Heterojunction Ohmic Contacts to Si using Ge/Si n + /n + Structures
- PDF / 197,972 Bytes
- 4 Pages / 420.48 x 639 pts Page_size
- 35 Downloads / 252 Views
Mat. Res. Soc. Symp. Proc. Vol. 54. ý 1986 Materials Research Society
512
-1
AI on n-type (100)
Material
10 -2
10 -3
CU
E
10
I
E
-C 0
-4 10
1O >-
i-.-
I'-
-5
10
w
U) U')
-6 10
a: z 0 U
10
-0 10 10is
1 020
(cm-
3)
Iigure 1. Theoretical results of contact resistivity as a function of Si electron concentration ND2 for conventional Al/n+Si contacts (dashed line) and Al/n+Ge/n+Si heterojunction contacts (solid curves) for different values of doping, NDl, in the Ge layer and for different positions of the Fermi level at the heterojunction interface.
513
barrier heights as a function of the carrier concentration in the Si. The barrier height calculations were then used to estimate interface charge necessary to produce the selected barrier heights. The results of our calculations are shown in Figure 1. The contact resistivity of conventional Al/n+Si ohmic contacts (dashed line) and of the Al/n+Ge/n+Si heterojunction ohmic contacts (solid lines) is presented as a function of carrier concentration N9 in the Si. Several values of carrier concentration in the Ge, N 1, are gi n along with two choices of the Fermi level (and resulting barri•r height) at the Ge/Si interface. The barrier height calculations show that if a Ge layer doped to 2 x 1020 cm- 3 can be obtained and if the interface charge density [6] is less than approximately 4 x 1013 cm- 2 the Fermi level at the interface is near the Ge conduction .66 eV) and the resulting barrier to electron flow band-edge (EF - EV 1 will be small. Thus the total contact resistivity could be quite small as indicated by the lower most curve in Fig. 1. For the calculations of Fig. 1 ý, (Al/Si) = 0.72 eV., m*(Ge) = 0.12 mi0 , and ýB(A]/Ge) = 0.43 eV., m*(Si)
= 0.33 m0 , where
mo
is the free electron mass.
EXPERIMENT Test samples containing Kelvin four-terminal devices [71 for measuring contact resistivity were processed concurrently for both Al/n+Ge/n+Si heterojunction and Al/n+Si ohmic contacts. Processing consisted mainly of conventional techniques: phosphorus diffusion, thermal oxidation, photolithography, and thermal evaporation. Selective-area MBE was used to grow epitaxial Ge in the contact windows of the Al/n+Ge/n+Si test devices [8]. Immediately prior to MBE, SiO was etched from the contact windows in an inert atmosphere. A Ge film 9f 3400 A thickness was grown at a substrate temperature of approximately 250°C and a rate of 2500 A/hr. with doping from an elemental As source. During growth a (2x2) RHEED pattern with sharp first and second order diffraction lines was observed. After removing polycrystalline Ge from areas outside the n+Ge/n+Si contact windows, Al was thermally evaporated simultaneously onto both heterojunction and conventional ohmic contact test samples. The best results obtained to date for contact resistance of jnsinter d Al/n+Ge/n+Si ohmic contacts show a contact resistivity of 2 x 10-0 ohm-cm•, one fifth of the contact resistivity of similarly processed Al/n+Si contacts. For these samples, spreading resistance measurements showe
Data Loading...
