Chapter 3 Experiments of Al/Ti/HfO 2 /Si MIS capacitors
3.5 MIS capacitors fabrication process
3.5.3 Dual plasma treatment for MIS capacitors
After standard initial RCA clean, wafers were put into chamber of the PECVD and plasma fluorination treats the Si surface. The plasma fluorination treatment conditions were in pure CF4 gas for 10 sec, 20 sec, 30 sec, and 40 sec respectively and the flow rate were 100 sccm and bias 20W in the 300℃ environment. After the plasma treatment were finished, the wafers were put into chamber and grew HfO2 layer with MOCVD.
After the thin films were deposited, most samples were annealed in a N2
25
ambient for 30 sec at 600℃ after deposition (PDA, Post-Deposition Anneal) by rapid thermal annealing (RTA). Then, we used the PECVD again to plasma nitridation treat the high-k surface. The plasma treatment conditions were in pure N2, NH3, and N2O gas for 120 sec, and respectively and the flow rate were 100 sccm and bias 20W in the 300℃
environment. Pure titanic was deposited on the HfO2 layer by e-gun evaporation system and aluminum films were evaporated on the top side of wafers. Mask defined the top electrode. Then, we used wet etching to etch undefined Al and Ti films. After patterning, backside native oxide was stripped with diluted HF solution, and Al was deposited as bottom electrode. The detailed fabrication process flow was listed as follows.
1. As shown in Fig 2-3
(1) Si substrate RCA clean
(2) Plasma fluorination treatment (3) Annealing by RTA
2. As shown in Fig 2-4
(1) 4nm HfO2 was deposited on the sub-Si by MOCVD.
(2) PDA by RTA 3. As shown in Fig 2-5
(1) Plasma nitridation treatment (2) PNA by RTA
4. As shown in Fig 2-6
20nm Ti was deposited on the HfO2 layer by E-gun evaporation system.
5. As shown in Fig 2-7
26
400nm Al was deposited on the Ti layer as top electrode by E-gun evaporation system.
6. As shown in Fig 2-8
Undefined Al was removed by wet etching.
7. As shown in Fig 2-9
Undefined Ti was removed by wet etching (1%HF).
8. As shown in Fig 2-10
Al was depodited on the back side of sub-Si as bottom electrode by E-gun evaporation system.
3.6 The MIS Capacitors measurement
After the Al/Ti/ HfO2 /Si MIS capacitors were prepared, we used semiconductor parameter analyzer (HP4156C) and C-V measurement (HP4284) to analysis electric characteristics (i.e. I-V, C-V, EOT, leakage current density etc.).
27
Chapter 4
Electrical characteristics of Al/Ti/HfO2/Si MIS capacitors In order to measure the C-V characteristics of our MIS capacitors, we used HP 4284A precision LCR meter in our experiments. We swept the gate bias from inversion region to accumulation region to obtain the curve at the frequency of 50 kHz from -2V to 1V. And the leakage current of our MIS capacitors were analyzed from the current-voltage (I-V) characteristics measured by an HP4156A semiconductor parameter analyzer. There are four kinds of plasma treatment with different souce gas (i.e. N2 , N2O , NH3) for different process time (i.e. 0 sec , 30sec , 60sec , 90sec , 120sec , 150sec , 180sec ) and CF4 plasma treatment for different process time (i.e. 0sec , 10sec , 20sec, 30sec , 40sec). Finally, we combine the two plasma process to obtain the optimal condition. Hence, the relationship of difference process time with CF4 in one kinds of plasma treatment will be discussed.
The hysteresis will also be discussed in this experiment. The name of Hysteresis was borrowed from electromagnetic. It is means that when a ferromagnetic material is magnetized in one direction, it will not relax back to zero magnetization when the applied magnetizing field is removed. It must be driven back to zero by the additional opposite direction magnetic field. If an alternating magnetic field is applied to the material, its magnetization will trace out a loop called a hysteresis loop [58]. The hysteresis phenomenon is similar to the C-V curve in the MIS capacitor device. When we apply a voltage in reverse, it will not fit the
28
original C-V curve measured previously. It is due to the interface traps which can trap charges to have impact on the flat band voltage and C-V curve. The C-V characteristics for hysteresis extraction were measured by sweeping the voltage from accumulation to inversion (-2V→1V) and then sweeping back (1V→-2V) at a frequency of 50kHz.
4.1 Electrical characteristics for HfO2 with nitridation plasma treatment
There are three kinds of plasma treatment with different source gas (i.e. N2, N2O, NH3) and they were treated for different process time (i.e.
30 sec, 60sec, 90sec, 120sec, 150sec and 180sec).
4.1.1 Electrical characteristics for HfO2 with N2 plasma treatment for different process time
Fig. 4-1 reveals the capacitance-voltage (C-V) characteristics of HfO2 gate dielectrics treated in N2 plasma and DC bias 50W for different process time. The capacitor treated in N2 for 120 sec shows the maximum capacitance density among these samples with different process time. In addition, other samples which treated in N2 plasma all present the larger values than the capacitors without whole plasma nitridation process. This phenomenon indicates that the N2 plasma treatment was workable to improve the capacitance. The factor of improvement might be from that the PDA process and the nitrogen incorporation in the HfO2 dielectrics,
29
which could enhance the electronic polarization as well as the ionic polarization, so the dielectric constant of the HfO2 thin films increases just as Hf-silicate thin film and SiO2 thin film. Besides, the capacitance density of the samples treated in N2 plasma for 150 sec and 180 sec are degraded. The reason could be the damage caused by the N2 plasma.
The J-V characteristics of the HfO2 capacitors treated by N2
plasma and DC bias 50W with different process time from 0V to -2V are described in Fig. 4-2. It can be observed that the samples which treated in N2 plasma all present the smaller leakage current density than the leakage current density without whole plasma nitridation process. The gate leakage current density treated in N2 for 120 sec shows the minimum current density among these conditions. The lower leakage shows that the weak structure of interface must be fixed by the plasma nitridaiton. In Fig.
4-1 and Fig. 4-2, it appears that the samples treated in N2 plasma for 120 sec display the most excellent value. While the nitridation process time is longer than 120 sec, the plasma damage from the plasma nitridation could cause the increase of the gate leakage density.
The hysteresis of C-V characteristics are shown in Figs. 4-3, 4-4, 4-5, 4-6, 4-7, 4-8, and 4-9 for the samples without treatment, and with 30, 60, 90, 120, 150, and 180 sec N2 plasma treatment , respectively. The hysteresis phenomenon of the C-V curves can be observed for all samples, which is caused by the existence of negative charges trapped in the dielectric defect states when the capacitors are stressed. The hysteresis characteristic could be improved by various plasma nitridation process. In Figs.4-3, 4-4, and 4-5, the hysteresis was a slightly reduced after N2
30
plasma treatment because the nitrogen incorporation in the HfO2
dielectrics and reduced the interface trap state, thus improving hysteresis.
In Fig.4-6, 4-7, 4-8, and 4-9, the hysteresis was a slightly enhanced more than other samples with N2 plasma treatment due to plasma damage.
However, these hysteresis voltage are very close to each other and can be acceptable.
4.1.2 Electrical characteristics for HfO2 with N2O plasma treatment for different process time
Fig4-10 reveals the capacitance-voltage (C-V) characteristics of HfO2
gate dielectrics treats in N2O plasma and DC bias 40W for different process time. The capacitor treated in N2O 10 sec shows the maximum capacitance among these samples with different process time, just like the group of N2 plasma treatment. In addition, other samples which treated in N2O plasma all present the larger values than the capacitors without whole plasma nitridation process. This phenomenon indicates that the N2O plasma treatment was workable to improve the capacitance. The factor of improvement might be from that the PDA process and the nitrogen incorporation in the HfO2 dielectrics, which could enhance the electronic polarization as well as the ionic polarization, so the dielectric constant of the HfO2 thin films increases just as Hf-silicate thin film and SiO2 thin film. Besides, the capacitance density of the samples treated in N2O plasma for 150 sec and 180 sec are degraded. The reason could be the damage caused by the N2 plasma.
31
The J-V characteristics of the HfO2 capacitors treated by N2O plasma and DC bias 40W with different process time from 0V to -2V are described in Fig. 4-11. It can be observed that the samples which treated in N2 plasma all present the smaller leakage current density than the leakage current density without whole plasma nitridation process. The gate leakage current density treated in N2O for 120 sec shows the minimum current density among these conditions. The lower leakage shows that the weak structure of interface must be fixed by the plasma nitridaiton. In Fig. 4-10 and Fig. 4-11, it appears that the samples treated in N2O plasma for 120 sec display the most excellent value. While the nitridation process time is longer than 120 sec, the plasma damage from the plasma nitridation could cause the increase of the gate leakage density.
The hysteresis of C-V characteristics are shown in Figs. 4-12, 4-13, 4-14, 4-15, 4-16, and 4-17 for the samples without treatment, and with 60, 90, 120, 150, and 180 sec N2 plasma treatment respectively. The hysteresis phenomenon of the C-V curves can be observed for all samples, which is caused by the existence of negative charges trapped in the dielectric defect states when the capacitors are stressed. The hysteresis characteristic could be improved by various plasma nitridation process.
It can be observed that the samples which treated in N2O plasma all present the smaller values than the leakage current density without whole plasma nitridation process. The hysteresis was a slightly reduced after N2 plasma treatment because the nitrogen incorporation in the HfO2
dielectrics and reduced the interface trap state, thus improving hysteresis.
32
However, these hysteresis voltage are very close to each other and can be acceptable.
4.1.3 Electrical characteristics for HfO2 with NH3 plasma treatment for different process time
Fig. 4-18 shows the capacitance-voltage (C-V) characteristics of HfO2 gate dielectrics treated in NH3 plasma and DC bias 40W for different process time. The capacitor treated in NH3 for 120sec shows the maximum capacitance density among these samples with different process times. In addition, other samples which treated in NH3 plasma all present the larger values than the capacitors without whole plasma nitridation process. This phenomenon indicates that the NH3 plasma treatment was workable to improve the capacitance. The factor of improvement might be from that the PDA process and the nitrogen incorporation in the HfO2 dielectrics, which could enhance the electronic polarization as well as the ionic polarization, so the dielectric constant of the HfO2 thin films increases just as Hf-silicate thin film and SiO2 thin film. Besides, the capacitance density of the samples treated in NH3
plasma for 150 sec and 180 sec are degraded. The reason could be the damage caused by the NH3 plasma.
The J-V characteristics of the HfO2 capacitors treated by NH3 plasma and DC bias 40W with different process time from 0V to -2V are described in Fig. 4-19. It can be observed that the samples which treated in N2 plasma all present the smaller leakage current density than the leakage current density without whole plasma nitridation process. The
33
gate leakage current density treated in NH3 for 120 sec shows the minimum current density among these conditions. The lower leakage shows that the weak structure of interface must be fixed by the plasma nitridaiton. In Fig. 4-18 and Fig. 4-19, it appears that the samples treated in NH3 plasma for 120 sec display the most excellent value. While the nitridation process time is longer than 120 sec, the plasma damage from the plasma nitridation could cause the increase of the gate leakage density.
The hysteresis of C-V characteristics are shown in Figs. 4-20, 4-21, 4-22, 4-23, 4-24, and 4-25 for the samples without treatment, and with 60, 90, 120, 150, and 180 sec NH3 plasma treatment respectively. The hysteresis phenomenon of the C-V curves can be observed for all samples, which is caused by the existence of negative charges trapped in the dielectric defect states when the capacitors are stressed. The hysteresis characteristic could be improved by various plasma nitridation process.
In Figs.4-20, 4-21, and 4-22, the hysteresis of NH3 plasma treatment for 60 sec and 90 sec were both larger than the sample of no plasma treatment. The reason possibly was the plasma process times not enough.
In Figs.4-23, 4-24, and 4-25, the hysteresis was a slightly reduced after NH3 plasma treatment because the nitrogen incorporation in the HfO2
dielectrics and reduced the interface trap state, thus improving hysteresis.
However, these hysteresis voltage are very close to each other and can be acceptable.
34
4.1.4 Short summary
Fig. 4-26 shows the capacitance-voltage (C-V) characteristics of HfO2 gate dielectrics treated in N2, N2O, and NH3 plasma at optimal condition. It is indicated that the capacitance treated in N2 plasma for 120 sec shows the most excellent value.
The J-V characteristics of the HfO2 capacitors treated in N2, N2O, and NH3 plasma at optimal condition in Fig. 4-27. It is indicated that the capacitance treated in N2O plasma for 120 sec shows the most excellent value.
By the compare of the samples which has the best capacitance in their own gas, we can realize the most suitable treatment condition which both has the best capacitance and lower leakage current. Hence, we significantly find a relative optimal condition among above discussion. It is proved that without thick oxidation layer, it can also reach the smallest leakage current when there is suitable time treatment.
4.2 Electrical characteristics for HfO2 with fluorination plasma treatment
Fig. 4-28 shows the capacitance-voltage (C-V) characteristics of Si surface (before HfO2 film deposited) treated in CF4 plasma and DC bias 20W for different process time. It indicated that the capacitance increases along with the time of plasma treatment. The capacitor treated in CF4 for 60 sec shows the maximum capacitance density among these samples with different process time. In addition, other samples which treated in
35
CF4 plasma all present the larger values than the capacitors without whole plasma fluorination process. This phenomenon indicates that the CF4
plasma treatment was workable to improve the capacitance.
The factor of improvement might be from that the PDA process and the fluorine incorporation at the interface.
The J-V characteristics of the HfO2 capacitors treated by CF4 plasma and DC bias 20W with different process time from 0V to -2V are described in Fig. 4-29. Compared with the control sample, the gate leakage current was decreased for the samples with CF4 plasma treated in 10 sec and 20 sec. The gate leakage current density treated in CF4 for 10 sec shows the minimum current density among these conditions. The lower leakage shows that the weak structure of interface must be fixed by the plasma fluorination. However, there is a non-ideal result that the gate leakage current increased for the sample with CF4 plasma treated in 30 sec and 40 sec. The reason for this phenomenon may be plasma damage and resulting in higher leakage current.
4.3 Electrical characteristics for HfO2 with dual plasma treatment for different process time
There are three kinds of plasma treatment with different source gas (i.e. CF4, N2, NH3) and they were treated for different process time (i.e.
10 sec, 20sec, 30sec, 40sec, and 120sec).
4.3.1 Electrical characteristics for HfO2 with N2 plasma
36
treatment 120 sec and CF4 plasma treatment for different process time
Fig. 4-30 reveals the capacitance-voltage (C-V) characteristics of MIS capacitor treated in CF4 for different time and N2 plasma at optimal condition. The capacitor treated in CF4 for 10 sec and N2 for 120 sec shows the maximum capacitance density among these samples with different process times. In addition, other samples which treated in CF4
plasma and N2 plasma all present the larger values than the capacitors without whole plasma process. This phenomenon indicates that dual plasma treatment was workable to improve the capacitance. The factor of improvement might be from that the fluorine and nitrogen can repair defect and dangling bonds.
The J-V characteristics of MIS capacitor treated in CF4 for different time and N2 plasma at optimal condition from 0V to -2V are described in Fig. 4-31. The gate leakage current density treated N2 plasma 120 sec shows the minimum current density among these conditions. The lower leakage shows that the weak structure of interface must be fixed by the plasma nitridaiton. However, there is a non-ideal result that the gate leakage current increased for the sample with CF4 plasma treated longer than 10 sec. The reason for this phenomenon may be plasma damage and resulting in higher leakage current.
The hysteresis of C-V characteristics are shown in Figs. 4-32, 4-33, 4-34, 4-35, and 4-36 for the samples without treatment, and with 30, 60, 90, 120, 150, and 180 sec N2 plasma treatment, respectively. The hysteresis phenomenon of the C-V curves can be observed for all samples,
37
which is caused by the existence of negative charges trapped in the dielectric defect states when the capacitors are stressed. The hysteresis characteristic could be improved by various plasma nitridation process.
It can be observed that the samples which treated in CF4 plasma and N2
plasma all present the smaller values than the leakage current density without whole plasma process. The hysteresis was a slightly reduced after dual plasma treatment because the nitrogen and fluorine incorporation in the HfO2 dielectrics and reduced the interface trap state, thus improving hysteresis. However, these hysteresis voltage are very close to each other and can be acceptable.
4.3.2 Electrical characteristics for HfO2 with NH3 plasma treatment 120 sec and CF4 plasma treatment for different process time
Fig. 4-37 reveals the capacitance-voltage (C-V) characteristics of MIS capacitor treated in CF4 for different time and NH3 plasma at optimal condition. The capacitor treated in CF4 for 10 sec and NH3 for 120 sec shows the maximum capacitance density among these samples with different process times. In addition, other samples which treated in CF4 plasma and NH3 plasma all present the larger values than the capacitors without whole plasma process. This phenomenon indicates that dual plasma treatment was workable to improve the capacitance. The factor of improvement might be from that the fluorine and nitrogen can repair defect and dangling bonds.
38
The J-V characteristics of MIS capacitor treated in CF4 for different times and N2 plasma at optimal condition from 0V to -2V are described in Fig. 4-38. The gate leakage current density treated NH3 plasma 120 sec shows the minimum current density among these conditions. The lower leakage shows that the weak structure of interface must be fixed by the plasma nitridaiton. However, there is a non-ideal result that the gate leakage current increased for the sample with CF4 plasma treated longer than 10 sec. The reason for this phenomenon may be plasma damage and resulting in higher leakage current.
The hysteresis of C-V characteristics are shown in Figs. 4-39, 4-40, 4-41, 4-42, and 4-43 for the samples without treatment, and with 30, 60, 90, 120, 150, and 180 sec NH3 plasma treatment respectively. The hysteresis phenomenon of the C-V curves can be observed for all samples, which is caused by the existence of negative charges trapped in the dielectric defect states when the capacitors are stressed. The hysteresis characteristic could be improved by various plasma nitridation process.
It can be observed that the samples which treated in CF4 plasma and NH3
plasma all present the smaller values than the leakage current density without whole plasma process. The hysteresis was a slightly reduced after dual plasma treatment because the nitrogen and fluorine incorporation in the HfO2 dielectrics and reduced the interface trap state, thus improving
plasma all present the smaller values than the leakage current density without whole plasma process. The hysteresis was a slightly reduced after dual plasma treatment because the nitrogen and fluorine incorporation in the HfO2 dielectrics and reduced the interface trap state, thus improving