The temperature dependence of the SBH has been mentioned in Section 4-1, and both the Ni/n-SiC and NiSi/n-Si junctions showed the temperature dependence of the SBH. However, the temperature dependence of the SBH is due to the IFBL effect, and the extracted SBHs without IFBL effect at different temperature are the same. Such issue has been discussed before, but most of the literatures extracted the SBH with considering the IFBL effect [57-60]. In other words, the temperature dependence of the SBH without considering the IFBL effect, the “true” temperature dependent SBH, is still unclear. By utilizing the extraction procedure with the TFE model in Section 4-2.2, we observed the temperature dependence of the SBH without considering the IFBL effect in the low temperature measurement, and following were the details. This section only discussed the PtSi/p-Si and NiSi/p-Si Schottky junctions and those with the BF2+ ion implantation to a dose of 11012 cm-2, and the extracted results are shown in Fig.4-11.
The cases without BF2+ ion implantation are discussed at first. For PtSi/p-Si, the possible reason why the BH0s lowered down from 0.25 to 0.17 eV as the temperature decreasing may be due to the lowering of the Fermi-level before the metal contacted.
Fig.4-12 shows that the interface state density is high near the valence band-edge and this is consistent for most cases. As the temperature decreasing, the Fermi energy was lowered down and the Fermi-level approached to the energy level of high interface
60
state density, and the donor-like states above Fermi-level would be depleted. The charge neutrality level (CNL) is close to the valance band. After the high work function metal contacted, electrons transfer from semiconductor side to metal side.
Because the high density of the donor-like states around the CNL, the pinning level is very close to the original CNT. Therefore, the SBH is low. The illustration is shown in Fig.4-13. The depleted donor-like states and the congregated electrons in the metal formed a dipole layer, and this dipole layer would lower down the SBH. However, the Fermi energy is higher and CNL is at the low interface state density at high temperature. After electrons transfer to metal, the pinning level is higher than that at low temperature, so is the lowering amount of the SBH. The reason why the temperature dependence of the NiSi/p-Si is not as apparent as PtSi/p-Si is because that the original pinning Fermi-level is at the energy level of low interface state density and the work function of NiSi is also low, so the charges in the dipole are only a little.
Following is the discussion of the cases with BF2+
ion implantation to a dose of 1
1012 cm-2. After the ion implantation the SBH of PtSi/p-Si is increased, the possible reason is that the distribution of interface states or the crystal defects of the silicon surface layer may be differed [61-63]. However, the SBH of NiSi/p-Si junction is lowered down after the ion implantation, and we believed it would be caused by the IFBL effect due to the addition carriers in the surface by comparing the results in Section 3-2.2. The high slope region in the log(J)-V curve due to the ion implantation was not measured, so the lowering amount due to the IFBL effect could not be separated from the SBH.61
4-4 Summary
The extraction procedure with the TFE model proposed in Chapter 2 is accurate from high to low SBH cases. High SBH cases of metal/SiC Schottky contacts are extracted correctly, and they are all consistent with the SBH reported in literature: the SBHs of Ni/n-SiC and Ti/n-SiC Schottky contacts are extracted as 1.42 eV and 0.85 eV. Middle SBH case of the NiSi/n-Si Schottky contact with carbon ion implantation is extracted as 0.65 eV which is also consistent with the literature. Moreover, the effect of carbon ion implantation is confirmed to act as the IFBL effect like what implanted donors act in the Schottky junction. Low SBH case of the NiGe/n-Ge Schottky contact is extracted as 0.47 eV, and the extraction with the TE model is believed to be inappropriate due to the raising of the non-ideality factor. Above all, the extraction procedure with the TFE model is also accurate for the Schottky junctions on different semiconductor materials.
For the exact low SBHs without BF2+ ion implantation cases, the extracted SBH of PtSi/p-Si ranges from 0.17 to 0.25 eV and the extracted SBH of NiSi/p-Si ranges from 0.42 to 0.45 eV. The exact low SBH without BF2+ ion implantation cases are extracted correctly with the TFE model. For the low SBHs with BF2+
ion implantation to a dose of 11012 cm-2 cases, the extracted SBHs of PtSi/p-Si are about 0.27 eV which are higher than that without BF2+
ion implantation by 0.02 eV, and the extracted SBHs of NiSi/p-Si ranged from 0.4 to 0.43 eV which are lower than that without BF2+
ion implantation. For the cases to higher doses, the dominant conducting mechanism is no long the TFE model, but the trend still can be observed: the rectifying effect is weaker as the implantation dose increases, and the Schottky junctions become purely ohmic with the BF2+ ion implantation to a dose of 31013 cm-2.
62
In section 4-3, the discussion of temperature dependence explains that the lowering of the BH0 is caused from the dipole layer, and the dipole layer is formed because of the depletion of the donor-like states and the congregation of the electrons in the metal. The SBH of PtSi/p-Si junction raises after the BF2+
ion implantation to a dose of 1
1012 cm-2, and it is believed that the distribution of interface states or the crystal defects of the silicon surface layer has been changed, so the SBH is pinned at different energy level. The SBH of NiSi/p-Si junction lowers down after the BF2+ ion implantation to a dose of 110
12 cm-2, and it is believed to be caused from the IFBL effect. The high slope region in the log(J)-V curve could not be measured, so the lowering amount of the IFBL effect could not be separated from the BH0.63
Table 4-1 Traditional TE extraction results of PtSi/p-Si Schottky junction without BF2+
ion implantation.
TE 100 K 125 K 150 K 175 K 200 K 225 K
BH (eV) 0.161 0.188 0.203 0.239
NA NA
N 1.96 2.12 1.14 ~1
NA NA
Table 4-2 TFE extraction results of PtSi/p-Si Schottky junction without BF2+
ion implantation.
TFE 100 K 125 K 150 K 175 K 200 K 225 K
BH0 (eV) 0.171 0.191 0.200 0.232 0.245 0.246
BH (eV) 0.161 0.181 0.191 0.221 0.233 0.232
N (cm-3) 1.771015 1.89
1015 1.881015 3.291015 4.18
1015 4.591015Table 4-3 Traditional TE extraction results of NiSi/p-Si Schottky junction without BF2+ ion implantation.
TE 125 K 150 K 175 K 200 K
BH (eV) 0.390 0.408 0.406 0.421
n ~1 1.07 1.01 1.02
TE 225 K 250 K 275 K 300 K
BH (eV) 0.424 0.437 0.436 0.448
n ~1 ~1 1.02 1.11
64
Table 4-4 TFE extraction results of NiSi/p-Si Schottky junction without BF2+ ion implantation.
TFE 125 K 150 K 175 K 200 K
BH0 (eV)
N.A.
0.426 0.425 0.423BH (eV)
N.A.
0.404 0.410 0.411N (cm-3)
N.A.
6.291015 4.091015 1.88
1015TFE 225 K 250 K 275 K 300 K
BH0 (eV) 0.433 0.447 0.441 0.449
BH (eV) 0.419 0.432 0.427 0.435
N (cm-3) 2.661015 3.83
1015 3.111015 3.37
1015Table 4-5 TFE extraction results of PtSi/p-Si Schottky junction with BF2+
ion implantation at a dose of 1
1012 cm-2.TFE 175 K 200 K 225 K 250 K
BH0 (eV) 0.263 0.269 0.263 0.268
BH (eV) 0.249 0.256 0.252 0.258
N (cm-3) 6.641015 5.22
1015 3.831015 3.27
1015Table 4-6 TFE extraction results of NiSi/p-Si Schottky junction with BF2+
ion implantation at a dose of 1
1012 cm-2.TFE 225 K 250 K 300 K
BH0 (eV) 0.404 0.415 0.425
BH (eV) 0.387 0.399 0.415
N (cm-3) 6.311015 5.93
1015 1.61101565
Fig.4-1 Electrical characteristics of Ni/n-SiC Schottky junction are measured at temperature of 423 and 448 K, including the extraction results by both TE and TFE model.
Fig.4-2 Electrical characteristics of Ti/n-SiC Schottky junction are measured at temperature of 300 K, including the extraction results by both TE and TFE model.
66
Fig.4-3 Electrical characteristics of NiSi/n-Si Schottky junction with carbon ion implantation are measured at temperature of 300, 325, and 325 K, including the extraction results by both TE and TFE model.
0 100 200
Fig.4-4 SIMS measurement before silicide formation shows the high concentration of carbon ion near the Schottky junction. SRP measurement shows that the carrier concentration near the junction is only about 1018 cm-3.
67
Fig.4-5 Electrical characteristics of NiGe/n-Ge Schottky junction are measured at temperature of 300 K, including the extraction results by both TE and TFE model.
-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5
Fig.4-6 Electrical characteristics of PtSi/p-Si Schottky junction without ion implantation are measured at the temperature from 100 to 225 K.
68
Fig.4-7 Electrical characteristics of NiSi/p-Si Schottky junction without ion implantation are measured at the temperature from 125 to 300 K.
-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 with BF2 ion implantation Dose=1012 cm-2
Fig.4-8 Electrical characteristics of PtSi/p-Si Schottky junction with BF2+
ion implantation at a dose of 1
1012 cm-2 are measured at the temperature from 100 to 225 K.69
with BF2 ion implantation Dose=1012 cm-2 with BF2 ion implantation Dose=6x1012 cm-270
with BF2 ion implantation Dose=6x1012 cm-271
with BF2 ion implantation Dose=3x1013 cm-2100 120 140 160 180 200 220 240 260 280 300 320 0.15
PtSi/p-Si with BF2+ ion implantation NiSi/p-Si with BF2+ ion implantation
Fig.4-11 The temperature dependence of the BH0 without considering the IFBL effect.
72
Fig.4-12 Illustration of the Fermi energy level and the donor-like interface state density before metal contact.
Fig.4-13 Illustration of the charge distribution in the Schottky junction.