Deposition and Annealing Studies of Diamond-Like Carbon Grown by Molecular Beam Deposition
- PDF / 958,306 Bytes
- 6 Pages / 417.6 x 639 pts Page_size
- 94 Downloads / 198 Views
3. RESULTS AND DISCUSSION
2
1n 101
as-depos ited
"7, E
10 0
0
0 "10
.-*--2000 C ananealed
•
A'
-v-
10"
300 0C anlnealed
400°C anlnealed 500°C anlnealed 49
10";
U I
I00,00
0,05
0,10
0,15
0,20
0,25
1IT (K1)
Figure 1: Changes in conductivity as a function of inverse temperature Figure 1 shows the conductivity of the as-deposited and annealed sample as a function of measurement temperature. Saxena and Bragg [2] have shown that the electrical conduction in glassy carbon can be expressed as a combination of variable range hopping and a strongly scattered metallic conductivity component independent of temperature. Later it has been found that the conduction in some pyrolyzed polymer films and highly graphitized poly(p-phenylene vinylene) films [2,3] exhibits also a combination of variable range hopping and strongly scattered metallic conduction. The overall conductivity can be expressed as: a(T)=a(O)+ aoexp[-[To/T] " 4]
(1)
In equation (1) a(O) is the conductivity at 0°K, a 0 is a constant factor and To is the characteristic temperature given by: (2)
To= [16&a]/keN(EF)
where kB is Boltzman's constant, N(EF) is the density of states and a-' is the radius of the localized state wave function. ca' is usually of the order of the crystallite size and thus gives an indication of the amount of disorder within the structure. For glassy carbon, an (x-1 value has been estimated of about 15-25A [2,3]. Sputtered amorphous carbon has an a'"value of around 12A[6]. As a first attempt in interpreting our results, a fitting of our data to equation (2) was made. cr(O) was determined from the intercept of In (a(T)) versus T plots. The
268
exponential factor To and the prefactor a 0 were determined by the slope and intercept of In (ca(T)-a(0)) versus T-"14 plots, respectively. In fitting our data to equation (2), we note that our implied assumption is that the modified layer is homogenous. An other attempt was also made to fit the data to a 3D variable range hopping model which has been found to work well for amorphous carbon and ion implanted diamond. This was done in order to check if all the samples have a strongly scattering metallic layer. The data were fitted to: a(T)=5oexp[-[To/T] 1/4 ]
(3)
and indeed straight lines (Figure 2 shows an example for the 200 0 C annealed sample) were obtained. Extraction of the gradient To (Figure 4) from equation 3 resulted for the as deposited and 500°C annealed samples in a value of 1021 states/eVcm 3 . 5,5
5,05 "
E 4,5 4,0
v3,5 C
0
0 - 3,02,50.2
0,3
0,4
0,5
0,6
0,7
08
lITr25 (K-1) 25 Figure 2: Logarithmic conductivity as a function of 1/To'
These values are reasonable. They were already found for amorphous carbon layers[4]. For the 2000 C, 3000 C and 400 0 C annealed sample we have found values in the range of 10 3 and 1024 states/eVcm 3 respectively. Such high values in the range of 1023 or 1024 states/eVcm 3 are not reasonable for carbon. The results of the as-deposited and 5000C samples indicate that annealing can not decrease the density of states. The
Data Loading...
