the same as the second one. In the sixth period the linear relation between the temperature drift rate of the calorimeter V f and the time is established exactly in a similar manner to that in the third period. The initial and the final temperatures of the calorimeter in the main (heating) period are calculated by extrapolating the linear dependencies of the drift rates V i and V f on time to the midpoint ( m ) temperature (Fig. 6). This method permits the heat interchange between the calorimeter and surroundings to be taken into account (Varushchenko et al., 1997a). The reliability of this method was proved by a congruence within (0.1 to 0.2) per cent of the heat capacity values of an empty calorimeter, measured in the temperature interval (90 to 110) К using different refrigerants: liquid helium and nitrogen. The values of the heat capacities, , C sm , are fitted with polynomials: The coefficients of the polynomials A i and  A i were estimated by the LSM.
The metrological characteristics of the calorimeter were tested by measuring the heat capacity of pure copper having a mass fraction of 0.99995 and n-heptane in the temperature intervals (from 8 to 372) K and (from 6 to 354) K, respectively. Obtained , C sm values of copper and n-heptane came to an agreement with the precise heat capacities of standard substances within (0.2 to 1.4) % below the temperature 70 К and decrease to (0.01 and 0.3) % above T=70 К.

Determination of thermodynamic properties of the phase transitions
The important characteristics of the substances: a triple point temperature, tp T , and a mole fraction of impurities, 2 N , were determined by calorimetrical method of the fractional melting study, developed by Mair, Glasgow and Rossini. A linear dependence between the reciprocal fractions of the sample melted, 1/F i , and the equilibrium fusion temperatures, T i , makes it possible to calculate both the tp T value and mole fraction of impurities, 2 N , by equations: Here  0 dT is the triple point temperature, tp T , of the pure compound, According to (Alexandrov, 1975), melting curves can be concave not only in the case of solid solutions, but also in the absence of equilibrium in the calorimeter at the onset of fusion, when the amount of the liquid phase is small and impurities can therefore be distributed no uniformly, and at the final stage of melting, when sedimentation of crystals to the bottom of container interferes with slow attainment of temperature equilibrium. In conformity with Alexandrov recommendation (Alexandrov, 1975), the tp  1960) and (Alexandrov et al., 1983): Here, A k is cryoscopic constant for the major substance; and k is a distribution coefficient of impurities between the solid and liquid phases. An insufficiency of this equation for calculating 2 N consists in the need to determine k coefficient by an independent method. By differentiating and finding the logarithm, equation (15) was transformed by (Alexandrov et al., 1983) to the form: Equation (16)  where  H tot is the total enthalpy absorbed in heating the calorimeter from initial temperature 1 T < T tp to final one 2 T > T tp ;  1 H and 2 H  are the changes of enthalpy www.intechopen.com calculated from the normal heat capacities of the crystal and liquid in the temperature intervals from 1 T to T tp and from T tp to 2 T , respectively; H emp is the enthalpy increment need for heating the empty calorimeter from 1 T to 2 T .

Crystal phase transitions and molecular dynamics
The solid state transitions revealed in the molecular crystals can be explained by different polymorphous transformations, caused by changing the crystal structure, different location of the molecules and their orientational and conformational disorder in the crystal lattice. In this chapter, some thermodynamic properties of solid state transitions and fusion are reviewed for some compounds, which were studied in the Luginin's Thermochemistry Laboratory of the Moscow State University and in some other thermodynamic Laboratories. An interpretation of the solid-state transitions in organic crystals was successfully fulfilled in a set of outstanding researcher's works (Westrum & McCullough, 1965;Kolesov, 1995;Adachi, et al., 1968) and the others. An interpretation of calorimetric measurements was carried out very often on the basis of the order -disorder concept. Understanding these processes requires sometimes exploring the molecular crystals by X-ray crystallography and IR and Raman spectroscopy. In this work, the solid state transitions will be discussed including some additional physico-chemical properties of the compounds. The values of thermodynamic properties of the phase transitions are given in Table 5. It was found by exploring IR-spectra of CF 2 ClCFCl 2 , that this substance comprises a mixture of trans-and gauche-conformers in solid (crystal I) and liquid states (Fig. 7). The sum = 20.52 JK -1 mol -1 for CF 2 ClCFCl 2 is small, while such sum used to be from (42 to 50) JK -1 mol -1 for organic crystals. By comparesing the calorimetric and spectroscopic 0 S m values, it was found that CF 2 ClCFCl 2 has residual entropy, (0) S = 10.1 JK -1 mol -1 at T= 198.15 K (Higgins & Lielmers, 1965;Kolesov, 1995).   A characteristic feature of solid state transition of organic crystals is a slow thermal equilibrium between co-existing phases which very often promote to formation of metastable phases existing in a wide temperature range. In Fig. 7(b), the heat capacity , C sm of 1,1-difluoro-1,2,2-trichloroethane, CF 2 ClCHCl 2 , is shown in the temperature interval studied. Similar  () , C f T sm dependence has been obtained for isomer of 1,1-difluoro-1,2,2trichloroethane: CFCl 2 CHFCl. Both isomers were in the forms of glasses, supercooled liquids, and partially crystalline states. The latter was attained after annealing the specimen at temperatures from (110 to 114) K during 3 days, followed by quenching it at  T 78 K over a period of 12 h. The heat capacity jumps, accompanying G-transitions, are observed on the  CT s curves of both freons. Taking into account this typical transitions for the glasses, the authors of reference (Adachi, et al., 1968) proposed a term "glassy crystal" for the frozen -in disordered states (AB) (Fig.7, (b)). The temperatures of the glass transition  T g 95.7 K and fusion,  T fus 123.1±0.4 K have been obtained. The degrees of crystallinity,  , appropriated to the mole fraction of crystalline samples, equal to 0.076 and 0.116 for CF 2 ClCHCl 2 and CFCl 2 CHFC , respectively, were calculated on the basis of calorimetric data on C s jumps by studying the G-transitions (Varushchenko et al., 1997b). Fig. 8 presents the  CT s curve of ferrocenyl-n-propane, which exhibits a fusion and a gradual solid-to-solid transition in the temperature range from (156 to 204) K. The temperature of the gradual transition of crystal II to crystal I of ferrocenyl-n-propane was ascribed to that of the maximum C s -value in the peak of solid state transition. A test of the calorimetric experiment showed that -anomaly was accompanied by decreasing the heat capacity, C s =-2.7 JK -1 mol -1 and continuance changing the enthalpy and entropy. Thus, the solid-state anomaly is the phase transition of the second order and can be interpreted as the "order-disorder" transformation. The changes of the enthalpy,  0 H m trs , and entropy,  0 S m trs of the thermal anomaly were evaluated by summing up these values in each experimental C s point with subtracting changes of appropriate functions for the empty calorimeter and those ones for the hypothetic normal parts of the crystals I and II. The nature of this transition was studied by the X-ray crystallography. Table 6 lists crystallographic data of n-propylferrocene in vicinity of the "order-disorder" transition. The structure of n-propylferrocene at 200 K contains a propyl-group disordered between two positions (in ca. 2:1 ratio) obviously due to thermal motion (Fig. 9). The transition of the crystal II (T = 150 K) to the crystal I (T = 200 K) occurred with significant changes of the lattice parameters: basis vectors, a, b, c, angle, , the volume, V, the number of molecule in the unit cell, Z, and the factors of the crystal structure quality, R 1 and wR 2 .  Table 6. Crystallographic data of n-propylferrocene crystal at the temperatures 150 K and 200 K While both a and b parameters became larger during phase transition II to I, the c parameter and angle slightly decreased (Table 6). Structure solution of the crystals I and II revealed that lower temperature modification of the n-propylferrocene molecule has only one orientation of the propyl group. Apparently, the transitions in reverse order occurred when cooling the crystals from (150 to 200) K. Thus, the solid-state anomaly of n-propylferrocene is caused by the onset of the internal rotation of propyl groups in the molecules and also by a small shift of the pentadienyl cycles around the axis passed through their centers. These variations led to some orientation disorder of the crystal phase II (Fig. 9). The order-disorder conception is successfully used in exploring the plastic crystals. Adamantane and some of its derivatives form disordered plastic crystalline phases. The fusion of such substances occurs some times in two stages. First, an orientational disorder proceeds in the crystalline lattice because of high mobility of the molecules, and then the plastic crystals fuse owing to a translational molecular motion at higher temperature. In this case, the magnitudes of enthalpy and entropy of the solid-to-solid transition are several times larger than those of the fusion. In accordance with the empirical Timmerman's criterion, the  0 S m fus values for the plastic crystals are usually less than 20 Jmol -1 K -1 . There are two modifications of the plastic crystals with different molecular reorientations, isotropic and anisotropic. Some substances are known to form both of reorientations. Fig. 10 presents  CT s curve of 1,3,5-trimethyladamantane (1,3,5-TMA) explored in this work (a) and the Raman spectrum of the substance (b). A spectroscopic investigation of 1,3,5-TMA was carried out together with that of 1,3-dimethyladamantane (1,3-DMA), which also formed plastic crystals. Both compounds have low values of the entropy of fusion,  0 S m fus = 8.1 (6.2) JK -1 mol -1 and narrow temperature intervals of existence of the high temperature crystal I, T = 21 (24.2) K, respectively. According to adiabatic calorimetry, these properties are typical for the plastic crystals. Distinct bands were observed in their spectra at the low temperatures. As it follows from joint spectra discussion, the number of bands for 1,3,5-TMA is about haft of that for the 1,3-DMA crystal. This suggests that more symmetric molecules of the former compound compose the lattice of higher symmetry and/or with the less quantity of units in a primitive cell. In the vicinity of the phase transitions, the changes in spectra become more pronounced. At the transition points, all the bands disappear and transform into the wing of a broad Rayleigh scattering. This implies that all of them were caused by the librational modes in the low-temperature crystals of 1,3-DMA and 1,3,5-TMA and that the character of molecular motion is different in high-temperature solid phases. The absence of the preferential axes of molecular reorientation in the latter implies that there are isotropic plastic crystals. Analysis of Raman www.intechopen.com spectra is in agreement with the low values of transition entropies measured by adiabatic calorimetry for alkylderivatives of adamantane. The cis-and trans-isomers of perfluorobicyclo(4,3,0)nonane are almost spherical "globular" molecules, which are able to form the plastic crystals due to unusually high molecular mobility. The thermodynamic properties of compounds are known to change when going from fluoroorganic compounds to their hydrogen containing analogous. With this in mind, the thermodynamic properties of the solid state transitions of perfluorocarbons are compared with those of appropriate hydrocarbons.  (Table 5) for transition of the crystal III-to-crystal II, this solid phase conversion can be attributed to an anisotropic molecular reorientation about preferential common C-C axis. According to the empirical Timmerman's criterion, small values of the fusion entropy,  0 S m fus = 9.34±0.09 Jmol -1 K -1 of cis-C 9 F 16 (Table 5) indicates the onset of the isotropic molecular reorientation and, thus the formation of the plastic crystals. As is seen from Table 5, the cis-bicyclo(4,3,0)nonane has thermodynamic properties of the solid-state transition, analogous to its perfluorated counterpart, and forms plastic crystals. The trans-perfluorobicyclo(4,3,0)nonane exhibits one solid-state transition and also forms the plastic crystals, but its hydrogen analogous, trans-bicyclo(4,3,0)nonane do not form the plastic crystals. These properties of the trans-isomers can be explained by their less spherical shapes as compared with the cis-isomers.
www.intechopen.com  Fig. 12) permits to interpret the influence of the structure and chemical nature on molecular mobility and thermodynamic properties of the solids. Comparesing the entropies of fusion shows that the mobility of the molecules increases in going from hydrocarbons to appropriate perfluorocarbons and from the cis-isomers to the trans-ones.
The larger molecular mobility of the perfluorocarbons can be explained by more weak intermolecular interactions for these compounds compared with the hydrocarbons. The greater ability of the molecules of cis-isomer to reorient in the solid state seems to be due to the steric factors. The nature of the solid-to-solid transitions in cis-[7] and trans- [8] bicyclodecanes and cis-[5] and trans-[6] perfluorobicyclodecanes (Fig. 12) were discussed in the order-disorder concept in reference (Kolesov, 1995).

Thermodynamic functions in the ideal gas states
The absolute entropy, changes of the enthalpy and Gibbs energy in three aggregate states are calculated on the basis of smoothed heat capacity values. The experimental data on the heat capacities of the substances under study in the temperature intervals from (6/8 to 372) K were fitted by polynomial (11). Extending the heat capacities to  0 T were performed by Debye equation: where D is the Debye function, n = 3, and  denote Debye characteristic temperature. Testing the s,m C values at  0 T was performed by fitting the heat capacities in small temperature interval below (10 to 12) K by equation: where  and  are coefficients. If  = 0, it can be accepted that s,m  0 C and extrapolation of the heat capacity to  0 T can be carried out by equation (18) (18) and (11) and adding the enthalpies and entropies of the solid-to-solid transition and fusion. The errors of thermodynamic functions were estimated by the law of random errors accumulation using the uncertainties of the heat capacity measurements. The ideal gas absolute entropy, 0 m () ST , the changes of the enthalpy and the free Gibbs energy at 298.15 K were calculated using the appropriate functions in the liquid state, enthalpies and entropy of vaporization and the entropy of the ideal gas compression,   ln{ ( ) /(101.325) } SR p T k P a calculated from the vapor pressure data.

Theoretical calculations of the thermodynamic functions
The ideal gas absolute entropy and heat capacity were calculated by statistical thermodynamics, additive principle (Poling et al., 2001;Domalski et al., 1993;Sabbe et al., 2008), and empirical difference method of group equations (Cohen & Benson, 1993). The statistical thermodynamic method was used with quantum mechanical [QM] calculation on the basis of the density functional theory [DFT]. The QM calculation was performed on the level B3LYP/6-31G(d,p) using the Gaussian 98 and 03 software packages (Frisch et al., 2003). As a result, the following constants can be calculated: the moments of inertia of the entire molecule, the moments of inertia for internal rotors, the normal vibrational frequencies, and the barrier to internal rotation. The potential functions of internal rotation were determined by scanning the torsion angles from (0 to 360) o at 10 o increments and allowing all other structural parameters to be optimized at the same level with the subsequent frequency calculation. The calculated potential energies were fitted to the cosine-based Fourier series: where ) ( V denotes potential energy function,  is torsional angle. The ideal gas entropies and heat capacities in dependence on the temperature were calculated by standard statistical thermodynamics formulae using the rigid-rotor harmonic oscillator [RRHO] approximation. To account for the internal rotation processes, the torsional frequencies were omitted in the calculation of thermodynamic function. A contribution of the internal rotation for each rotor was calculated by direct summation over the energy levels obtained by diagonalization of the one-dimensional Hamiltonian matrix associated with potential function from equation (20). The RRHO approximation, is known, results in overestimated entropy values for flexible molecules due to coupling the internal rotations. One-dimensional hindered rotor correction has been applied by (Vansteenkiste et al., 2003;Van Speybroeck et al., 2000) assuming decoupled internal rotations. The method of group equation is suitable for calculation of some additive properties of a compound, namely  () The calculation of the absolute entropy by additive methods requires taking into account corrections for the symmetry and the optical isomerism of the molecule. Otherwise, the principle of group additivity can be broken if these parameters alter when changing the structure of the molecules in the series of compounds. Therefore, an additive calculation of the absolute entropy, 0 m S , is conducted by using so-called intrinsic entropy, 0 m,int S , which allows to exclude an influence of the rotary components, depending on the symmetry and optical activity of the molecule: , are applied for prediction of the missing data, verification of their reliability, and mutual congruence in the series of the same type compounds, or homologous. Below, a critical analysis of the ideal gas entropies is performed for some series of functional organic compounds. Freons and cloroalkanes. Table 7 -122]. A characteristic feature of these compounds is availability of residual entropies caused by orientational or conformational disorders. The mixtures of the trans-and two gauche-conformers of these freons were probably frozen at low helium temperatures. The entropy change caused by disorder of this type can be evaluated by the formula , where 2 N and www.intechopen.com 1 N are the number of states of disordered and ordered phases. Thus, the residual entropy of CFCl 2 CHFCl and CF 2 ClCHCl 2 is ) 0 ( o m S = R·ln(3) = 9.1 J·K -1 ·mol -1 . The value ) 0 ( o m S = 10.1 J·K -1 ·mol -1 of CF 2 ClCFCl 2 , obtained in (Higgins & Lielmers, 1965), is in good agreement with the R ·ln(3) value. After taking into account residual entropies of these freons, the calorimetric and theoretical values of absolute entropies agree within errors limits from (0.1 to 1.3) %. At the same time, testing the low-temperature , C p m values of CFCl 2 CHFCl and CF 2 ClCHCl 2 by equation (19) showed an absence of the residual entropies for these freons. Their heat capacities in the temperature interval from (5 to 8) K obeyed Debye cub's law with experimental error of , C p m value, 2 %. A disagreement of two methods evaluation of the residual entropy can not be explained on the basis of available physico-chemical data of these freons.   (Frenkel et al., 1994). Cyclic hydrocarbons and perfluorocarbons. The absolute entropies obtained by the third thermodynamic law from the experimental data and statistical thermodynamics for some cyclic and bicyclic hydrocarbons and perfluorocarbons are listed in Table 8. The values of the absolute entropies determined by independent methods agree within errors limits from (0.1 to 1) % that proves their mutual conformity and also the reliability of the all experimental data used for computing the value of ( )( ) o Sg m e x p t , namely the heat capacities, enthalpies of vaporization, and saturation vapor pressures. Ferrocene derivatives [FD]. The experimental data on thermodynamics of the phase equilibriums for FD are very scarce. Therefore, critical analysis of the available data on thermodynamic properties and using them for science prognosis of failing properties are the urgent problems. A verification of reliability and mutual congruence of the properties of some derivatives of ferrocene were carried out on the basis of absolute entropies, which were computed by the third thermodynamic law from the experimental data and by statistical thermodynamics and group equation method. Comprising the o S m and  C p,m values determined by both theoretical methods showed that they agree within errors limits,  1.2 %. The values of absolute entropies in homologous series of alkyl-and acyl-ferrocenes obtained in this work and available in the literature are presented in Table 9.  (Emel'yanenko et al., 2010); e (Karyakin, et al., 2003). values for ferrocene, n-propyl-and iso-butylferrocene agree to within their errors that proves the reliability of these values and all the thermodynamic data used for their calculation. The mutual consistency of the absolute entropies of iso-butyrylferrocene and iso-butylferrocene in the liquid state was revealed using the Benson additive method. The difference between the entropies of these compounds, S= 2 J·K -1 ·mol -1 , is close to the entropy increment   S = 1.4 J·K -1 ·mol -1 , which fit for replacing the СН 2 by CO group in the passage from iso-butyl-to iso-butyrylferrocene.  values and can be explained by errors in calorimetric heat capacities, obtained in (Karyakin, et al., 2003). After excluding of this value, the coefficient R 2 of linear correlation increased from 0.9147 to 0.9992, and the  s value decreased to 4 J·K -1 ·mol -1 . Thus, the value Correlations for the absolute entropies of acylferrocenes obtained for the condensed and ideal gas states (Table 9) are shown in Fig. 13 butylferrocene, acetylferrocene, propionylferrocene, and iso-butyrylferocene are reliable to within errors. The entropies of ethylferrocene in the condensed state and n-butylferrocene and formylferrocene in the ideal gas states are not reliable because of errors in EF heat capacity measurements and errors of the vapor pressures and enthalpy of vaporization and sublimation for n-BF and FF, respectively. On the basis of the critical analyses of the absolute entropies of the ferrocene derivatives, the recommended reliable ( )( ) o S g recom m values have been presented in Table 9.

Extrapolation of vapor pressure to entire range of liquid phase
A saturation vapor pressure of substances for entire liquid region was obtained by calculation methods on the basis of the experimental pT data of moderate "atmospheric range" from 2 to 101.6 kPa. Extending the pT parameters to the region of low pressures is performed by simultaneous processing the vapor pressure and differences between the heat capacities of the ideal gas and liquid. The vapor pressures above 100 kPa are calculated by the empirical bimodal equation obtained by processing the pT parameters and densities of liquids using the one-parameter corresponding states law in Filippov's version. The extrapolation of vapor pressure to the low-temperature region down to the triple point is carried out by a system of two equations: where <p> is the vapor pressure at the average temperature of experimental range, and A', B', C', and D' are coefficients of simultaneous processing of the pT data with  , C p m differences. Mathematical processing by a system of equations (23) requires the use of heat capacities of the ideal gas and liquid in the neighbourhood of the triple point, where the linear temperature dependence of , C p m is valid.

www.intechopen.com
The method of extrapolation of the vapor pressure was tested with standard substances 1,1,1-trifluoro-2,2-dichloroethane and n-decane for which precision pT parameters and heat capacities of the ideal gas and liquid are available in a wide temperature range.  (Boublik et al., 1984) and most reliable calculated values (Ruzicka & Majer, 1994) were evaluated using equation (4) (curves 2 and 3) and the system of equations (23) (curve 1). Curves 2 and 3 were obtained for the complete and shortened temperature ranges of the pT data T = (80 and 50) K, respectively. In so doing, the deviation of the calculated p values from experimental ones,  s p , increase for equation (4) from (5 to 40) % in dependence on the temperature range of extrapolation , T extr , from (50 to 120) K. The appropriate deviations for the system of equations (23) are much smaller and equal from (2 to 4) % for this range of extrapolation. The results of approximation of pT data for l,l,l-trifluoro-2,2-dichloroethane (Weber, 1992) by individual equation (4) and by system of equations (23) are compared in Table 10 . The data were processed by the least squares method using the vapor pressure in the temperature range from 256 to 299 K and the heat capacities of the ideal gas (Frenkel et al., 1994) and liquid  in the range from 150 to 240 K. In so doing, the temperature range of extrapolation of vapor pressure from initial temperature 256 K to T tp = 145.68 K equals T extr = 111 K. Tentative errors of approximation coefficients were evaluated only for revealing the change of extrapolation prediction in going from equation (4) to that of (23). As is seen from Table 10, errors of coefficients of equations (23) are approximately ten times lower than those of equation (4), which explains increasing the extrapolation capabilities of the system (23). The resultant value of the vapor pressure at triple point p tp is about 25% lower than that of equation (4). Reducing the temperature range of initial pT data, T  , from (25 to 50) K (accordingly, extending the range of extrapolation of vapor pressure to ~ 130 K) leads to the variation of p tp value by only 4-6%.  (4) and (23) and saturated vapor pressure of 1,1,1trifluoro-2,2-dichloroethane at the triple point, T tp = 145.68±0.02 K (Weber, 1992) Analysis of calculated data for n-decane and l,l,l-trifluoro-2,2-dichloroethane revealed that the use of the system of equations (23) for extrapolation of the vapor pressure in a wide temperature range   T extr (120 to130) K enabled one to obtain () p T tp values with errors of  10%. Extending the saturation vapor pressure to the critical region and computing the critical parameters of the substances were performed by the corresponding states law of Filippov's version (Filippov, 1988 Few literature data are available for critical parameters of organic compounds. The law of corresponding states in Filippov's version enables one both to obtain the required critical parameters using precise pT data and density  of liquid and to calculate numerous thermophysical properties on the basis of known values of , T c , V c and  A F . Generalized equations for the calculation of the critical parameters and of the thermophysical properties were obtained by empirical method using the array of experimental data on the appropriate properties for studied compounds (Filippov, 1988). The critical parameters were calculated by algorithms (Filippov, 1988) and (Varushchenko et. al., 1987), in which the following pairs of input data were used: The derivative ln( ) / (1 / ) dpd T was determined by differentiation of the equations  ln( ) ( ) p FT for investigated substances. A random sampling of 14 compounds, having reliable data on the vapor pressure, density, and critical parameters was used to demonstrate (Varushchenko et. al., 1987) that 1) the employed algorithm is capable to produce adequate results in computing , T c , V c and  A F and 2) the calculation algorithm originally developed for compounds with the similarity criterion 1    A F 4 is suitable for a wider class of compounds, including polyatomic molecules with the criterion 0.5   A F 4.
A prediction of the thermophysical properties by the OLCS of Filippov (Filippov, 1988) requires following basic quantities: , T c , V c and  A F . The errors of the latter calculation are within ±1-2%; the appropriate errors of , P c are higher and amount to ±3-5%. The saturation vapor pressure are extrapolated to the critical region using the empirical binodal equation with pseudocritical parameters   TT c and   PP c : where a = 3.9726, b= 0.3252, and c = 0.40529. Two-parameter formula (25) weakly depends on similarity criterion  A F ; therefore, it may be employed for curvilinear extrapolation of saturated vapor pressure using only two pairs of the pT data. The formula was derived by the similarity conversion, namely using superposition of curves    log() /l o g() which have the same slope for different substances. As a result, the family of curves reduces to a single curve which has segments of vapor pressure in the supercritical region. The pair of the    PT parameters is of importance for remote point on the binodal curve, which enables one to calculate the vapor pressure of a substance in the region from the normal boiling temperature to the critical point. The error of the pressure calculation is  (3 to 5) % depending on the temperature range of pT data extending and their reliabi lity. Therefore, analysis of equations (4) and (23) demonstrated that equation (4) obtained by approximation of vapor pressure of the "atmospheric range" gives precision results when it used as an interpolation equation. It is further employed for extrapolation of the vapor pressure in the temperature range of  50 K with an error of (1 to 2) %. Extending the vapor pressure to the entire liquid phase from the critical to the triple points temperatures is performed by means of one-parameter law of corresponding states with errors from (3 to 5) % and, respectively, by simultaneous processing of the pT parameters and the differences between low-temperature heat capacities   00 () ( ) ,, , CC g C l i q pm pm pm with uncertainties of  10 %, respectively. The data on extrapolation of the vapor pressure are suitable for many technological calculations.

Conclusion
The fundamental thermodynamic investigations of the phase equilibriums of many functional organic compounds were carried out by experimental and calculation methods in Luginin's Thermochemistry Laboratory of the Moscow State University. Modified setups have been created for precise determination of the saturation vapor pressures by www.intechopen.com comparative ebulliometry, the enthalpies of vaporization by evaporation calorimetry, and the low temperature heat capacity and phase transitions by vacuum adiabatic calorimetry. The saturation vapor pressures were determined in moderate range of pressure 2  (p/kPa)  101.6 with accuracy of the temperature   0.01 K, and pressure,   26 Pa, which correspond to the modern precision levels. The temperature dependences of the saturation vapor pressure,  ln( ) ( ) p FT , and the enthalpies of vaporization   () H f T vap were obtained by mathematical processing of the pT parameters by equation (1), derived on the basis of Clapeyron equation using LSM with orthogonal functions. The latter allow one to calculate the errors of H vap values, which are urgent problem because the indirect method is the main source for determination of the enthalpies of vaporization. An agreement of the H vap values obtained by direct (calorimetric) and indirect (calculation) methods proves their reliability. The precise saturated vapor pressure data are extended to entire region of the liquids under study. Extrapolation of the pT parameters down to the triple point temperature are carried out by simultaneous processing the vapor pressures and low-temperature differences   00 () ( ) ,, , CC g C l i q pm pm pm , which are the second derivatives of the vapor pressure upon the temperature. Extending the pT parameters to the critical region and calculation of the critical quantities are performed by Filippov's one-parameter law of the corresponding states. The latter enables us both to calculate the critical parameters on the basis of more readily available pT -data and density of liquids and to predict numerous thermo-physical properties of the equilibrium liquid -vapor by means of the known critical quantities , T c , V c and criterion of similarity  A F . The low temperature heat capacity in the temperature region (5 to 373) K, molecular motion in crystal and metastable phases, and solid state transitions and fusion were investigated by the adiabatic calorimetry. Uncertainties of the , C p m measurements are on the average ~0.2 % which correspond up-to-date precision level. An accurate calorimetric study of the solid states of the functional organic compounds revealed different polymorphic modifications of the molecules, order -disorder transitions involving orientational and conformational disorder, glass-like transitions, and plastic crystalline phases with anisotropic and isotropic reorientations of the molecules. For interpretation of these transformations, the X-ray crystallography, infrared and Raman molecular spectroscopy were got involved in investigation.
The main thermodynamic functions in three aggregate states: the absolute entropy by the 3 d law of thermodynamics, the changes of the enthalpy and free Gibbs energy are derived on the basis of the heat capacity and vapor pressure measurements. A critical analysis and verification of the reliability of obtained data are very significant parts of the thermodynamic investigation. With this in mind, the experimental thermodynamic functions are compared with calculated ones by additive principles and by statistical thermodynamics coupled with quantum mechanical (QM) calculation on the basis of DFT method. The QM calculation are performed on the level B3LYP/6-31G(d,p) by Gaussian 98 and 03 software packages. A qualitative analysis of thermodynamic properties in dependence on some parameters responsible for intermolecular interactions and short range order of the liquid phase has been carried out for verification of the mutual consistency of the properties in homologous series and the series of the same type of compounds. A quantitative verification of the The agreement between these quantities confirms their reliability, as well as all experimental and calculation data used in computing process.

Acknowledgment
We are grateful to Professor O. Dorofeeva for providing Gaussian programs and assistance in quantum-chemical calculations of the ideal gas thermodynamic functions. Many thanks to Professor S. Verevkin for helping in determination of the vapor pressures of some derivatives of ferrocene and Dr. L. Pashchenko for taking part in determination of the vapor pressures of some freons. Special thanks are to Post-graduate student E.