Accurate Monte Carlo Simulation of Ion Implantation into Arbitrary 1D/2D/3D Structures for Silicon Technology
- PDF / 284,708 Bytes
- 6 Pages / 595 x 842 pts (A4) Page_size
- 17 Downloads / 251 Views
C6.5.1
Accurate Monte Carlo Simulation of Ion Implantation into Arbitrary 1D/2D/3D Structures for Silicon Technology Shiyang Tian1 , Victor Moroz2 , and Norbert Strecker2 1
2
Synopsys, Inc., 14911 Quorum Dr., Dallas, Texas 75254 Synopsys, Inc., 700 East Middlefield Road, Mountain View, CA 94043
ABSTRACT We present an integrated Monte Carlo implant simulator which is capable of accurately simulating ion implantation into any amorphous materials and crystalline Si for 1D/2D/3D structures with arbitrary geometry and topography. With this simulator, we investigate some practical examples which reveal interesting 2D/3D effects, and demonstrate the importance of channeling for sub-100nm silicon technology. INTRODUCTION The predictive power of Monte Carlo (MC) ion implantation simulations has been well recognized in the microelectronics industry through the widespread use of the popular Monte Carlo implant simulators such as ut-marlowe [1], imsil [2], and Crystal-trim [3]. However, the full potential of MC implant has not been realized partly because these simulators were designed primarily for 1D structures, or limited to 2D structures [4], or 2D/3D structures with limited capability [5, 6]. As the industry moves toward high-tilt implants (such as halo implants), and 3D structures (such as FinFET), coupled with “diffusion-less” (< 3 nm) activation for 65 nm node and beyond [7], accurate prediction of the dopant placement is becoming vitally important. On the other hand, stand-alone MC simulators are not suitable for implants with complex geometry and topography. In order to overcome this problem, as part of our process simulator Taurus-Process [8] we have developed an integrated MC implant simulator, which is capable of simulating the implants into any amorphous materials and crystalline Si with arbitrary geometry and topography. It was shown that with the same set of parameters for each ion-target combination our model is capable of predicting dopant distributions from sub-keV to 10 MeV for common dopant species, and is valid for different implant directions including random direction, , , and channeling directions [9]. We implemented trajectory split and lateral trajectory replication algorithms which greatly reduce the simulation time. With the improved efficiency and seamless interface with diffusion simulations, MC simulation as a better alternative to analytic implant in a complete process flow becomes realistic for low to medium energy implants. The goal of this paper is to explore in the predictive mode some 2D and 3D effects that can not be measured directly, but may be of critical importance for the 65-nm technology node and below. MODEL DESCRIPTION Our model is based on the classical binary collision approximation (BCA) such as that used by trim for amorphous materials and ut-marlowe for crystalline silicon. The basic idea of MC simulation is to follow the motion of a large number of individual ions in a target. The ion loses its energy as a result of nuclear and electronic stopping which are assumed to be
Data Loading...
