J. W. Matthews

Evidence for pseudomorphic growth of iron on copper

Abstract Iron was deposited at room temperature in ultra-high vacuum (5 × 10−8 Torr) onto single-crystal films of copper oriented with (001) parallel to their plane. A shutter was moved between source and substrate so that iron deposits in which there were thickness gradients were obtained. Transmission electron micrographs and diffraction patterns from the thin (less than about 20 A) parts of the iron deposits showed that they were face-centred cubic and were strained to exactly match the copper substrates. The thicker parts of the deposits contained dislocations to accommodate part of the misfit between face-centred cubic iron and copper. They also contained small nuclei of body-centred cubic iron.

Pseudomorphic deposits of cobalt on copper

Abstract Cobalt was deposited in ultra-high vacuum onto (001) copper surfaces prepared inside the vacuum chamber. If the copper substrate was at room temperature during film growth then the cobalt grew approximately as a monolayer. The structure and lattice parameter of the cobalt remained the same as the copper until its thickness reached about 20 A. At this thickness long dislocations to accommodate part of the difference between the lattice parameters of cobalt and copper were generated. Many of these dislocations were unusual in that they were imperfect. The stacking faults associated with them converted a little of the f.c.c. cobalt into the stable h.c.p. structure. If the copper was hot (350°C) during the deposition of cobalt, then film growth began with the formation of three-dimensional nuclei. Nuclei less than about 375 A in radius were f.c.c. and were strained to match the copper lattice.

Pseudomorphic growth of iron on hot copper

Abstract An iron deposit in which there was a thickness gradient was grown in ultra-high vacuum (< 10−7 torr) on a hot (400°C) copper surface prepared inside the vacuum chamber. Electron micrographs of the iron-copper bicrystal revealed that the growth of iron began with the generation of three-dimensional iron nuclei. These nuclei were f.c.c. and their lattices were strained to give a coherent copper-iron interface. The nuclei grew in size as the thickness of the iron increased and dislocations to accommodate part of the misfit between γ-iron and copper were generated. The dislocations were in mixed orientation and had Burgers vectors which were inclined to the iron-copper interface. A further increase in deposit thickness was accompanied by the nucleation of b.c.c. (α) iron.

A mechanism for the formation of twins in evaporated face-centred cubic metal films

Abstract Face-centred cubic metal films that are prepared by evaporation onto sodium chloride contain numerous micro-twins. It is suggested that these twins are formed when two nuclei, which happen to be approximately twins of one another, coalesce and rotate into precisely twin relationship. The rotation takes place in a manner similar to the rotation of nuclei into parallel alignment that was observed by Bassett. It is shown that this mechanism is consistent with the mode of growth of films, and with the size, shape and arrangement of the twins. The mechanism is also able to account for the formation of twins early in film growth, for the rarity of twins of twins, and for the presence of the {112} orientation observed by Gottsche and Kehoe.

Defects in silver films prepared by evaporation of the metal onto mica

Abstract Silver deposits, with the (111) silver plane parallel to the plane of the deposit, were prepared by evaporation of the metal on to the cleavage face of hot muscovite mica. A number of films of different thicknesses were examined in a transmission electron microscope and were found to contain dislocations, plane stacking faults, bent stacking faults, tetrahedra of stacking fault, and twins whose twinning plane was parallel to (111). The lateral boundaries to the twins was parallel to the {211} planes. Twin boundaries parallel to (111) often contained dislocations, some of which may have been twinning dislocations. Mechanisms by which these defects could have been formed during film growth are described.

A study of the F.C.C. to B.C.C. transformation in films of iron on nickel

Abstract Iron was deposited in high vacuum (10−7 to 10−8torr) onto (001) nickel surfaces prepared inside the vacuum chamber. It was found that continuous iron films were formed early in film growth, and that these films were f.c.c. and strained to match the nickel lattice. When the thickness of the iron deposits reached 15±3 A small elongated nuclei of b.c.c. iron appeared. Those nuclei were in the orientations described by Pitsch (1959) and grew in size as film growth proceeded. Misfit dislocations were not observed at any stage in the growth of 150±204 A iron films. It is suggested that they were absent because much of the misfit between f.c.c. iron and nickel was accommodated by the lattice distortion associated with the formation of b.c.c. grains in the Pitsch orientation.

Pseudomorphic deposits of chromium on nickel

Abstract Chromium films were prepared by evaporation of the metal in ultra-high vacuum onto the (001) surfaces of nickel. A shutter was moved between source and substrate during film growth so that films in which there were thickness gradients were obtained. The very thin (roughly 10 A) parts of the chromium deposits were face-centred cubic, and dislocations to accommodate part of the misfit between the lattices of nickel and f.c.c. chromium were present. Thicker deposits contained small nuclei of chromium in the stable b.c.c. structure.

The climb and glide of misfit dislocations

Abstract Direct observations of dislocations that accommodate the misfit between crystals with slightly different lattice parameters have recently been made (Matthews 1961, Delavignette et al. 1961) using the transmission electron microscope technique (Hirsch et al. 1960). In one of these (Matthews 1961) the misfit dislocations lay in an (001) interface between single crystal films of PbS and PbSe. The dislocations were in edge orientation and had +a(llO) Burgers vectors which lay in (001).

Dislocations in evaporated lead sulphide films

Abstract Single-crystal films of PbS, oriented with (001) parallel to the film plane, were prepared by evaporation into rocksalt. They were examined in a transmission electron microscope and were found to contain dislocations of three kinds. (i) Screw dislocations with long lines approximately parallel to the ⟨110⟩ directions in the film plane. (ii) Dislocations that took the shortest route through the films and appeared as dots. The majority of these had 1/2a⟨110⟩ Burgers vectors and the Burgers vectors of the remainder were a⟨100⟩. A large fraction of the Burgers vectors lay in the film plane. Some of the dislocations were arranged in low-angle grain boundaries. (iii) Dislocations with lines whose length and orientation suggested that they lay in {110} planos inclined at 45° to (001). Motion of dislocations that corresponded to the operation of the ⟨110⟩ {001} and ⟨110⟩{110}, slip systems occurred while specimens were under observation. Examination of very thin deposits showed that film growth took plac...

Influence of substrate steps on the arrangement of defects in evaporated films

Abstract Silver films grown on mica or molybdenite substrates often contain defects which are lined up with one another. Evidence is presented that this alignment is caused by steps on the substrate surface. It is shown that the alignment of defects along steps is not surprising if the defects are formed by the coalescence of nuclei.