Engineering Materials and Processes phần 8 pdf

Integration Issues 89

30 40 50 60 70 80 90

6

7

8

9

10

11

12

13

14

Ag(200nm)/Al(30nm) on SiO2

400o

C

500o

C

600o

C

Thickness (nm)

time1/2 (s1/2)

Figure 6.6. Al oxide thickness as a square root function of time for an Ag(200nm)/Al(30

nm) bilayer system [8]

6.1.3.2 Growth Kinetics of Oxide Surface Layer

In order to investigate the growth kinetics, the thickness of either aluminum or

silver was changed. The influence of the initial Al thickness on the oxide growth

kinetics was studied by considering the following two bilayers structures: Ag(200

nm)/Al(20 nm) and Ag(200 nm)/Al(30 nm). The AlxOy thickness (x) derived from

the RBS data presented in the previous sections was plotted as a function of the

square root of annealing time (t1/2) for the range 15 to 120 minutes (Figures 6.5 and

6.6). Note the square root of annealing time is expressed in seconds in Figures 6.5

and 6.6. The plots of thickness (x) versus square root of time are straight lines,

which imply that the oxide growth follows a parabolic growth behavior (x2

~t). In

Figure 6.5, the slopes of the 500 and 600°C are almost parallel compare to the

400°C anneal.

For the thicker Al(30 nm), the lines are almost all parallel to each other (Figure

6.6). The diffusion coefficient D for the different bilayers Ag(200 nm)/Al(20 nm)

and Ag(200 nm)/Al(30 nm) systems annealed at different temperatures were

determined by taking the squares of the slopes of the plots in Figures 6.5 and 6.6.

The results of these diffusion coefficients are given in Table 6.1 and reflect the

behavior of the plots of thickness versus time1/2. The diffusion coefficient increases

as a function of temperature and is the highest for the 600°C anneal. It was found

that the growth rates are much higher in Ag(200 nm)/Al(20 nm) than those in

Ag(200 nm)/Al(30 nm) bilayers.