The BN Phase Diagram

In 1963 a phase diagram was established by Bundy and Wentorf [19], based on data of Wentorf [42] and experiments carried out at pressures higher than 4 GPa. This phase diagram described c-BN as the stable phase at standard temperature and pressure (Fig. 6). In 1975 a new phase diagram was published by Corrigan and Bundy [11], showing the c-BN/h-BN equilibrium line similar to the graphite/diamond line in the carbon system. 

The phase diagram of Corrigan and Bundy had been considered to be correct until 1987 when Leonidov et al. [43] published fluoro-calorimetric results for burning c-BN, and additional calculations from Solozhenko and Leonidov [44] followed in 1988. These papers described c-BN as the stable phase - up to 1300 °C. A comparison of the data for burning c-BN and h-BN confirm the c-BN stability  

Further results were reported by Maki et al. in 1991 [46] and by Solozhenko in 1993 [47]. However, the thermodynamic data show large discrepancies (Table 2 [44, 47], Table 3 [43, 45, 46, 48-50]) and therefore the differences in the phase diagrams are easy to explain. 

It can be summarized that c-BN is the stable phase at room temperature but there are still discrepancies about the phase transformation lines and thermodynamic data. Several review articles draw conclusions from the available results (Fig. 7) [11, 41, 51-54] but there are also new results describing c-BN as metastable at standard conditions [55]. 

---

Parts prepared of h-BN as well as c-BN are of great interest for industrial applications but also for materials science. The thermodynamic data for c-BN and the BN-phase diagrams found in literature are not in agreement. After the first high pressure experiments the B-N phase diagram was designed, and after some modifications c-BN was described as metastable phase at room temperature. Contrary to this opinion in 1988 it was reported that c-BN is the stable phase. Many experiments have confirmed this result, but exact thermodynamic data are still not available. 

A rather interesting discovery was made by Kobayashi et al. [174]. At 1470-1870 K and 4.2 GPa, c-BN was transformed into h-BN. According to the earlier BN phase diagrams this should not have happened, because these conditions are located in the region where c-BN should be stable. 

Suenaga et al. (1999) [16] produced nanoparticles and nanotubes by arc discharging and found by means of electron energy loss spectroscopy (EELS) that such materials consist of immiscible BN and C-layers. The growth mechanism was explained by using the calculated phase diagram sections in the B-C-N system according to [33, 244]. 

E-BN (E = explosion) is described as high pressure phase by a few scientists. For synthesis shock wave methods [25, 26] were used and also reactions at normal pressure with photon [27] or electron [28, 29] assistance. In a special three-dimensional phase-diagram (pressure, temperature, electrical field) the existence of the metastable E-BN was described [30]. 

Fig. 4.4. Schematic diagram to illustrate the growth process of the ArBs layer in the A-AiBn reaction couple due to the counter diffusion of both components and the partial decomposition of the A/Bn phase at interface 3.

The phase diagram of BN has been extended to 14 kbar and 4000 Apparently the wurtzite-like phase is stable at high pressures and low temperatures. 

Solozhenko (1988) [176] concluded that kinetic factors influenced the transformation significantly and that the true equilibrium line can only be determined by a thermodynamic approach. From data on heat capacities [177-189], relative enthalpies (heat contents) [190-199], enthalpies of formation data [177, 201-207], equations of state and thermal expansion data for all BN modifications, Solozhenko [176] derived a calculated new phase diagram which significantly differed from the Corrigan and Bimdy (1975) [174] version. An overview of somces of thermodynamic data is given in Table 15. A review of calorimetric studies was given by Gavrichev et al. (1994) [212]. Vaporization studies of boron nitride were made by [208-211]. 

Experimental work on the phase diagram of this system is meager. Some information on selected phase equilibria can be derived indirectly from work on the development of commercial materials in the Si-B-N system. Gugel et al. (1972) [223] suggested a tentative phase equilibrium diagram. Equilibria between silicon Sorides and silicon nitride were assumed. According to this result, silicon does not show an equilibrium with BN. Kato et al. [227]... 

The ternary system was calculated by extrapolation from the binary subsystems (Kasper, 1996) [33]. The calculations cover phase equilibria at one bar and do not assume any solubilities as no experimental evidence for stable sohd solutions between B4+5C and BN or a-BN and graphite exist. The section between graphite and boron nitride including the invariant reactions Uj, Dj and U2 (Fig. 23) is shown in Fig. 22. A calculated potential phase diagram (logpN2-T) can be found in [244], The complete Scheil reaction scheme (P = 1 bar) is shown in Fig. 23. 

In section two of tUs paper we describe the technical aspects of C02-laser heating in a DAC. The third section focuses on the methods for measuring melting temperatures at variable pressures, the fourth section on the determination of high pressure and temperature phase diagrams, and in the fifth section some experiments focusing on the synthesis of diamond and cubic BN from organic precursors will be described. 

Figure 10. P,T phase diagram of boron nitride (a) catalytic conversion of hBN intocBN (b) metastable region for the vapor phase formation of cBN. An alternative phase diagram of BN suggested previously [111] is shown by dashed lines.

No boron-nitrogen phase diagram is presendy available but a tentative pressure-temperature diagram is shown in Fig. 12.1.1 1 This diagram suggests that c-BN (widi the zincblende structure) is stable below 225°C while h-BN is stable above 225 C and melts at approximately 3000°C. A second cubic structure (with the wurtzite structure) is stable only at pressures >11 GPa. 

Very little is known about the equilibrium pressure-temperature phase diagram of BN (23,66) because of the retardation of the transformation at low temperatures and difficulty in carrying out high-temperature experiments at high pressures. Tentative equilibrium phase boundaries are shown by the dashed lines in Fig. 6. 

Our phase diagram is meant to apply to a type I superconductor. The letters s and n label the superconducting and the normal conducting phases, respectively. In the s-region we have therefore Bs = 0. If in addition we use the linear relation Bn = iJ-rH, where fir is the magnetic permeability, then Eq.(4.51)... 

According to the data of [1986Smi] one can assume that BN-Fe and BN-FeB could be the quasibinary sections but their phase diagrams are not investigated. 

FIGURE 18.6 Calculated partial pressure diagram for Si-B-C-N ceramics. The calculations predict that SiC/SijN4/BN/C composites behave thermally like boron-free SiC/SijNyC materials (cf. Figure 18.3). Boron-containing phases (here BN) do not directly influence the thermal stability of Si3N4. ... 

The shift in the order-order boundaries is easily explained by the conformational asymmetry. The argument is most transparent for layers. In these diagrams, bs < a, which implies that the A block is naturally more extended than the B block. If the diblocks have equal volumes,/a =/b, then, in the lamellar morphology, the A and B subdomains must have the same thickness. However, in this case of bn < bA, this means that the B block is stretched beyond its unperturbed Rg more than the A block. A corollary to this is that equal stretching will occur for copolymers with > /b-This means that the tendency for these polymers to be driven toward a B-centered morphology is reduced by this conformational asymmetry. Hence, the L/C phase boundary moves toward greater /a. Similar arguments apply to the other order-order phase boundaries, which all shift in the same direction. 

---