Use of Small Gate Voltage Pulses for the Extraction of the Interface Trap Densities in MOS Structures Using The Charge P
- PDF / 112,004 Bytes
- 6 Pages / 612 x 792 pts (letter) Page_size
- 85 Downloads / 164 Views
Use of Small Gate Voltage Pulses for the Extraction of the Interface Trap Densities in MOS Structures Using The Charge Pumping Technique Eugène Lin, Eric Moussy, and Daniel Bauza, Institut de Microélectronique, Electromagnétisme et Photonique, UMR CNRS 5130, INPG, ENSERG, 23, rue des Martyrs, BP 257, 38016 Grenoble Cedex 1, FRANCE. ABSTRACT A general model for charge pumping is derived and is used for extracting interface trap densities using small gate voltage pulses. In MOS structures with ultrathin oxides, this strongly reduces the leakage current and prevents the oxide from any degradation. Interface trap densities from devices with oxides from 2.3 to 1.3 nm thick are reported for the first time. INTRODUCTION Charge pumping (CP) is the leading interface trap characterization technique but as for all the other methods, the measurement of interface trap densities, Dit, on MOS structures with ultrathin oxides is difficult or even unfeasible as soon as the oxide thickness is few nanometer thick, due to currents flowing through the oxide. Around 3-4 nm, the gate bias excursion allowed is reduced due to Fowler-Nordheim injection that makes the measurements difficult and produces degradation of both the Si-SiO2 interface and the oxide bulk. Below 2.5 nm direct tunneling also occurs. In conventional CP, Dit is measured using gate bias swing, Vsw, values much greater than (Vth – Vfb), Vth and Vfb being the device threshold and flat band voltages [1]. In this article, the great interest of using small gate pulses is demonstrated.
A GENERAL MODEL FOR CHARGE PUMPING When repeatedly switching a MOS transistor between strong inversion to strong accumulation, the energy at the Si-SiO2 interface in the upper and lower half of the band gap, Eem,e and Eem,h, where emission occurs is given by [2]:
E i - E em, e (0)
kT
E em, h (0) − E i
kT
E em,e (t em, e ) − Ei = −kTln (n iσ n v th t em,e + exp(
and
E em, h (t em, h ) − E i = + kTln (niσ p v th t em, h + exp(
))
(1),
))
(2).
Ei and ni are the intrinsic level and carrier density, σn and σp and tem,e and tem,h are the trap cross section and the emission time, for electrons and holes, vth is the carrier thermal velocity, Eem,e(0) and Eem,h(0) being the onset of non steady state electron and hole emission [2,3]. B7.3.1
At large Vsw, Eem,e(0) and Eem,h(0) are such that the variation with time of the quasi-Fermi level of the traps, Et, and of the Fermi level, Ef, at the interface, are equal [4]:
and
dE f dE t E - Ei = = − kTe n = − kTn iσ n v th exp( t ) dt dt kT
(3),
dE f dE t E -E = = kTe p = kTn iσ p v th exp( i t ) dt dt kT
(4).
Similarly, the onset of steady state capture for electrons or holes, Ecap,e or Ecap,h, is such that the variation with time of Et and Ef are equal [4]:
and
dE f dE t E - Ei = = − kTc n = − kTn iσ n v th exp( t ) dt dt kT
(5),
dE f dE t E - Et = = kTc p = kTn iσ p v th exp( i ) dt dt kT
(6).
From equations 3 and 5, Eem,e(0) = Ecap,e, and from equations 4 and 6, Eem,h(0) = Ecap,h. In the literature, these q
Data Loading...
