Patent Application
20070270085
A preferred embodiment of the present invention will be described in the following, but the present invention should not be construed to be limited to the present embodiment. Examples of the method of producing dry type composite particles comprising organic host particles and inorganic guest particles can be found in the compounding methods according to the Mechanofusion System of Hosokawa Micron Co., Ltd. and the Hybridizing System of Nara Machinery Co., Ltd. These compounding methods are intended to produce composite particles by combining different ingredient particles on molecular levels mechanochemically by applying frictions, pressures and shear forces based on mechanical energy. These methods are simpler in the processes than the method of producing wet type composite particles and have higher degrees of freedom of combination.
Next, the polishing test method conducted to prove the operating effect of the present embodiment of the invention. A polishing device by Engis Corporation, IMRTECH 10DVT, was used to measure the polishing speed on the silicon oxide (SiO2) film by gluing a polishing pad (IC1000/Suba 400 by Nitta Haas Inc.) on top of a polishing platen (diameter: 200 mm) and another pad on the bottom of a guide ring. A silicon oxide wafer, which was the material to be polished, was set on the bottom of the polishing weight which was then placed on the polishing pad inside the guide ring, allowing the guide ring and the polishing weight to rotate in the same direction as that of the platen by the friction resistance of the contract surface in accordance with the rotation of the platen thus to create polishing of the material The polishing load was adjusted by the number of polishing weights, while setting the weight to 30 kPa, the rotating speed of the platen to 150 rev/minute, and the polishing time to 2 minutes. During the polishing, the slurry was supplied continuously by dropping (15 ml/min) by means of a tube pump. The silicon oxide film wafer was cleaned by ultrasonic cleaning for 10 minutes after the polishing using pure water and then dried. The film thickness of the silicon oxide film was measured using an optical interferometer type film thickness measuring deice to obtain the difference in the film thickness before and after the polishing in order to calculate the polishing speed.
In all of the embodiments 1-3 according to the present invention, the slurry for CMP was based on composite particle abrasive grains produced by the dry type composite producing method compounding organic host particles of poly(methylmethacrylate) (PMMA) mono-dispersion particles (5 μm) with inorganic particles of ceria (CeO2) particles (14 nm) in such a way that the zeta potential becomes a negative potential. Since it is difficult to measure the zeta potentials of the composite particle abrasive grains thus produced as the high concentration slurry, large-sized particles and composite particles cannot be measured easily by the conventional laser Doppler method, it was measured by Matec Applied Sciences' instrument, ESA-9800, that measures the zeta potential by measuring the pressure amplitude of the high frequency vibrations caused by electrophoresis. The zeta potentials of the composite particles were −40 mV in embodiment 1, −20 mV in embodiment 2, and −5 mV in embodiment 3. The slurries of these embodiments had composite particles with negative potentials as a result of adjusting the zeta potentials by means of changing the concentrations introducing carboxyl groups to the organic host particles. The above polishing tests were conducted by adjusting the concentration of abrasive grains (weight ratio of the abrasive grains relative to water) to 1 weight percentage. In order to check the polishing selectivity, a Si3N4 film wafer was used as the polish stopper film in addition to the SiO2 wafer as the workpiece to be polished. The polishing speed and the polishing selectivity (ratio of polishing speeds of the SiO2 film and the Si3N4 film) were evaluated.
Comparative examples 1 and 2 were based on composite particles consisting of conventional poly(methylmethacrylate) (“PMMA”) mono-dispersion particles and PMMA mono-dispersion particles to which hydroxyl groups were introduced in order to achieve composite particles of positive potential. Example 3 is the sample for polishing speed comparison prepared by using inorganic host particles consisting of nano ceria alone.
Table 1 shows the evaluation results of CMP slurries of embodiments 1 to 3 and comparative examples 1 to 3.
As shown in table 1, the slurries for CMP of embodies 1 through 3 showed much faster SiO2 polishing speeds compared to that of comparative example 3 using nano ceria alone, indicating excellent polishing capabilities, also indicating that better polishing selectivity was achieved when the absolute value of the negative potential of the zeta potentials of the composite particles increased. Since their zeta potentials were positive in comparative examples 1 and 2, adequate polishing selectivity was not achieved in both cases although the polishing speeds were very high.
Although PMMA particles implanted with carboxyl groups were used as the organic host particles in embodiments 1 through 3 in order to make the zeta potential of the composite particles negative, organic host particles having functional groups implanted with sulfonyl groups or other functional groups that cause a negative potential can be used as well. Although PMMA particles were used as the base material for the organic host particles in embodiments 1 through 3, we note here that similar effects can be achieved by using polystyrene particles implanted with functional groups as the organic host particles. While the aforementioned dry type composite particles were used in embodiments 1-3, it was learned that it is also possible to achieve high speed polishing and sufficient polishing selectivity using wed type composite particles produced by hetero aggregation. Although it is written in this specification that the abrasive grains used in the embodiments are totally made up of composite particles alone for the simplicity sake, some nano ceria particles that are not compounded exist in those particles in reality, as they are not 100% compounded in the dry type composite particle producing method. If the surface coverage rate of the composite particles is the same, the more uncompounded nano ceria particles exist, the faster the polishing speed of the SiO2 film. Thus, the same effect can be achieved in the slurry using abrasive grains where the composite particles are added with nano ceria. Although HDP-TEOS (High Density Plasma-Tetra Ethoxy Silane that is capable of lowering the softening temperature and achieving gettering action as the SIO2 film, which is the workpiece to be polished in these embodiments, thermally oxidizing films and silicon oxide films such as 03-TEOS and SOG (Spin On Glass).
Next, the relation between the concentration of the abrasive grains of composite particles and the polishing speed in the slurry for CMP was evaluated. The polishing test was conducted under similar conditions as in embodiments 1 through 3. The slurry for CMP used in embodiments 4 through 7 were composite particle abrasive grains consisting of mono-dispersion particles (5 μm) as the organic host particles of PMMA and CeO2 particles (14 nm) as the inorganic guest particles same as in embodiment 2, having the zeta potential of −20 mV. The concentration of the abrasive grains relative to water was 0.2 wt % in embodiment 4, 1.5 wt % in embodiment 5, 5 wt % in embodiment 6, and 10 wt % in embodiment 7.
While the abrasive grains of composite particles used in embodiment 2 were used both for comparative examples 4 and 5, the concentration of the abrasive grains relative to water was 0.1 wt % in embodiment 4, and 20 wt % in embodiment 5 in the polishing tests conducted similarly to embodiments 1 through 3. Table 2 shows the evaluation results of embodiments 4 through 7 as well as comparative examples 4 and 5.
As can be seen from Table 2, the slurries of embodiments 4 through 7 showed very high SiO2 film polishing speeds and excellent polishing capabilities in comparison with comparative example 4. In contrast to this, comparative example 4 with the abrasive grain concentration of 0.1 wt % showed a much slower polishing speed and sufficient polishing speeds for practical use could not be achieved. Comparative example 5 with the abrasive grain concentration of 20 wt % showed almost no difference and was in a saturated condition even in comparison with embodiment 7 with the abrasive grain concentration of 10 wt %. Furthermore, the dispersion state of the abrasive grains of comparative example 5 was very poor although it is not shown in the table.
Next, the relation between the concentration of the abrasive grains of composite particles and the polishing speed in the slurry for CMP was evaluated. The polishing test was conducted under the same conditions. The slurries for CMP used in embodiments 8 through 11 were composite particle abrasive grains consisting of organic host particles of PMMA mono-dispersion particles and inorganic guest particles of CeO2 particles (14 nm), having the abrasive concentration of 1 wt %, while the average particle size of the PMMA mono-dispersion particles used as organic host particles was 0.3 μm in embodiment 8, 1.5 μm in embodiment 9, 5 μm in embodiment 10, and 10 μm in embodiment 11. Since the average particle diameter of the inorganic guest particles was 14 nm, sufficiently smaller than that of the host particles, so that the average particle diameter of the abovementioned PMMA mono-dispersion particles can be regarded as the average particle diameter of the composite particles, which were the abrasive grains. Since carboxyl groups were introduced into PMMA, the zeta potentials of all the composite particles thus produced were negative potentials. In addition to the polishing speed, the damages of the wafers after polishing were observed.
Comparative examples 6 and 7 were identical to embodiments 8 through 11 in that they were based pm composite particle abrasive grains consisting of PMMA mono-dispersion particles used as the organic host particles and CeO2 particles (14 nm) as the inorganic guest particles, with the concentration of the abrasive grains of 1 wt % and the zeta potential of the abrasive particles of a negative potential, except that the particle diameter mono-dispersant PMMA particles of the organic host particles, which is essentially equal to that of the composite particles, was 0.15 μm in comparative example 6 and 20 μm in comparative example 7.
Table 3 shows the evaluation results of embodiments 8 through 11 and comparative examples 6 and 7, which were conducted in a similar manner as the previous polishing test method.
As shown in Table 3, the slurries for CMP in embodiments 8-11 showed very fast SiO2 polishing speed, no damages on wafers after polish, and excellent performances overall. The fasted polishing speed was noted in embodiment 9 where the average diameter of the composite particles was 1.5 μm. In contrast to that, the polishing speed dropped substantially and sufficient polishing speeds could not be achieved in comparative example 6 where the average particle diameter of the composite particles was 0.15 μm. Furthermore, although a sufficient polishing speed was achieved, many damages were observed on the polished surface, and the slurry dispersion state was very poor in comparative example 7 where the average particle diameter of the composite particles was 20 μm. In the dry composite particle producing method, it is not easy to produce composite particles if the average particle diameter of the organic host particles is less than 1 μm, so that it should preferably be 1-7 μm. The average particle diameter of the composite particle abrasive grains relates very much to the type of the polishing pad, especially to the pattern of the polishing pad which is made of porous urethane plastic and its surface roughness, so that it is mandatory to select the particle diameter suited for the polishing pad to be used.
The slurries for CMP added with planarization additives are evaluated here. Embodiments 12 through 14 were based on the slurries for CMP used in embodiments 1 through 3 added with poly(methyl)acrylic acid ammonium salt as an planarization additive. The composition of the additive relative to water was 0.3 wt %, the pH value of the slurries was adjusted to 5 using ammonia, the abrasive grain concentration relative to water was 1.0 wt %, and the zeta potential prior to the addition of the planarization additive was −40 mV in embodiment 12, −20 mV in embodiment 13, and −5 mV in embodiment 14.
Comparative examples 8 through 10 were based on the slurries used in comparative embodiments 1 through 3 added with poly(methyl)acrylic acid ammonium salt as an planarization additive. The composition of the additive relative to water was 0.3 wt %, the pH value of the slurries was adjusted to 5 using ammonia, and the abrasive grain concentration relative to water was 10 wt %.
Table 4 shows the evaluation results of embodiments 12 through 14 and comparative examples 8 through 10, which were conducted in a similar manner as the previous polishing test method.
As can be seen from Table 4, the slurries of embodiments 12 through 14 showed very high SiO2 film polishing speeds and excellent polishing capabilities in comparison with comparative example 10 with nano ceria only. Moreover, comparative examples 8 and 9 used composite particles with positive zeta potentials showed substantial drops in the polishing speeds as a result of adding planarization additives which work to prevent polishing. In contrast to that, the slurries for CMV of embodiments 12 through 14 using composite particles with negative zeta potentials caused only small polishing speed reductions even though planarization additives are added, and could sustain adequate polishing selectivity as well.
Although PMMA particles implanted with carboxyl groups were used as the organic host particles in embodiments 12 through 14 in order to make the zeta potential of the composite particles negative, organic host particles having functional groups implanted with sulfonyl groups or other functional groups that cause a negative potential can be used as well. Although PMMA particles were used as the base material for the organic host particles in embodiments 12 through 14, we note here that similar effects can be achieved by using polystyrene particles implanted with functional groups as the organic host particles. Embodiments 12 through 14 used slurries added with planarization additives with the pH value of 5, it was observed that similar results as in embodiments 12 through 14 can be achieved if the pH value is within 4-8. The planarization additives used in the present invention are not limited to poly(methyl)acrylic acid ammonium salt, but rather can be any kind of additives that adhere to inorganic dielectric films placed as workpieces to be polished such as SiO2 and Si3N4 and separate when polishing surface pressure increases, when they are incorporated as additives.
Next, the relation between the concentration of the abrasive grains of composite particles added with planarization additives and the polishing speed in the slurry for CMP was evaluated. The slurry for CMP used in embodiments 15 through 18 was made from composite particle abrasive grains consisting of PMMA mono-dispersion particles (5 μm) as the organic host particles and CeO2 particles (14 nm) as the inorganic guest particles same as in embodiments 2 and 13, the zeta potential was −20 mV, and poly(methyl)acrylic acid ammonium salt was used as the planarization additive. The composition of the additive relative to water was 0.3 wt %, the pH value of the slurry was adjusted to 5 using ammonia, the concentration of the abrasive grains relative to water was 0.2 wt % in embodiment 15, 1.0 wt % in embodiment 16, 5 wt % in embodiment 17, and 10 wt % in embodiment 18.
The aforementioned abrasive grains of composite particles used in embodiments 2 and 13 were also used in comparative examples 11 and 12, and the concentration of the abrasive grains relative to water was 0.1 wt % in comparative example 11 and 20 wt % in comparative example 12.
Table 5 shows the evaluation results of embodiments 15 through 18 and comparative examples 11 and 12, which were conducted in a similar manner as the previous polishing test method.
As can be seen from Table 5, the slurries of embodiments 15 through 18 showed very high SiO2 polishing speeds and excellent polishing capabilities in comparison with comparative example 11. When the concentration of the abrasive grains was less than 0.2 wt %, the polishing speed dropped sharply and sufficient speeds for practical use could not be achieved as shown in comparative example 11. On the other hand, when the abrasive concentration was as much as 20 wt % as shown in comparative example 12, no difference in the polishing speed was observed even in comparison with comparative example 18 whose grain concentration was 10 wt % and the polishing speed was almost saturated. Furthermore, in comparative example 12 with the concentration of the abrasive grains of 20 wt %, the dispersion state of the abrasive grains was very poor although it does not appear on the table. Moreover, according to the evaluation of the effect of the average particle diameter of the composite particles concerning the slurries added with the planarization additives, the polishing speed dropped substantially and sufficient polishing speed for practical use could not be achieved when the average particle size of the composite particle abrasive grains was less than 0.3 μm. Also, when the average particle diameter of the composite particle abrasive grains was 20 μm, many damages were noted on the workpiece surface and the dispersion state was very poor although the polishing speed was sufficiently high. From these results, it was learned that the average particle diameter of the composite particle abrasive grains should preferably be within the range of 0.3-10 μm, and that it is preferable to select slurries for CMP using abrasive grains of composite particles of the optimum particle diameter that matches the type of polishing pad, in particular, the pattern and surface roughness of polishing pads made of porous urethane plastics.
Next, the relation between the concentration of planarization additive and polishing selectivity of the slurry for CMP was evaluated. In order to check the polishing selectivity, the SiO2 film wafer was chosen as the workpiece to be polished and the Si3N4 film wafer was chosen as the polishing stopper film. The slurry for CMP used in embodiments 19-22 was made from composite particle abrasive grains used in embodiment 13 consisting of PMMA mono-dispersion particles (5 μm) as the organic host particles and CeO2 particles (14 nm) as the inorganic guest particles, wherein the zeta potential was −20 mV, the concentration of the abrasive grains relative to water was 1.0 wt %, poly(methyl)acrylic acid ammonium salt was used as an planarization additive, and the pH value of the slurry was adjusted to 5 using ammonia. The concentration of the planarization additive, i.e., poly(methyl)acrylic acid ammonium salt, relative to water was 0.05 wt % in embodiment 19, 0.3 wt % in embodiment 20, 1.0 wt % in embodiment 21, and 5.0 wt % in embodiment 22.
The composite particle abrasive grains having the same planarization additive concentration as that of embodiments 19 through 22 were used in comparative examples 13 and 14, wherein the concentration of the abrasive grains relative to water was 1.0 wt %, the concentration of poly(methyl)acrylic acid ammonium salt used as the planarization additive was 0.001 wt % in comparative example 13 and 10 wt % in comparative embodiment 14, and the slurry's pH values was adjusted to 5 using ammonia.
Tests were conducted under the same condition of the aforementioned polishing test method for embodiments 19 through 22 and comparative examples 13 and 14 in order to check the operating effects of the embodiments. Table 6 shows the results of the tests.
As can be seen from Table 6, it was learned that the slurries of embodiments 19 through 22 show very high polishing selectivity as they presented very high polishing speeds for the SIO2 film and slow speeds for the Si3N4 film as the stopper film, thus achieving high planarity in Chemical Mechanical Polishing. In contrast to that, it was learned that comparative example 13, whose planarization additive's concentration was 0.01 wt %, shows that it has a very fast polishing speed for the SiO2 film, but polishes through the Si3N4 film, which is the stopper film, as well, so that it has a shortcoming of not being able to achieve sufficient planarity. It was also learned that comparative example 14 whose planarization additive's concentration was 10 wt %, has high polishing selectivity and provides high planarity, but has a problem that its polishing speed for the SiO2 film is very slow so that its process is very poor.
From the polishing tests of various embodiments, it is evident that the slurry for Chemical Mechanical Polishing, the Chemical Mechanical Polishing method using said slurry, and the method of producing electronic devices using said method as recited in the claims of the present invention have an excellent polishing capability for SiO2 film, and provide effects that fully satisfy low scratch characteristics and polishing selectivity. Although the descriptions about the present embodiments were made concerning CMP of the STI method, the present invention can be applied to abradants for other processes where reduction of damages are required such as CMP in Inter Layer Dielectric (ILD) films.
The present invention can be applied to abradants used in processes where damage reduction is required such as CMP (Inter Layer Dielectric film CMP, etc., in addition to STI-CMP) for semiconductor devices, which are electronic devices.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2006-139787 | May 2006 | JP | national |
