TY - GEN
T1 - Effect of sputtered TiW deposition conditions on barrier properties for submicron metallization
AU - Georgiou, G. E.
AU - Baker, M.
AU - Eshraghi, S. A.
PY - 1991/1/1
Y1 - 1991/1/1
N2 - We report the results of an orthogonal array study of the step coverage of magnetron sputtered TiW. We concentrate on the effect of target power (l-2.5kW), Ar pressure (4-2lmT), and substrate temperature (100-450°C) on coverage into near vertical wall windows (85° wall angle) with aspect ratio -1.5 (h/w=l.3/0.85) reactive ion etched in a deposited oxide. Scanning electron microscopy (SEM) indicates that coverage defined as the ratio of TiW -1000A thick on top of oxide to that in the bottom of the windows, varies between -25-35%. Coverage improves with moderate power ~2kW and temperature ~300°C. Ar pressure has a somewhat smaller effect on coverage than either power or temperature. TiW resistivity and stress magnitude also decrease to 0.65-0.7μΩ-cm and low 109 dynes/cm2 respectively, for the sputtering conditions used to maximize step coverage. Reverse bias diode leakage (-3000Å deep n+/p with -1000Å TiSi2 and window contacts) does not degrade for Al/TiW or Al/CVD W/TiW metallization of the window contacts are annealed at 450oC,30min., the highest anneal cycle used in this study. Low leakage was measured for a nominal 1000Å thick TiW deposited with the conditions giving the best and worst window coverage, indicating that -300Å is an acceptable barrier at ≤450°C.
AB - We report the results of an orthogonal array study of the step coverage of magnetron sputtered TiW. We concentrate on the effect of target power (l-2.5kW), Ar pressure (4-2lmT), and substrate temperature (100-450°C) on coverage into near vertical wall windows (85° wall angle) with aspect ratio -1.5 (h/w=l.3/0.85) reactive ion etched in a deposited oxide. Scanning electron microscopy (SEM) indicates that coverage defined as the ratio of TiW -1000A thick on top of oxide to that in the bottom of the windows, varies between -25-35%. Coverage improves with moderate power ~2kW and temperature ~300°C. Ar pressure has a somewhat smaller effect on coverage than either power or temperature. TiW resistivity and stress magnitude also decrease to 0.65-0.7μΩ-cm and low 109 dynes/cm2 respectively, for the sputtering conditions used to maximize step coverage. Reverse bias diode leakage (-3000Å deep n+/p with -1000Å TiSi2 and window contacts) does not degrade for Al/TiW or Al/CVD W/TiW metallization of the window contacts are annealed at 450oC,30min., the highest anneal cycle used in this study. Low leakage was measured for a nominal 1000Å thick TiW deposited with the conditions giving the best and worst window coverage, indicating that -300Å is an acceptable barrier at ≤450°C.
UR - http://www.scopus.com/inward/record.url?scp=85067848054&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85067848054&partnerID=8YFLogxK
U2 - 10.1109/VMIC.1991.153044
DO - 10.1109/VMIC.1991.153044
M3 - Conference contribution
T3 - 1991 Proceedings 8th International IEEE VLSI Multilevel Interconnection Conference, VMIC 1991
SP - 420
EP - 422
BT - 1991 Proceedings 8th International IEEE VLSI Multilevel Interconnection Conference, VMIC 1991
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 8th International IEEE VLSI Multilevel Interconnection Conference, VMIC 1991
Y2 - 11 June 1991 through 12 June 1991
ER -