Modelling the Influence of the Silicon Doping Profile on a Silicide Well Ohmic Contact
- PDF / 2,043,545 Bytes
- 6 Pages / 417.6 x 639 pts Page_size
- 47 Downloads / 191 Views
Continued reductions in device geometry have required commensurate reductions in device contact resistance in order not to compromise down-scaled device performance. Reductions in contact resistance for ULSI devices are achieved in a variety of ways. A common approach is the inclusion of a low resistivity silicide between the metal and heavily doped semiconductor. Further reductions in Rc are obtained by growing the silicide as a well into the heavily doped silicon of the drain/source region, and by extending the silicide laterally so that it surrounds the immediate metal contact via. Fig.1 illustrates a cross-section of part of a MOSFET device, showing the relevant features of a silicided source/drain contact. The depth of the heavily doped silicon is t, the silicide well depth is h and the overlap of the silicide around the metal via is 5. Monitoring the electrical properties of the device contact is commonly done by the measurement of the interfacial metal-semiconductor resistance and the subsequent derivation of the specific contact resistance pc(Q.cm 2 ) using a Cross Kelvin Resistor (CKR) test structure [1]. Corrections to Pc for test structure parasitic resistance are generally made using twodimensional (2-D) models (ie the semiconductor layer is assumed to have zero depth) of the Kelvin structure. However a 2-D model is inappropriate for a silicide-well structure [1] where the depth of the silicide and the doping profile of the silicon need to be modelled in order to calculate their influence on contact properties. A 3-D CKR test structure, which corresponds to 171
Mat. Res. Soc. Symp. Proc. Vol. 564 © 1999 Materials Research Society
the device contact in fig. 1, is shown in fig.2(a) while in 2(b) a crosssection of the contact region of fig.2(a) is illustrated using the same DIELECTRIC
.
o
°namely
__________________________________ Fig. 1. Device cross-section with a silicide-well contact. CONTACT
geometrical parameters as in fig.l. Fig.2(b) also shows the two relevant material interfaces with their own specific contact resistance values, the metal-silicide interface (Pca) and the silicide-silicon interface (Pcu).
IV
VOLTAGE FINITEMETAL :SEMICONDUCTOIR •
OTG
SILICON
SIENCIDE
tm lSILICIDE
Pcu (b) contact cross-section of the silicide Fig. 2. (a) CKR test structure of silicide well contact, (Petu) contact interfaces and geometry. well showing metal-silicide (Pea) and silicide-silicon FINITE ELEMENT MODEL
A 3-D Finite Element (FE) model of the CKR structure of fig.2 was used to obtain equipotentials and current density within the structure. Three different vertical doping profiles for the heavily doped n-type silicon region of the contact are modelled. They are (a) Constant doping profile (P1: lxI02°cm"3 to a depth of 150nm) (b) Phosphorus implant (P2: 40 KeV Phosphorous at 2 x 1015 cm- 2 plus a RTA) (c) Phosphorus implant (P3: 25 KeV Phosphorous at 1 x 1015 cm- 2 plus a RTA) Profiles P2 and P3 were derived using an ion implantation simulation program for a phosphorus implant into silicon. The
Data Loading...
