annual report annual report annual report annual report 2005

124
MATERIAL ENGINEERING, STRUCTURAL STUDIES, DIAGNOSTICS
gism of TiO 2 bound with WG-QAC. Taking into
account preliminary results, TiO 2
materials coated
with WG-QAC seem to be very perspective to be
applied as photocatalysts and photobiocatalysts.
References
[1]. Nilius N., Ernst N., Freund H.J.: Chem. Phys. Lett.,
349, 5-6, 351-357 (2001).
[2]. Kim K.D., Lee T.J., Kim H.T.: Colloids. Surf. A, 224,
1-3, 1-9 (2003).
[3]. Pierre A.C.: Am. Ceram. Soc. Bull., 70, 1281 (1991).
[4]. Hoffmann M.R. et al.: Chem. Rev., 95, 69 (1995).
[5]. Siemon U., Bahnemann D.W.: J. Photochem. Photobiol.
A, 148, 247 (2002).
[6]. Łukasiewicz A., Chmielewska D., Waliś L., Krajewski
K.: Ekologia, 29, 3, 36-37 (2005), in Polish.
METALLIC AND/OR OXYGEN ION IMPLANTATION
INTO AlN CERAMICS AS A METHOD OF PREPARATION
FOR ITS DIRECT BONDING WITH COPPER
Marek Barlak 1/ , Wiesława Olesińska 2/ , Jerzy Piekoszewski 1,3 /, Zbigniew Werner 1,4/ ,
Marcin Chmielewski 2/ , Jacek Jagielski 1,2/ , Dariusz Kaliński 2/ , Bożena Sartowska 3/ ,
Katarzyna Borkowska 1/
1/
The Andrzej Soltan Institute for Nuclear Studies, Świerk, Poland
2/
Institute of Electronic Materials Technology, Warszawa, Poland
3/
Institute of Nuclear Chemistry and Technology, Warszawa, Poland
4/
Institute of Physical Chemistry, Polish Academy of Sciences, Warszawa, Poland
Direct bonding (DB) process is recently getting an
increasing interest as a method for producing high
quality joints between aluminum nitride (AlN)
ceramics and copper. Beneficial features of AlN
as a substrate material for AlN-Cu joints in high
power semiconductor devices are: high thermal
conductivity (about 10 times higher than that for
Al 2 O 3 and only about 17% lower than for toxic BeO
ceramic), good electrical insulation, thermal expansion
similar to silicon and non-toxicity. To form
AlN-Cu joint, two prerequisites must be satisfied.
Firstly, oxygen must be present in the system to
allow the formation of Cu/Cu 2 O eutectic liquid
(about 0.4 wt% oxygen in copper) with melting
point 1065 o C, i.e. slightly below the melting point
of copper. Secondly, since the surface of pure AlN
is not wettable by this eutectic, this surface must
be suitably modified to ensure sufficiently strong
bond. Conventionally, AlN is oxidized in a furnace
to convert its surface into Al 2 O 3 type, e.g. [1,2],
which is wettable by Cu/Cu 2 O eutectic. Recently,
we undertook a new approach to AlN pretreatment
prior to DBC using ion implantation technique to
form an intermediate layer between the joined elements.
In the present work, we summarize the
results of systematic experiments in which with
such ion elements as: O, Ti, Fe, Cr and Cu were
implanted in four versions into the commercial
AlN substrates. For each of them, the calculations
of the Gibbs free energy ∆G o was performed to
facilitate the prediction of the formation of a given
compound. The results are as follows:
(i) Oxygen implantation
+ 3/2 (O) → 1/2 + (N)
∆G o 298K=-396.89 kJ/mol
+ 3/2 (O) → 1/2 Al 2 O 3 + 1/2 (N 2 )
∆G o 298K=-852.56 kJ/mol
+ (O) → + (O)
∆G o 298K=0 kJ/mol
+ (O) → + 1/2 (O 2 )
∆G o 298K=-262.45 kJ/mol
(ii) Titanium implantation
+ (Ti) → +
∆G o 298K=-425.57 kJ/mol
In addition, it is likely that the presence of titanium
in AlN substrate can lead to creation of TiN.
The value of ∆G o 298K
computed for such reaction
seems to confirm this possibility:
+ (Ti) → +
∆G o 298K=-445.91 kJ/mol
(iii) O+Ti implantation
Assuming that oxygen implantation may lead to
Al 2 O 3 as in (i), one can expect the formation of
TiO, which is well wettable by copper. Then, the
calculation of Gibbs free energy gives:
+ 3 → 3 + 2
∆G o 298K=-1232.66 kJ/mol
From these calculations, it appears that the probability
of formation of titanium lean oxides, which
are not wettable, is 2 to 3 times smaller. Obviously,
the formation of TiN is also possible as it was shown
in (ii).
(iv) Ti+O implantation
In the fourth approach we intended to check
whether the sequence of implantation of titanium
and oxygen influences the quality of the joints. The
reasoning behind this approach stems from considering
case (ii), where titanium implantation
leads to the formation of TiN and hence oxygen
implantation as the second step will promote the
formation of TiO according to the reactions:
+ (O) → + (N)
∆G o 298K=18.26 kJ/mol
+ (O) → + (N 2 )
∆G o 298K=-195.39 kJ/mol
The metallic ions were implanted using an
MEVVA type TITAN implanter with unseparated
beam, described in detail elsewhere. Oxygen ions
were implanted using a semi-industrial ion implanter
without mass separation equipped with a gaseous
ion source. The substrate temperature did not exceed
200 o C. Ions were implanted at two accelera-