Chapter 3 Investigation of the Characteristics of 3D-Stacked Single Grain Boundary
3.2 Electrical Characteristics of 3D-SSGB-TFTs Fabricated by Elevated Channel
3.2.2 Electrical Characteristics of SGB-TFTs in Individual Layers of
3.2.2.1 Electrical Characteristics of SGB-TFTs
It has been demonstrated that large and longitudinal grains could be formed in the channel region by the elevated channel method in section 2.2. The grain structure would have a significant influence on the electrical characteristics of the fabricated TFTs. In section 2.4.2, it was found that the SGB channels could be achieved when the channel length was 1μm with the laser energy density was from 460mJ/cm2 to 500mJ/cm2 and the number of laser shots was 20 (ie. 95% overlapping). As a result, the optimized laser process conditions in this thesis for SGB-TFTs with the channel length of 1μm would be thought as the laser energy density of 460mJ/cm2 and 20 laser shots due to the concern of laser energy efficiency. Besides, there were four kinds of structure conditions for each type of SGB-TFTs in the 3D-SSGB TFTs since there were both two conditions for separation oxide thicknesses and dopant arrangements.
Fig. 3-2 and Fig. 3-3 showed the typical transfer characteristics of n-channel SGB-TFTs with both channel length and width of 1μm, in which the thickness of gate oxide was 1000Å. Fig. 3-4 and Fig. 3-5 showed the typical transfer characteristics of p-channel SGB-TFTs with both channel length and width of 1μm, in which the thickness of gate oxide was 1000Å. The laser process conditions were optimized. The threshold voltage was defined as the gate voltage required to achieve a normalized drain current of I
d = (W/L)×10-8A at |V
ds| =0.1V. The subthreshold swing was defined as the constant minimum value ofthe subthreshold curve before the curve takes off, and the equivalent was d(log Id)/d(Vg) at |V
ds| =0.1V. The equivalent field-effect mobility was extracted from the maximum transconductance in the linear region of I
d-V
g characteristics at |V
d| = 0.1V (i.e., the formula of μ=g
m/[(W/L)V
dsCox] ). The on/off current ratio was defined as the ratio of maximum drain current over minimum drain current at |V
d| =3V. Several important electrical characteristics of the SGB-TFTs were summarized in Table 3-1 and Table 3-2 for channel type of N-type and P-type, respectively.
It is notable that the definition of field-effect mobility of double gate TFTs was analogous to HARA et al. in this thesis [3.1]. The definition of field-effect mobility could be approached by two aspects. One was the physical meaning of carrier transport capability of polycrystalline silicon layer, which was defined by the following equation:
μ = g
m / [ (2W/L)．Cox．Vds ] ……….(1),
while the channel width was 2W due to the number of field induced channel was two.
The other definition was corresponding to the carrier transport capability in the occupied area of active region, which was defined by the following equation:
μ* = g
m / [ (W/L)．Cox．Vds ]………(2),
while the channel width was W because of the occupied width of active region was only W.
In this thesis, the latter definition was adopted; the name of “equivalent field-effective mobility” and the symbol of “μ*” was used to avoid the confusion with the first definition.
According to Fig. 3-2~Fig. 3-5 and Table 3-1~Table 3-2, both the n-channel and p-channel SGB-TFTs exhibited good electrical characteristics in all kinds of structure conditions for the 3D-SSGB TFTs. Take the dimension of L=W=1μm for example, n-channel SGB-TFTs with equivalent field-effect mobility of about 320 cm2/V-s could be achieved by using the elevated channel method while the mobility of conventional SPC n-channel TFTs having the same device dimension as the SGB-TFTs was only about 34 cm2/V-s. In the same dimension as above, p-channel SGB-TFTs with equivalent field-effect mobility of about 140 cm2/V-s could be achieved by using the elevated channel method while the mobility of conventional SPC p-channel TFT having the same device dimension as the SGB-TFTs was only about 12 cm2/V-s. Furthermore, In L=W=1μm devices, we obtained subthreshold swing of the n-channel SGB-TFTs about 0.45V/decade, while that of conventional SPC n-channel TFTs having the same device dimension as the SGB-TFTs was only about 2.3V/decade. Similarly, the on/off current ratio of the n-channel and p-channel SGB-TFTs with L=W=1μm was about 107 to 108, while that of the conventional SPC TFTs was under 107. It was found that SGB-TFTs with the same dopant type showed almost the same well transfer characteristics no matter what layer they were in and what the thickness of separation oxide was, which corresponded to the results of SEM analysis in section 2.4.2.
Fig. 3-6~Fig. 3-9 displayed the typical output characteristics of the SGB-TFTs with device layer and dopant type of bottom layer/N-type, top layer/N-type, bottom layer/P-type, and top layer/P-type, respectively. It was found again that SGB-TFTs with the same dopant type exhibited almost the same well output characteristics no matter in what kind of structure condition. Therefore, high-performance SGB-TFTs in both the top and bottom layers of the 3D-SSGB-TFTs were successfully achieved due to the good crystallinity SGB silicon
channel crystallized by the elevated channel method. Since the electrical characteristics of the same type SGB-TFTs in different structure conditions was very close to each other, the uniformity of SGB-TFTs in different device layers of 3D-SSGB-TFTs would be discussed in the next section.
3.2.2.2 Uniformity Investigation of SGB-TFTs
Fig. 3-10 and Fig. 3-11 showed the dependence of field-effect mobility on the device dimension for TFTs in different device layers of 3D-Stacked structure with different separation oxide thicknesses and dopant arrangements. Twenty TFTs for each device dimension were measured with optimal laser irradiation condition for maximum field-effect mobility to investigate the device-to-device uniformity. The vertical bars in the figures indicated the minimum and maximum characteristic values obtained at the specific laser energy density, and the symbols were the average calculated characteristic values. The mobility of each type of TFTs increased as the channel length decreased, which indicated that the grain boundaries perpendicular to the direction of current flow acted as strong trapping centers which degraded the performance of TFTs resulting from grain boundary potential barrier height [3.2]-[3.4]. Besides, the field-effect mobility decreased significantly when the device dimension was over 1.5μm, indicating that channel lengths above 1.5μm were too long to achieve the SGB by elevated channel method as mentioned in section 2.4.2.
Fig. 3-12~Fig. 3-17 displayed the dependence of several electrical characteristics, including the field-effect mobility, the subthreshold swing, and the threshold voltage, on laser energy densities for the SGB-TFTs with L=W=1μm. For the same dopant type SGB-TFTs in different structure conditions, the dependence of all the electrical characteristics as mentioned above on laser energy density was very weak and similar to each other. For the N-type SGB-TFTs in different structure conditions, the mobility, the subthreshold swing, and the threshold voltage were always around 300cm2/V-s, 0.5V/decade, and -1V. And for the P-type SGB-TFTs in different structure conditions, the mobility, the subthreshold swing, and the threshold voltage were always around 140cm2/V-s, 0.7V/decade, and -3V.
Therefore, excellent device-to-device uniformity was achieved due to the large process window of the elevated channel method (460mJ/cm2 ~ 500mJ/cm2 at least), which was shown in different structure conditions of 3D-SSGB-TFTs. What is more, take the dimension of L=W=1μm with optimized laser process conditions for example, Fig.
3-18~Fig. 3-23 showed the comparison of equivalent field-effect mobility, subthreshold swing, and threshold voltage for the N-type and P-type SGB-TFTs respectively.
Twenty-two SGB-TFTs were measured to investigate the device-to-device variation for each structure condition, then twenty of which were chosen for the exclusion of unexpected extreme value under statistics. Referring to Table 3-3 and Table 3-4, the values of the average and standard deviation of each electrical characteristic of the same type SGB-TFTs in different structure conditions were almost the same. For the N-type SGB-TFTs in different structure conditions, average values of the mobility, the subthreshold swing, and the threshold voltage were always around 313cm2/V-s, 0.52V/decade, and -0.73V. And for the P-type SGB-TFTs in different structure conditions, average values of the mobility, the subthreshold swing, and the threshold voltage were always around 140cm2/V-s, 0.6V/decade, and -3V. Moreover, for each electrical characteristic of the same type SGB-TFTs in different structure conditions, the standard deviation values were always much smaller than the average values. For the N-type SGB-TFTs in different structure conditions, standard deviation values of the mobility, the subthreshold swing, and the threshold voltage were always around 10cm2/V-s, 0.08V/decade, and 0.26V. And for the P-type SGB-TFTs in different structure conditions, standard deviation values of the mobility, the subthreshold swing, and the threshold voltage were always around 4cm2/V-s, 0.16V/decade, and 0.5V. As a result, we could conclude that high performance multi-layer SGB-TFTs with good uniformity could be fabricated easily using elevated channel method in different structure conditions of the 3D-SSGB TFTs.