Oxidation of Epitaxial- and Polycrystalline-SiGe Alloys

  • PDF / 470,911 Bytes
  • 6 Pages / 414.72 x 648 pts Page_size
  • 111 Downloads / 234 Views

DOWNLOAD

REPORT


"3000

rn*

. .

\ 1000°C for 60minI * 900TC for 60min * 800TC for 60min A 700°C for 60min

.

0< 3000

1000"C 900TC 8000C 700TC

Q)* +A

2000

for for for for

60min 60min 60mrin 60min

..• 2000

S1000 0

721000

(a) 10

20

0

30

(b) 0

10

20

30

Ge content(%) Ge content(%) Figure 1 : Oxide thickness as a function of Ge content after wet oxidation between 700 and 1000°C for 60 minutes. (a) for epitaxial-SiGe and (b) for poly-SiGe.

228

4000

4000

0 @ Sio.70Geo.30 30 3000

Z- A Si 0 80 Ge0 20 *Si 0 1

0

*

U 2000

Soo:

1

0.0ooo

I I 0O1 Sio.70Geo. 30 A SiooGeo.20 *r Sio.9 oGeo.10

CJ2000

/

5

.

)

O*(a)

S0.5

. . . . . - - 'Sý

"oGe3000 3000

0(b)

1.01/2 ((hr) 1.5 1/2 ) 2.0 (Time)

0.0

2.5

1.01/2 ((hr) 1.5 1/2) 2.0 0.5 (Time)

2.5

Figure 2 : Oxide thickness versus square root of time for 700'C (filled symbols) and 1,000 0C (open symbols) oxidized samples with Ge concentration between 10 and 30%, (a) for epitaxial-

SiGe and (b) for poly-SiGe. Only data down to 10% are shown. Figure 2(a) shows the oxide thickness versus the square root of time for oxidations at 700 and 1,000TC for the samples of epi with Ge concentrations of 10 to 30%. In epi samples, the

oxide thicknesses do not linearly increase with the square root of the oxidation time at 1,000'C and seem to be suppressed. At 700TC, however, the thickness is proportional to the square root of the oxidation time, indicating that the oxidation process at 700"C is diffusion limited. Figures 3(a) and 3(b) show RBS spectra before and after oxidation at 700'C and 800'C for epi films with Ge concentrations of 20 and 30% respectively. At 700'C Ge is fully incorporated into the oxide and the oxide is of uniform composition. At 800°C, no Ge is incorporated into the oxide for 20% sample, resulting in Ge piling up just behind the oxide(Fig.

3(a)), while some Ge is incorporated into the oxide for 30% sample and the oxide layer consists of three-layers of different composition. A pure SiO 2 is formed near the free surface, a Ge-rich mixed (Si, Ge)0 2 layer is formed underneath, and a Si-rich (Ge, Si)0 2 or a pure SiO 2 film is formed near the oxide/SiGe interface (Fig. 3(b)). At -9000C, no Ge is observed in the oxide for any epi samples with 5-30% Ge.

Our observation of oxidation of epi-SiGe films roughly agree with the previous findings"-a, but add the following new findings : (1) At 700 0C, the thickness of oxides grown on 20, 25, and 30% Ge exceeds those of the corresponding oxides grown at _800 0C. (2) A three-layer oxide is formed for Si0.70Ge030 sample after 800 0C oxidation while Eugene et al.1 2 reported that a twolayer oxide is formed for the sample with _>50% Ge after 9000C wet oxidation, consisting of a mixed (Si, Ge)0 2 layer near the surface and a pure Si0 2 layer underneath. To consistently explain the above observations we propose a model based on the following assumptions: (1) the oxidation rate can be controlled by either the Ge removal from the reaction front or the diffusion of the oxidizing species through the oxide, or both. (2) At 7000