UV irradiation effects in Al chemical vapor deposition on titanium nitride

Article abstract of DOI:10.1063/1.118053

UV irradiation effects on Al chemical vapor deposition on titanium nitride (TiN) was investigated by using dimethylaluminum hydride at 150°C. Al films grew thermally at a rate of 6.3 nm/min, while the UV light generated by a deuterium lamp reduced the rate to 5.2 nm/min. When TiN surfaces were oxidized, Al films started to grow only under UV irradiation. Using x-ray photoelectron spectroscopy (XPS), we showed that the adsorbates formed on the oxidized surfaces could be dissociated only when the UV light was irradiated. The XPS results also suggested involvement of photoinduced desorption in reducing the growth rate.

Full text of DOI:10.1063/1.118053

UV irradiation effects in Al chemical vapor deposition on titanium nitride
Toshinari Nitta and Mitsugu Hanabusa
Citation: Applied Physics Letters 69, 340 (1996); doi: 10.1063/1.118053
View online: http://dx.doi.org/10.1063/1.118053
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/69/3?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Sticking coefficient and growth rate during Al chemical vapor deposition
Appl. Phys. Lett. 69, 4191 (1996); 10.1063/1.116982
Early stages of diamond growth by chemicalvapor deposition monitored both by electron spectroscopies and
microstructural probes
J. Appl. Phys. 80, 1830 (1996); 10.1063/1.362995
Microstructure and deposition rate of aluminum thin films from chemical vapor deposition with
dimethylethylamine alane
Appl. Phys. Lett. 68, 3567 (1996); 10.1063/1.116639
Texture formation in titanium nitride films prepared by chemical vapor deposition
J. Appl. Phys. 79, 8047 (1996); 10.1063/1.362358
Enhancement of titanium nitride barrier metal properties by nitrogen radical assisted metalorganic chemical
vapor deposition
Appl. Phys. Lett. 68, 2580 (1996); 10.1063/1.116190
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:
137.189.170.231 On: Fri, 19 Dec 2014 19:36:54

UV irradiation effects in Al chemical vapor deposition on titanium nitride
Toshinari Nitta and Mitsugu Hanabusaa)
Department of Electrical and Electronic Engineering, Toyohashi University of Technology,
Tenpaku, Toyohashi 441, Japan
͑Received 20 February 1996; accepted for publication 15 May 1996͒
UV irradiation effects on Al chemical vapor deposition on titanium nitride ͑TiN͒ was investigated
by using dimethylaluminum hydride at 150 °C. Al films grew thermally at a rate of 6.3 nm/min,
while the UV light generated by a deuterium lamp reduced the rate to 5.2 nm/min. When TiN
surfaces were oxidized, Al films started to grow only under UV irradiation. Using x-ray
photoelectron spectroscopy ͑XPS͒, we showed that the adsorbates formed on the oxidized surfaces
could be dissociated only when the UV light was irradiated. The XPS results also suggested
involvement of photoinduced desorption in reducing the growth rate. © 1996 American Institute
of Physics. ͓S0003-6951͑96͒00829-7͔
Currently, Al chemical vapor deposition ͑CVD͒ is under
active investigation as a technique for fabrication of various
metallization structures of advanced microelectronics de-
vices. As the source gas, dimethylaluminum hydride
͑DMAH͒ is often used.1 It can produce carbon-free Al films
above 230 °C. The temperature required for the Al deposi-
tion can be lowered below 230 °C by coating the substrate
with catalytic materials.2–5 As an alternative, we have used
photochemical vapor deposition ͑photo-CVD͒.6 The UV
light generated by a deuterium lamp or an ArF laser helped
to lower the substrate temperature required for Al CVD to
about 100 °C.
In the present work we investigated experimentally Al
photo-CVD on titanium nitride ͑TiN͒. TiN serves as a barrier
material that reduces interfacial mixture of Si with Al, and
active investigations are under way to develop techniques to
deposit Al films on TiN.7 Al CVD on TiN had been investi-
gated by other groups,4,7 but the effects induced by UV irra-
diation remained unknown. We showed in this letter that
DMAH was dissociated thermally at 150 °C owing to cata-
lytic activity of TiN, while irradiation of the UV light gen-
erated by a deuterium lamp reduced Al growth rates. On the
other hand, after exposing TiN to oxygen, Al films started to
grow only when the UV light was present.
The experiment was carried out at a low DMAH pres-
sure in order to emphasize a surface reaction.8 According to
x-ray photoelectron spectroscopy ͑XPS͒, the adsorbate that
plays a key role in the surface reaction was not identical on
the surfaces of TiN and its oxidized layer ͑TiON͒. Then,
subsequent dissociation of the adsorbates differed on the two
surfaces. XPS spectra showed thermal dissociation at 150 °C
on TiN, while the adsorbates on TiON could not be dissoci-
ated unless the UV light was present. The XPS study also
showed that photoinduced desorption of adsorbates on TiN
competed with the dissociation at the initial deposition stage.
We explained the observed UV-induced suppression of Al
growth rates on growing Al layers by the photoinduced de-
sorption as well. Thus, XPS studies helped us to explain the
evolution of film thickness observed by a profilometer.
TiN layers, about 15 nm thick, were deposited on
Si͑100͒ substrates by reactive evaporation in a stainless-steel
chamber filled with nitrogen ͑99.995%͒ at a pressure of 5
ϫ10Ϫ4 Torr. We used a tungsten melting pot to evaporate
99.5% pure Ti metal at 1500 °C for 80 s. The electrical re-
sistivity of TiN layers was 100 ␮⍀ cm. TiON layers were
formed by exposing the TiN films to 1 atm pure oxygen
͑99.995%͒ for 1 h at room temperature in the evaporation
chamber. The Ti 2p3/2 XPS peak was observed at binding
energy of 455 eV in agreement with literature.9 The peak
shifted to 458 eV, after exposing to oxygen.
Al films were deposited in a CVD chamber that was
separated by a gate valve from the evaporation chamber de-
scribed above. The chamber, 270 mm long and 152 mm in
diameter, was evacuated to a base pressure of 6ϫ10Ϫ8
Torr by a turbomolecular pump. Throughout the present ex-
periment the substrate temperature was maintained at 150 °C
by an electric heater. DMAH ͑2 Torr at room temperature͒
was blown toward the center of the substrate through a 1/8
in. stainless-steel tube without any carrier gas. We set the
DMAH pressure at 4.5ϫ10Ϫ5 Torr, which was low enough
to warrant the surface reaction.8
As in our previous works, the UV light, generated by a
150 W deuterium lamp ͑Hamamatsu L1853͒ with two broad
emission bands near 160 and 240 nm, was introduced into
the CVD chamber through a CaF 2 lens. We focused the UV
light to a spot size of about 5 mm in diameter on the sub-
strate to increase the power density at the center of the fo-
cused light to about 140 mW/cm2 for the 160 nm band. Hy-
drogen gas was blown against the CaF 2 lens, which served
as an optical window, in an attempt to reduce contamination
of the lens surface. Then, the total pressure in the CVD
chamber increased to about 1.5ϫ10Ϫ2 Torr during the Al
deposition. We used an XPS instrument ͑VSW Scientific In-
struments͒ equipped with an Al K␣ characteristic x-ray
source ͑1486.6 eV͒ in another vacuum chamber connected to
the CVD chamber and evacuated to 10Ϫ9 Torr. With the slit
of an electron energy analyzer set to 4 mmϫ10 mm, energy
resolution was about 1.3 eV, full width at half-maximum.
The takeoff angle for photoelectron was adjusted to 30°.
First, we show in Fig. 1 how Al films grew in thickness
on TiN layers. The film thickness was measured by a con-
ventional stylus-type profilometer. The Al growth was com-
pared at the center of the irradiated area and outside the
a͒
Electronic mail: hanabusa@eee.tut.ac.jp
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:
340 Appl. Phys. Lett. 69 (3), 15 July 1996 0003-6951/96/69(3)/340/3/$10.00 © 1996 American Institute of Physics
137.189.170.231 On: Fri, 19 Dec 2014 19:36:54

FIG. 1. Evolution of Al film thickness observed in ͑a͒ the irradiated area and
͑b͒ the nonirradiated area on TiN layers at 150 °C.
irradiated area on the same substrate. In both cases the Al
films started to grow right after the introduction of DMAH.
The growth rate, measured by the slope of the straight line in
Fig. 1, was 6.3 nm/min for CVD, which was reduced to 5.2
nm/min for photo-CVD.
FIG. 3. XPS spectra showing Al 2p peaks observed on TiN layers without
UV irradiation at 20, 90, and 480 s. The broad signal observed at 90 s was
decomposed into two peaks ͑dotted curves͒, as explained in the text. The
metallic Al position is marked at 72.9 eV. The magnification factor is also
shown.
On the contrary, Al grew on TiON layers only with the
help of UV irradiation, as shown in Fig. 2. The Al films
became detectable with the profilometer only after 6 min.
Similar evolution of Al film thickness was observed in Al
photo-CVD on Si,10 SiO2,11,12 TiO2,13 and Pd.14 During this
initial period, Al islands started to appear on substrate sur-
faces and grew in size, according to a scanning tunneling
microscope study.13 On TiON the Al islands were formed
only when the UV light was irradiated, thus initiating the Al
film growth. The Al films deposited on both TiN and TiON
were highly conductive with the electrical resistivity of about
6 ␮⍀ cm.
while a single, strong peak was observed at 72.9 eV at 480 s,
which coincided exactly with the signal observed from me-
tallic Al films.15 We assigned the 74.1 eV peak to Al con-
tained in adsorbates formed on TiN. We subtracted the 72.9
eV peak ͑after adjusting the peak height properly͒ from the
broad peak observed at 90 s, and obtained a second peak at
74.1 eV. Thus, at 20 s the undissociated adsorbates existed
together with metallic Al on TiN.
Next, the Al 2p XPS spectra observed under UV irradia-
tion were taken for comparison. As shown in Fig. 4, we
observed the 74.1 eV peak at 20 s again. On the other hand,
XPS Al 2p spectra observed at the initial stage of the Al
deposition of TiN without UV irradiation are shown in Fig.
3. At 20 s the Al 2p peak was observed with binding energy
of 74.1 eV. The spectrum changed to a broad peak at 90 s,
FIG. 2. Evolution of Al film thickness observed in ͑a͒ the irradiated area and
͑b͒ the nonirradiated area on TiON layers at 150 °C.
FIG. 4. XPS spectra showing Al 2p peaks observed on TiN layers under UV
irradiation at 20, 90, and 480 s.
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:
Appl. Phys. Lett., Vol. 69, No. 3, 15 July 1996 T. Nitta and M. Hanabusa 341
137.189.170.231 On: Fri, 19 Dec 2014 19:36:54

TiON, which shifted by 1.0 eV and therefore differed from
those formed on TiN. The adsorbates formed on TiON can-
not be dissociated so easily, even under UV irradiation, as
those on TiN. Without UV irradiation they remained undis-
sociated even at 20 min. The UV light helped to dissociate
the adsorbates on TiON and form metallic Al at 10 min. An
additional peak observed at 75.5 eV coincided with the sig-
nal we observed from Al2O3. The reduction of TiON was
confirmed from the shift of Ti 2p binding energy back to 455
eV.
In summary, we studied Al film growth on TiN and
TiON surfaces, using DMAH at 150 °C. Al films grew ther-
mally at a rate of 6.3 nm/min on TiN, which was reduced to
a lower rate of 5.2 nm/min under UV irradiation from the
deuterium lamp. On TiON, Al did not start the growth with-
out the UV light. XPS studies showed that the adsorbates
formed on TiON differed from those on TiN. Part of the
adsorbates formed on TiN was lost through photoinduced
desorption, which explained the reduced growth rates under
UV irradiation. On the other hand, the adsorbates formed on
TiON could be dissociated only with the help of the UV
light, as shown by the thickness and XPS studies.
FIG. 5. XPS spectra showing Al 2p observed under UV irradiation on TiON
layers at ͑a͒ 5 min, ͑b͒ 10 min, and ͑c͒ 20 min. The spectrum observed
without UV irradiation at 20 min is shown in ͑d͒.
We thank Kazuo Kawahito for technical assistance. This
work was partly supported by a Grant-in-Aid from the Min-
istry of Education, Science, Sports, and Culture.
only 72.9 eV metallic Al peak was observed at 90 s. We
cannot ascribe the absence of the 74.1 eV peak at 90 s to the
dissociation of the adsorbates, because this interpretation
contradicts the observation that the Al films grew more
slowly under UV irradiation than without UV irradiation
͑Fig. 1͒. Instead, we propose that the adsorbates were lost
partly by UV-induced desorption on the TiN surface, thus
slowing down the Al growth.
The UV-induced desorption may offer an explanation
for the reduced Al growth rate observed on TiN at 150 °C.
Namely, the desorption competes with the dissociation,
while their relative importance changes with temperature. In-
deed, at 200 °C the UV-induced suppression was more
prominent than in the present case.11 This was observed on
Si surfaces, but when Al was growing on Al underlayers the
substrate surface should not have affected the outcome. At
120 °C the UV light helped to enhance the growth rate on
Al.13 These results suggested greater importance of the
photoinduced desorption at higher temperatures.
1 R. Bhat, M. A. Koza, C. C. Chang, and S. A. Schwarz, J. Cryst. Growth
77, 7 ͑1986͒.
2 O. Lehmann and M. Stuke, Appl. Phys. Lett. 61, 2027 ͑1992͒.
3 K. Sugai, T. Shinzawa, S. Kishida, and H. Okabayashi, Digest of Papers,
4th MicroProcess Conference, Kanazawa, Japan, 1991 ͑unpublished͒, pp.
206–207.
4 K. Sugai, H. Okabayashi, T. Shinzawa, S. Kishida, T. Kobayashi, N.
Hosokawa, T. Yako, H. Kadokura, M. Isemura, and M. Naruse, Proceed-
ings of the Symposia on Interconnects, Contact Metallization, and Multi-
level Metallization/Reliability for Semiconductor Devices, Interconnects,
and Thin Insulator Materials, edited by T. O. Herndon, K. Okabayashi, N.
Alvi, H. S. Rathore, R. A. Susko, and M. Kashiwagi ͑Electrochemical
Society, Pennington, NJ, 1993͒, Vol. 93-25, pp. 290–298.
5 K. Sugai, H. Okabayashi, T. Shinzawa, S. Kishida, A. Kobayashi, T.
Yako, and H. Kadokura, J. Vac. Sci. Technol. B 13, 2115 ͑1995͒.
6 M. Hanabusa, A. Oikawa, and Peng Ying Cai, J. Appl. Phys. 66, 3268
͑1989͒; 67, 3208E ͑1990͒.
7 M. Eizenberg, MRS. Bull. XX ͑11͒, 38 ͑1995͒.
8 M. Hanabusa and M. Ikeda, Appl. Organomet. Chem. 5, 289 ͑1991͒.
9 D. Briggs and M. P. Seah, Practical Surface Analysis, 2nd ed. ͑Wiley,
Chichester, 1990͒, p. 606.
Finally, Fig. 5 shows the Al 2p spectra taken for the
TiON surface at the initial stage of the Al deposition. Even
under UV irradiation the metallic Al signal was missing at 5
min. An identical peak was observed at 20 min without UV
irradiation ͓Fig. 5͑d͔͒. These results agreed with those ob-
tained by the profilometer ͑Fig. 2͒. The peak observed at
75.1 eV is assumed to represent the adsorbates formed on
10 M. Hanabusa and A. Komatsu, Appl. Surf. Sci. 86, 428 ͑1995͒.
11 T. Kawai, M. Okawa, A. Komatsu, T. Shimada, and M. Hanabusa, Appl.
Surf. Sci. 79/80, 385 ͑1994͒.
12 T. Kawai and M. Hanabusa, Jpn. J. Appl. Phys. 32, 4690 ͑1993͒.
13 T. Nitta, K. Nishitani, and M. Hanabusa, Jpn. J. Appl. Phys. 34, L1500
͑1995͒.
14 M. Hanabusa and A. Komatsu, Appl. Phys. Lett. 65, 1826 ͑1994͒.
15 Reference 9, p. 601.
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:
342 Appl. Phys. Lett., Vol. 69, No. 3, 15 July 1996 T. Nitta and M. Hanabusa
137.189.170.231 On: Fri, 19 Dec 2014 19:36:54