found to have elongated blocklike shape ϳ600 nm in width
and 930 nm in length. Some small Al grains, which were
formed at the nucleation stage, were found near the interface.
In the Al grains, there exist many tangled dislocations, and a
few small particles with a diameter of about 50 nm which
were identified as Al O phase by selected area diffraction
2
3
analysis. Al O particles cannot nucleate and grow to such
2
3
size in the Al grains within 30 min at 190 °C. It was also
found that there is not any definite orientational relationship
between the Al O particles and Al matrix. Therefore, a pos-
2
3
sible reason for the existence of Al O particles within the Al
2
3
grains is that Al O particles formed in gas phase because of
2
3
residual O or H O and then deposited together with Al onto
2
2
specimen surface. DMEAA is unstable in the atmosphere and
it readily decomposes into alumina when exposed to air. Be-
cause the volume fraction of Al O particles is small, it is
2
3
impossible to detect them by XRD. The existence of Al O3
2
particles in the Al grains is also one of the reasons why there
exist many tangled dislocations in the grains. Compared with
the AFM results, it was found that the rugged mountains in
the AFM images do not correspond to the grains. It is desir-
able that grain size should be large and uniformly distributed.
Also, the film should be free from impurities. Other research-
8
9
ers showed that impurities like hydrogen, nitrogen, and
oxygen10 have bad effects on the migration resistance of
metal lines.
In summary, the TiN film sputtered on the Si has a pre-
ferred orientation along the growth direction with the ͗111͘
of the film parallel to the Si͗111͘. Sputtering of the TiN film
on the Si induced strains at the interface. The TiN/Si inter-
face is flat while the Al/TiN interface is rough. There exist
many dislocations at the Al/TiN interface. Al O3 phase
2
formed at the Al/TiN interface during the early stages of Al
deposition. In the Al grains, there exist many tangled dislo-
cations and a few Al O particles. With increasing deposition
2
3
time, the Al film surface roughness increases.
This research was supported by LG Semiconductor Co.
and support of Engineering Research Center for Interface
Science and Technology of Materials is also acknowledged.
FIG. 4. Bright field cross-sectional TEM images of ͑a͒ the Al film deposited
on the TiN/Si substrate for 15 min and ͑b͒ the Al/TiN interface, respectively.
1
N. Takeyasu, Y. Kawano, E. Kondoh, T. Katagiri, H. Yamamoto, H. Shin-
angle of about 66°. This direction is parallel to the Si͗111͘.
From the XRD results, it is suggested that the growth direc-
tion of the TiN film is ͗111͘. Therefore, there is a preferred
orientation of the TiN film along the growth direction with
the film ͗111͘ parallel to the Si͗111͘. This is probably be-
cause atomic arrangement and interatomic distance of the
TiN͑111͒ plane are close to those of the Si͑111͒. Obvious
changes in contrast were observed in the Si near the interface
riki, and T. Ohta, Jpn. J. Appl. Phys. 33, 424 ͑1994͒.
J. Drucker, R. Sharma, and K. Weiss, J. Appl. Phys. 76, 8198 ͑1994͒.
Y. Matsumiya, K. Kitahara, N. Ohtsuka, and K. Nakajima, Jpn. J. Appl.
Phys. 34, L17 ͑1995͒.
M. E. Gross, C. G. Fleming, K. P. Cheung, and L. A. Heimbrook, J. Appl.
Phys. 69, 2589 ͑1991͒.
M. G. Simmonds, I. Taupin, and W. L. Gladfelter, Chem. Mater. 6, 935
2
3
4
5
͑
1994͒.
6
7
T. Kaizuka, H. Shinriki, N. Takeyasu, and T. Ohta, Jpn. J. Appl. Phys. 33,
70 ͑1994͒.
4
͓Fig. 4͑a͔͒, indicating that the sputtering of TiN film on the
S. R. Ryu, D. S. Shin, J. E. Oh, J. S. Choi, S. H. Paek, S. I. Lee, J. K. Lee,
T. U. Sim, J. G. Lee, and G. T. Sheng, Appl. Phys. Lett. 62, 579 ͑1993͒.
K. Tokunaga and K. Sugawara, J. Electrochem. Soc. 138, 176 ͑1991͒.
J. Klema, R. Pyle, and E. Domangue, Proceedings of the 22nd Interna-
tional Reliability Physics Symposium ͑IEEE, New York, 1990͒, p. 216
H. Okabayashi and K. Aizawa, Proceedings of the 2nd International
Stress-Induced Phenomena in Metallization ͑AIP, New York, 1994͒, p. 33.
Si induced strains at the interface. The Al/TiN interface is
relatively rough, and there exist many dislocations ͓Fig.
8
9
4
4
͑b͔͒. An intermediate layer indicated by the arrow in Fig.
͑b͒ seems to exist at the Al/TiN interface, which might be
10
the Al O phase found by XRD and AES. The Al grains were
2
3
3428
Appl. Phys. Lett., Vol. 67, No. 23, 4 December 1995
Li, Kim, Rhee
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
37.149.200.5 On: Sat, 29 Nov 2014 14:41:39
1