Effect of the Surface Temperature on Formation of Low-Temperature Phase in Magnetron-Sputtered Chromium Films Leonid R. SHAGINYAN 1;2 , Jeon Geon HAN 1 and Nikolay V. BRITUN 1 1 Center for Advanced Plasma Surface Technology, SungKyunKwan University, 300 Chun-Chun-Dong, Jangan-gu, Suwon 440-746, Korea 2 Institute for Materials Science Problems of National Academy of Sciences of Ukraine, Krzhizhanovsky str., 3, Kiev-03142, Ukraine (Received July 8, 2004; accepted January 14, 2005; published May 10, 2005) The existence of a low-temperature (LT) chromium phase with the primitive cubic lattice in chromium films is questionable. We revealed that the main condition necessary for LT phase formation in magnetron sputtered Cr films is the deposition of the film from low-energy atomic flux, which is accomplished during deposition at a high working gas pressure. The films deposited from high-energy atomic flux at a low gas pressure crystallized in the ordinary bcc phase independently of the substrate temperature. This effect is explained by assuming the existence of a ‘‘hot’’ layer on the growth surface whose temperature can be noticeably higher than the substrate temperature. [DOI: 10.1143/JJAP.44.3200] KEYWORDS: chromium film, sputtering, transmission electron microscopy, structure, surface temperature 1. Introduction The low-temperature chromium phase (LT) has been discovered and is described in ref. 1. Fine micro-sized particles of chromium produced by evaporation and the subsequent condensation of chromium in argon at low pressures have crystallized in a cubic primitive lattice with the lattice constant a of 4.588  A. This new modification of chromium transforms into the ordinary bcc form at higher than 400  C. To the best of our knowledge, no additional information has appeared in the literature about the formation of the LT phase in Cr films. In view of the high technological and scientific importance of the question concerning the formation of this phase is a significant issue and remains open. A key parameter that defines the formation of the LT phase is the temperature at which the film forms. Substrates in magnetron sputtering systems are heated to high temper- atures due to the energy delivered to the substrate surface by sputtered atoms and from the plasma. 2,3) To reduce the substrate temperature, one should decrease the energy of condensing particles (thermalize them) and reduce plasma radiation. One of the ways of controlling the substrate temperature is the variation of the energy flux delivered to the substrate surface; this can be achieved by varying working gas pressure during film deposition. 4,5) In this paper, we present results demonstrating that the LT phase can be obtained in Cr films deposited by conventional magnetron sputtering. We explain this phenomenon in terms of the temperature that develops within the thin uppermost layer of the growing film during deposition. This temper- ature is substantially higher than conventional substrate temperature due to the energy delivered to the growth surface by sputtered atoms and from the plasma. 2. Experimental Details Cr films were deposited on single crystal silicon wafers via magnetron sputtering of a Cr target with a diameter of 50 mm in argon plasma. The substrate holder was equipped with a chromel-alumel thermocouple mechanically attached to the back surface of the substrate, and silver paste was used to provide the thermal contact between the thermocouple and the substrate. The fixed deposition parameters were: a discharge current of I d ¼ 0:4 A and a substrate-to-target distance of 50 mm. The variables were Ar pressure, P Ar , and substrate temperature, T s . The target voltage was dependent on P Ar and changed from 450 V at P Ar ¼ 5 Pa to 500 V at P Ar ¼ 0:2 Pa. The substrate either was not heated or was heated to the required value before the deposition using resistance heat. The film thickness was in limits of 1–6 mm and the deposition rate was in the range of 4–5 nm/s, depending on the deposition conditions. The cross-sections of Cr films deposited on Si wafers for structural investigations were prepared using a two-step procedure including mechanical polishing and subsequent ion milling. The structures of three parts of the samples (the lower part of the film adjacent to the substrate and the middle and upper parts of the film) were studied using selected area electron diffraction (SAED) and high resolu- tion transmission electron microscopy (HRTEM). For this purpose a high-resolution (2.0  A by point, 300 kV accelerat- ing voltage) electron microscope, JEOL-3010, was utilized. Special measures were taken to increase the accuracy of the lattice parameter (a) measurements from SAED data. First, SAED patterns from different parts of the film were taken along with the SAED pattern from a silicon substrate, which served as a built-in standard for the accurate measurement of a camera constant. Second, a for each film part was calculated as an average of all interplanar distances available from the SAED pattern from this film part. To verify the influence of the energy input delivered by the atoms and from the plasma to the condensation surface on the growth temperature, Cr films were deposited at two Ar pressures P Ar ¼ 5 Pa (high-pressure conditions) and 0.2 Pa (low-pressure conditions). 3. Results 3.1 Films deposited from low-energy atomic fluxes (high- pressure conditions) To ensure that the film has actually grown at the lowest substrate temperature (i.e., from fully thermallized atoms and minimal plasma influence), the film was deposited on the back surface of a free-hanging Si wafer. For this purpose the Si wafer was suspended using a special holder in place of commonly used substrate holder so that the film could be deposited on both sides of the substrate. In this case the film  E-mail address: shagin@skku.edu Japanese Journal of Applied Physics Vol. 44, No. 5A, 2005, pp. 3200–3204 #2005 The Japan Society of Applied Physics 3200