Post-Deposition N 2 O Plasma Treatment
2.3 Results and Discussion
2.3.4 Current Transport Mechanism
Using the carrier separation method, the carrier type is investigated for the fresh devices.
The carrier of gate leakage can be separated into holes and electrons. Figs. 2-16 (a) and (b) are carrier separation results for the control sample under inversion and accumulation regions, respectively. It is shown that the S/D current dominates the gate leakage for the inversion region; while the substrate current dominates the gate leakage for the accumulation region.
The carrier separation results for the N2O- plasma-treated sample are shown in Fig. 2-17. The
case for the accumulation region is similar to the control sample, i.e., electrons from the substrate dominate the gate leakage. However, the case for the inversion region is different from the control sample, where ISD is suppressed by almost 1.5 order and the IB remains almost unchanged. These trends can be explained by the band diagram shown in Fig. 2-18. In the inversion region, the S/D current (hole current) is formed by the carrier in the inversion layer, whereas the substrate current IB originates from electrons tunneling from the gate terminal. In addition, we can see that the magnitude of the leakage current in accumulation is about two orders larger than that in inversion. The tendency is ascribed to the asymmetric band as shown in Fig. 2-18. In inversion mode, both electrons from the poly gate electrode and holes from the inversion layer tunnel through the gate stack. On the other hand, in accumulation mode, electrons from the substrate only face the tunneling barrier of ~1nm interfacial oxide layer. Therefore, the asymmetric band diagram provides a plausible explanation why the leakage current in accumulation is much larger than that in inversion.
Fig. 2-19 (a) and Fig. 2-20 show gate current Ig as a function of Vg for the HfO2/SiON gate stacks measured at several different temperatures up to 125℃ in the inversion and accumulation regions, respectively. All currents are dependent on temperature irrespective of Ig at the range of Vg=0V to -3.5V. To determine the conduction mechanism for samples with and without N2O plasma treatment, numerical fitting was conducted. Base on the equation of Frankel-Poole (F-P):
height, E is the electric field in the HfO2 film, ε0 is the free space permittivity, εk is the dielectric constant of HfO2, k is Boltzmann constant, T is the temperature measured in Kelvin.
Excellent fitting curves are shown in the plot of ln J E
⎛ ⎞⎜ ⎟
⎝ ⎠ versus E0.5of IG, indicating that the conduction mechanisms both under gate and substrate injections for the SiON/HfO2
gate stacks, with and without post-N2O plasma treatment, are F-P type in nature. The barrier heightΦB and the dielectric constantεk of SiON/HfO2 gate stacks can be calculated from the intercept of y axis and the slope of the fitting curves according to (2.6). The εk value is found to be around 16 for the control sample and around 17 for the sample with N2O treatment. The value is very close to the estimated value from HRTEM image, which is 13.4 for control sample and 14.6 for the sample with N2O treatment. In the high-k gate dielectric, two current components could dominate the gate leakage current. Recalling from the carrier separation results, ISD current is mainly contributed by the carrier in the inversion layer, i.e., holes, while IB current is mainly contributed by the minority carrier in the gate electrode, i.e., electrons. Fig. 2-21 shows the F-P plot for the source/drain current in the inversion region. Fig.
2-22 shows the F-P plot for the substrate current in the inversion region. The solid lines are fitting curves for all temperatures. An excellent fitting can be obtained, indicating that the conduction mechanism for the high voltage ISD and IB are indeed Frenkel-Poole-type in nature.
The fitting parameters for the hole and electron barrier heights are 1.17eV and 1.05eV, respectively, for the control sample. Results for post-N2O plasma samples are shown in Fig.
3-23 and Fig. 3-24. Again, good fitting curves can be seen, indicating that F-P is the right mechanism for both ISD and IB. The barrier heights are 1.14eV and 1.18eV for electrons and holes, respectively, for the N2O-treated sample. Note that the barrier height for electrons has changed from 1.05eV for the control to 1.14eV for the N2O-treated sample, indicating that the trap position has moved closer to the conduction band of the poly Si gate after post-N2O plasma treatment. The band diagrams are shown in Fig. 2-26 and Fig. 2-25 for the sample
with and without post-N2O plasma treatment, respectively. The effective barrier heights for IG
in the accumulation region are 1.389eV and 1.32eV for samples with and without N2O treatment, respectively.
2.4 Summaries
Improvements in the electrical characteristics of the p+-poly gate pMOSFETs with HfO2/SiON gate stacks by post-deposition N2O plasma treatments have been demonstrated in this work. We have found that improvements are achieved in many aspects, such as reduced leakage current, better subthreshold swing, enhanced normalized transconductance, and higher driving current. These improvements are ascribed to the lower interface states and bulk traps, as confirmed by various types of charge pumping measurement. Although charge pumping is a powerful technique to evaluate the interface state densities and bulk traps present in high-k films, care attention should be paid when dealing with dielectric that depicts larger gate leakage current, since the charge pumping current will be severely influenced in such case. Note that this phenomenon is entirely different for th nMOSFETs, where the leakage current will contribute to the source/drain current, but not the substrate current. In evaluation of the reliability, we found that the degradation caused by the voltage stress is dominated by the charge trapping in the bulk of HfO2 films, rather than interface state generation, irrespectively of whether post-N2O plasma treatment was performed or not. In addition, it was observed that the electron is the main trapped species during stressing for the N2O-treated samples, which is very different from the hole trapping observed in samples without post-N2O plasma treatment.