Experimental Study of Physical-Vapor-Deposited Titanium Nitride Gate with An n⁺-Polycrystalline Silicon Capping Layer and Its Application to 20 nm Fin-Type Double-Gate Metal–Oxide–Semiconductor Field-Effect Transistors

Abstract

We have comparatively investigated the electrical characteristics including threshold voltage (V th ) variability and mobility by fabricating n þpolycrystalline silicon (poly-Si) gate and physical-vapor-deposited (PVD) titanium nitride (TiN) gate fin-type double-gate metal-oxidesemiconductor field-effect transistors (FinFETs), and demonstrated 20-nm-thick PVD-TiN gate FinFETs with a symmetrical V th . It is experimentally found that the gate stack of a 20-nm-thick PVD-TiN layer capped with a 100-nm-thick n þ -poly-Si layer is very effective for setting a symmetrical V th for undoped FinFETs keeping almost the same V th variability and mobility as those in the case of the n þ -poly-Si gate only. On the other hand, mobility degradation was observed in the case of pure 50-nm-thick PVD-TiN gates. These results indicate that mobility degradation probably caused by the thick metal gate induced mechanical stress can be effectively suppressed by reducing the PVD-TiN thickness to 20 nm or less. #

Suppression of threshold voltage variability of double-gate fin field-effect transistors using amorphous metal gate with uniform work function

“…1,20 The V t variation increases with the gate size scaling. As commonly reported, 1,7,8,14,15,[19][20][21] the linear relationship is recognized for both the MG cases. The linear relationship in the Pelgrom plot shows that the variability is caused by the gate stack origins such as the WFV and is not due to the SCE variation caused by the LER according to Ref.…”

“…Since the constant current criteria is the most commonly used technique for V t definition 1 and has been conveniently used for measuring V t of huge number of transistors due to its simplicity, 17,18 the constant current criteria are appropriate for benchmarking of the V t variability with regard to the previously reported V t variability of FinFETs. 7,8,14,19,21,22 As clearly seen in Figs. 2 and 3, the TaSiN MG exhibits smaller V t fluctuation than the TiN MG does both for the long and short channel cases.…”

“…The TaSiN film was confirmed to be amorphous by X-ray diffraction (XRD) analysis and plane-view transmission electron microscopy (TEM) after rapid thermal annealing at 860 C for 2 s equivalent to the thermal budget of source/drain (S/D) dopant activation. A poly-crystalline TiN gate, which is commonly investigated as a gate-first MG, [13][14][15][16] was also used for the comparison. These MG FinFETs were fabricated through the gate-first process as follows.…”

“…As clearly seen, the TaSiN MG significantly suppresses the slope value in comparison to the TiN MG case and exhibits the A Vt value of 1.34 mV lm, which is the smallest one reported so far for MG-FinFETs. 7,8,14,19,21,22 Validity of the constant current criteria for the V t variability analysis is discussed as follows. We have previously confirmed that the V t variation measured by the constant current criteria is independent from the fluctuation of the parasitic resistance.…”

Drift-Diffusion Versus Monte Carlo Simulated ON-Current Variability in Nanowire FETs

“…Furthermore, the MGG profile is characterized by a grain size (GS) and by a work function value (WFV) [48]. For the current study, we have chosen the titanium nitride (TiN) which is commonly used as a gate material [51]. The metal has experimentally observed WFVs of 4.6 eV and 4.4 eV with a probability of 60% and 40% formation, respectively [52].…”