4-3-1 Effect of Deposition Temperature
Fig.4-1 shows the film thickness after implantation process. The thickness is measured by n&k analyzer. Because the optical properties of HfO2 film may change seriously with the heavy implantation dose and the original characteristics of non-implanted HfO2 films may also change after implantation process, Timp is different from Torg and the difference depends on implantation species and dosages.
The intention of the implant process is to damage HfO2 films so that the films could be wet etched efficiently by HF/IPA solution. As shown in Fig. 4-2, the etch rates of the implanted HfO2 films deposited at 500°C and 550°C increases with increase of implantation dose. The etching rate of the 400°C deposited sample is almost independent on implantation dose and is higher than that of all 500 and 550°C deposited samples. The main difference between the 400°C deposited sample and 500°C or 550°C deposited samples is the degree of crystallization, which has been reported previously [53]. It is proposed that the Hf-O bond is etched by un-dissociated HF [38]. The more amorphous HfO2 film causes higher wet etch rate due to the more Hf-O dangling bonds and imperfect Hf-O bonds. For the 400℃ deposited amorphous film, there are many Hf-O bonds exists inside; therefore, the wet etch rate is high. In the 500℃and 550℃ deposited films, since there are fewer Hf-O bonds existed inside,
the wet etching rate is almost zero. When heavier implant dose gets more disorder of HfO2 film, the higher wet etching rate could be obtained. However, the higher deposition temperature produces fewer Hf-O bonds, the wet etching rate decreases under the same implant dose.
The similar result is revealed in the experiment of etching rate selectivity. It could support that the implanted samples can be more easily etched by the reaction of Hf-O dangling bonds and imperfect Hf-O bonds with HF radical. For the selectivity experiment, some regions were damaged by the implant ions and some were not.
After the wet etching process, if the samples reveal very clear etching pattern, some regions are well etched, and selectivity of sample is defined as excellent. If sample reveals light pattern, the etching rate selectivity is defined as good. If sample reveals almost no pattern, two conditions are possible. One is that both of the implanted and non-implanted regions were well etched. The other one is that both were not etched.
To distinguish the two conditions, the thickness of HfO2 film is checked by n&k measurement. The results are summarized in Table 4-1. The symbols “╳”,“△” and
“ˇ” represent poor selectivity, good selectivity, and excellent selectivity, respectively.
For the amorphous structure of 400℃ deposited film, no pattern could be found after the implantation and wet etching process. It’s identified that both the implanted and non-implanted regions were well etched. The pattern could be found easily in the 500
℃ deposited film since the damaged region has more Hf-O dangling bonds and is easier to be wet etched by the HF/IPA solution. The 550℃ deposited samples also present excellent selectivity, except for the condition of 5×1013 cm-2 dose. It is presumably that the insufficient dose cannot produce sufficient damage to the high temperature deposited films to enable wet etching ability. Therefore, the dosage condition should be considered together with the deposition condition of the HfO2
film.
4-3-2 Effect of PDA Temperature
The thickness change of HfO2 film after PDA process is shown in Fig.4-3. The negative value shows the thinner film thickness measured after PDA process. It is believed that the loose structure deposited at lower deposition temperature is porous but the structure would become dense after PDA process. Therefore, the thickness change appears. At higher deposition temperature, the thickness of HfO2 film before and after PDA is almost the same. This observation confirms that the structure of high temperature deposited film is tight. After PDA process, the implantation process was followed. The thickness changes of the 400℃ and 500℃ deposited films were shown in Fig.4-4 and Fig.4-5, respectively, for reference purpose.
In the wet etching rate experiment, the etching rate of the 400℃ deposited film without PDA process is very high. The wet etching rate of 400℃ deposited film is shown in Fig 4-6. As discussed above, Hf-O dangling bonds are repaired with the high temperature process. However, after the implantation process, the film structure is damaged and Hf-O dangling bonds produce. Therefore, with the higher implantation dose, more Hf-O dangling bonds induce higher wet etching rate. On the other hand, with the same implantation dose but different PDA temperature, the higher PDA temperature repairs more Hf-O dangling bonds so that the wet etching rate is lower.
This condition is also presented in 500℃ deposited film and is shown in Fig.4-7.
As shown in Fig.4-4 to 4-7, BF2+ and Ar+ exhibit the same effect as As+ in the wet etching rate experiment. There are also some samples prepared for the selectivity experiment. Table 4-2 shows the outcome of this experiment. The poor selectivity was obtained in 400℃ deposited film due to many Hf-O dangling bonds existed inside the non-implanted region. But in the 500℃ deposited film, less Hf-O bonds exist inside soothe film is quite hard to be wet etched if the implantation process does not proceed.