Impact of Low-Temperature Anneals of Electroplated Copper Films on Copper CMP Removal Rates
- PDF / 561,475 Bytes
- 6 Pages / 417.6 x 639 pts Page_size
- 76 Downloads / 174 Views
Mat. Res. Soc. Symp. Proc. Vol. 566 02000 Materials Research Society
Hardness measurements were taken at vartious time intervals. During the measurement, a known load was applied and a permanent indentation was created on the material under measurement. The hardness value was extracted by measuring the size of the indentation using an optical system. Since our Copper films were thin and soft, the applied deformation load was small. The applied load ramped from 0.04 to 0.4 grams using 15 load and unload steps with a time interval of 4 seconds between steps. Fig. I displays these hardness measurements as a function of storage time. For reference, the resistance transformation is displayed on the same graph (dashed line). It was observed that the film hardness dropped 43% over time of 50-60 hours. It started with an initial hardness reading of 280 Hv and dropped to 160 Hv* in the same time frame of resistivity transformation. It is necessary to find out how much this hardness reduction will impact the CMP removal rate since it has been reported that film hardness can have a profound influence on CMP removal rate [4]. Hardness Change of 1.6pjm EP Cu Film 310
2.250
290
/-
-' -
"-
270
Hardness
--.
2.150
resistivity
2.050 = 0
230
"210--
1.950
•
190-1.850
170
---
---
1 50 1
0
t
10
20
30
40
50
60
70
-
80
-
90
4-I•1.750
100
Time (hours)
Fig. 1 Cu film hardness change as a function of time for a 1.6.tm electrodeposited Cu film at room temperature. To investigate CMP removal rate change as Cu films self anneal, a batch of blanket wafers was electroplated under the same condition. As Cu films were being aged, wafers were pulled at different times to go through an identical CMP process. Cu removal for each wafer was extracted by measuring the initial and final thickness using cross-sectional SEM and also independently measuring wafer weight loss with a balance. In order to identity the stage of transformation, a wafer from the same batch was pulled and monitored regularly for sheet resistance change through the entire period. Fig.2 displays the normalized Cu removal rate as a function of storage time. For reference, the sheet resistance self-annealing curve is also displayed in the graph. As the graph demonstrates, during film transformation, the CMP removal rate increased by -35%. The increase of Cu removal rate over time was in part caused by the 43% mechanical softening of the
" Vickers Hardness Hv is defined as the ratio of the applied load to the indented "unrecovered" projected area. higher the Hv value, the harder the film is.
144
The
Cu film. It is obvious that room temperature self-annealing would result in high variations in CMP process in a real production environment, which is unacceptable for IC manufacturing. In order to ensure consistent Cu film properties and stable CMP removal rates, the electroplated Cu films have to be stabilized. CMP Polishing of 1.6um EP Cn Rim 1.40
2.3
_ , /-
1.35-
0
2.2] 0-.
1.30 1.25 >
2.1
1.20 -
2.0
E
0 .
S1.15-_
o Z
1.05
normaliz
Data Loading...
