Metastable phases in thin films and nanolayered materials

Although there is a great body of literature dealing with bulk solid-state phase transformations in different materials, considerably less information is available on the same subject in thin films. One interesting topic related to phase transformations is the formation of metastable phases in thin films. Metastable phases can be formed in thin films whereas they are hardly seen in bulk counterparts. Metastable phases can be formed either in as-deposited thin films (deposited by physical vapor deposition) or after heating immiscible nanolayered thin films.

I. Metastable phases formed during physical vapor deposition.

Bulk austenitic stainless steels (SS) have stable face-centered-cubic (fcc) structure. However, sputter deposited films synthesized using austenitic SS targets usually exhibit metastable body-centered-cubic (bcc) structure or a mixture of fcc and bcc phases. It was argued that rapid quenching, from liquid to solid or from vapor to solid, results in the formation of metastable bcc SS. However, our studies suggest that chemistry will play an important role in the phase transformations.

It was shown in the literature that bcc 304 SS thin films annealed at 700^oC/1h will completely transform to the stable fcc 304 SS. However, after annealing of our bcc 304 SS films at 800^oC/1h, fcc phase remains as shown by XRD and TEM.

It is known that 304 SS has the minimum Ni content needed (8wt%) to stabilize single fcc phase at room temperature. A slight loss of Ni (during magnetron sputtering) could push the alloy into a two-phase region at room temperature. Our as-sputtered 304 SS films, although depleted with Ni (only 7wt%), could presumably maintain a single metastable bcc phase due to rapid quenching effect. This metastable phase will transform to an equilibrium state, the mixture of fcc and bcc phases after elevated temperature annealing.

Our studies shows that sputtered 330 SS (with 36wt% Ni) has a stable fcc structure. Incorporating literature data, we obtained an empirical plot showing the phases observed in sputtered austenitic SS films as a function of substrate temperature during deposition and Ni concentration of the target. It clearly shows that the formation of single fcc phase in 304 SS (8wt% Ni) needs a very high substrate temperature, > 500^oC, while fcc single phase can be formed at room temperature in as-sputtered 330 SS (36wt%) films. The situation is somewhat mixed in the cases where Ni concentration is between that in 304 SS and 330 SS. Two straight lines are drawn in this figure to show possible boundaries that separate the three regions: bcc, mixture of bcc and fcc, and fcc phases. These studies strongly suggest that the composition of Ni and substrate temperature during deposition are the two dominating factors that determine the phase formation in as-sputtered SS films.

Empirical plot of the phases obtained in sputtered austenitic stainless steel (304SS - 8wt% Ni, 316SS – 13 wt% Ni, 330SS – 36wt% from this study, 310SS – 20wt% Ni and others) as a function of temperature and Ni concentration.

II. Metastable phases formed after heating thin films.

Metastable phases have been observed in nanolayered thin films where the adjacent constituents are immiscible. Nanolayered thin films provide considerable interfacial energies associated surfaces and interfaces and a greater abundance of heterogeneous nucleation sites, such as surfaces and interfaces. Very often a higher diffusivity is achieved in thin film which will expedite phase transformations. The kinetics of phase transformations will also be influenced due to the large amount of surface and interfaces in thin films.

Metastable phases can also be formed as a result of epitaxial growth. For instance bcc Cu can be stabilized with Mo under layers. There are a large number of literatures in this aspect.

References

“Critical Factors that Determine Face-centere-cubic to Body-centered-cubicPhase Transformations in Sputter-deposited Austenitic Stainless Steel Films”, Journal of Materials Research, 19 (2004) 1696. (PDF)