The C, H and N analyses were performed on a CARBO-ERBA-11008 rapid elemental analyzer. IR spectra were recorded as KBr pellets on a Perkin-Elmer FT-IR spectrometer (RX1, FT-IR) in the region of 4000 - 450 cm⁻¹. Fast atom bombardment mass spectra (FABMS) of the nickel complexes were recorded by employing a JEOL SX 102/DA-6000 spectrometer using argon as the FAB gas with an accelerating voltage of 10 kV and the spectra were recorded at room temperature. M-nitrobenzyl alcohol (NBA) was used as the matrix unless specified otherwise. The thermal analyses were carried out using Perkin-Elmer, Pyris Diamond TG/DTA at a linear heating rate of 0.17 Ks⁻¹. High purity nitrogen (purity > 99.99%) dried by passing through refrigerated molecular sieves (Linde 4A) was used as the purge gas at a flow rate of 12 dm³∙h⁻¹.

A horizontal thermal analyzer was adapted as a transpiration setup for vapor pressure measurements. The configuration of the horizontal dual arm with a single narrow furnace chamber minimizes errors arising from convection, buoyancy, thermo molecular and electrostatic charge effects. The arms of the thermo balance served as the temperature-cum-DTA sensors.

The block diagram of the thermal analyzer, modification for its functioning in the transpiration mode, including precise flow calibration for the carrier gas using a capillary glass flow meter and corrections for apparent weight losses in isothermal mode were the same as reported [15] [16] . The choice of 6 dm³∙h⁻¹ for N₂ gas was made for the isothermal equilibrium vaporization on the basis of the existence of a plateau in the plot of the equilibrium vapor pressure against the flow rate. The vapor pressure measurements were carried out by rapid heating 0.17 Ks⁻¹ and after allowing for temperature stabilization, subsequent changes in isothermal steps were done at a heating rate of 0.03 Ks⁻¹. It turned out that the vapor pressure derived from the TG-based transpiration method was reliable to 10% accuracy [15] [16] .

Experiments were performed under non-isothermal conditions at a programmed linear heating rate of 10˚C∙min⁻¹ for methyl to pentyl homologues. Among the various methods for the kinetics evaluation of TG weight loss data, Arrhenius method was followed in the present investigation to study the vaporization kinetics.

The elemental analyses and their composition are presented in Table 1. The melting point endotherms (Figure 1) were calculated accurately using the built-in Pyris software and the values 280˚C, 271˚C, 205˚C, 176˚C and 175˚C for methyl to pentyl homologues respectively which exhibited a decreasing trend with increasing length of the alkyl chain. The non-isothermal TG curves (Figure 2) of bis(2-hydroxonapthaldehydato)nickel(II) indicated a 37.8% residue at 700 K making it unsuitable for CVD applications. The TG curves of bis(N-alkyl-2-hy- droxonapthaldimine)nickel(II) homologues (where N-alkyl = methyl to pentyl) showed a single step weight loss commencing from 700 K, leading to a negligible amount of residue qualifying themselves as precursors for CVD.

The relevant frequencies such as, , and (Table 2) support the struc-

ture. The FABMS patterns (Table 3) of bis[N-alkyl-2-hydroxonapthaldimine)]nickel(II) (where N-alkyl = methyl to pentyl) complexes revealed the molecular masses (M) of the complexes and their fragmentation pattern. The values of molecular ion peaks were observed at (Da) = 430, 454, 483, 510 and 539 respectively, for methyl to pentyl homologues as clusters due to the isotopic abundances of natural nickel (⁵⁸Ni: 67.76%, ⁶⁰Ni: 26.16%, ⁶¹Ni: 1.25%, ⁶²Ni: 3.66%, ⁶⁴Ni: 1.16%).

Effective control of the process of any type of CVD and to monitor the composition, thickness and microstructure of thin films, relevant information on precursor chemistry, possible fragmentation pattern by the decomposition of precursors and vapor pressures are needed. Therefore vapor pressure measurement was deemed as essential for these complexes to be used for MOCVD.

If W is the mass loss of the sample (Table 4) at the isothermal temperature caused by the flow of dm³ of the carrier gas (measured at), the vapor pressure could be calculated using Dalton’s law of partial pressure for ideal gas mixtures as

where is the molar mass of the complexes, and are the temperature and volume of the carrier gas respectively. The molecular mass of the congruently vaporizing species was obtained from FABMS. The observed mass loss and the calculated (using Equation (1)) (Table 4) and versus 1000/T (K) are shown in Figure 3. The values of slope and intercept of the Clausius-Clapeyron equation obtained from the plot for the nickel homologues (Table 5) along with the values of enthalpy of vaporization and entropy of vaporization are given.

Enthalpy of vaporization is obtained by multiplying the slope in Equation (2) with −2.303 R. The least- square expressions from the plots are included in Table 5 along with the temperature ranges of their experimental measurements employing the TG-based transpiration technique. The values as calculated from the slope of the curve were found to be 153.1 (±1.9), 122.9 (±0.3), 147.6 (±10.7), 151.8 (±10.9) and 114.7 (±5.3) k∙Jmol⁻¹

respectively. The entropies of vaporization for these complexes as calculated from the intercept of the

linear fit expression were found to be 319.7 (±3.9), 229.9 (±5.8), 317.8 (±17.2), 319.7 (±19.1) and 254.6 (±9.6) Jmol⁻¹∙K⁻¹. The vapor pressure of the complexes will be helpful for fixing the metal organic chemical vapor deposition (MOCVD) process parameters for getting the desired phase and rate of deposition of nickel and composite materials.

The rate constant for the vaporization enthalpy of the complexes was determined in the temperature range of 400 - 680 K for every 10% weight loss of the complex. The expression (3) for is given by

where is the derivative of the fraction vaporized with respect to time and is the rate constant of va- porization. For every 10% weight loss, was calculated by the expression (4) as

where is the per cent weight at any time and and respectively, are the initial and final percent sample weights [17] . The Arrhenius expression (5) is

And the plot of versus 1000/T (K) (Figure 4) is found to be linear. From the slope, the activation energy for the vaporization of the complexes was calculated. The activation energy values were found to be 106 ± 4, 111 ± 6, 114 ± 7, 115 ± 7 and 121 ± 4 k∙Jmol⁻¹ respectively for methyl to pentyl homologues.

The kinetics of the complexes was followed by employing the Coats-Redfern approximation which gives the expression (6).

A plot of versus 1000/T (K) gives (Figure 5) a straight line when the correct func-

tion is used in the equation. The function describes the mechanism of the reaction [18] . Straight lines with high-correlation coefficient and low standard deviation were selected to represent the possible controlling mechanism. The corresponding kinetic parameters were then calculated and are shown in Table 6. The best fit for the methyl complex is obtained using A3 Avrami-Erofe’ev Equation (3). For the ethyl and pentyl complexes, best fit was obtained with R2, contracting area model. For the propyl and butyl complexes, best fit was obtained with A2, two dimensional Avrami-Erofe’ev Equations (2) model. The activation energy values were found to be 112 (±5), 115 (±5), 105 (±6), 108 (±10) and 126 (±5) k∙Jmol⁻¹ respectively for methyl to pentyl homologues. Analysis of these data show that the activation energies which are in good agreement with that obtained using Arrhenius’s method.