Random Dopant Distribution Research Articles

Discrete Dopant Fluctuations in 20-nm/15-nm-Gate Planar CMOS

We experimentally quantified, for the first time, the random dopant distribution (RDD)-induced threshold voltage standard deviation up to 40 mV for 20-nm-gate planar complementary metal-oxide-semiconductor (CMOS) field-effect transistors. Discrete dopants have been statistically positioned in the 3-D channel region to examine the associated carrier transportation characteristics, concurrently capturing ldquodopant concentration variationrdquo and ldquodopant position fluctuation.rdquo As the gate length further scales down to 15 nm, the newly developed discrete dopant scheme features an effective solution to suppress the 3-sigma-edge single-digit dopant-induced variation by the gate work function modulation. The results of this paper may postpone the scaling limit projected for planar CMOS.

High-performance yellow ceramic pigments Zr(Ti 1− x− ySn x− yV yM y)O 4 (M = Al, In, Y): Crystal structure, colouring mechanism and technological properties

Zirconium titanate-stannate doped with V with co-dopants Al, In or Y was synthesised by solid state reaction and its structural (XRD, SEM), optical (DRS) and technological properties were determined to assess its potential use as ceramic pigment. These compounds have a srilankite-type, disordered orthorhombic structure, implying a random distribution of Zr, Ti, Sn and dopants in a single, strongly distorted octahedral site. Doping caused an increase of unit-cell dimensions, metal–oxygen distances and octahedron distortion. Optical spectra show crystal field electronic transitions of V 4+ as well as intense bands in the blue-UV range due to V 4+–V 5+ intervalence charge transfer and/or to V–O charge transfer. The formation of oxygen vacancies is supposed to compensate the occurrence of V 4+ ensuring the lattice charge neutrality. These srilankite-type oxides develop a deep and brilliant yellow shade with colourimetric parameters close to those of industrial ceramic pigments. Technological tests in several ceramic applications proved that zirconium titanate-stannate is very stable at high temperature, exhibiting an excellent performance in the 1200–1250 °C range, even better than praseodymium-doped zircon.

Reduced backscattering in potassium-doped nanotubes:Ab initioand semiempirical simulations

By means of combined ab initio and semiempirical simulations, the charge transmission properties of potassium- and nitrogen-doped nanotubes are compared. The backscattering efficiency of adsorbed potassium is shown to be much weaker, owing to a shallower impurity potential well that weakly traps the induced quasibound states. The difference in conductance and elastic mean free paths is further investigated for nanotubes with lengths of several tens of micrometers and random distribution of dopants. Our results provide clear criteria for favoring doping by physisorption instead of chemical substitutions.

On the statistical background of the correlation of the magnetic and electrical properties as well as of the creation of a spin-glass state in spinels with chromium

The relation of the magnetic coupling constants and the random distribution of dopants and defects under the percolation threshold is studied. The relation of the 2nd non-linear static magnetic susceptibility and the 2nd harmonic of the dynamic magnetic susceptibility is used for the interpretation of experimental data concerning the spin-glass state obtained as a result of different types of the competitions of the magnetic interactions. Statistical background of the electrical properties as a result of their correlation with the magnetic properties is discussed.

Enhancing semiconductor device performance using ordered dopant arrays

As the size of semiconductor devices continues to shrink, the normally random distribution of the individual dopant atoms within the semiconductor becomes a critical factor in determining device performance--homogeneity can no longer be assumed. Here we report the fabrication of semiconductor devices in which both the number and position of the dopant atoms are precisely controlled. To achieve this, we make use of a recently developed single-ion implantation technique, which enables us to implant dopant ions one-by-one into a fine semiconductor region until the desired number is reached. Electrical measurements of the resulting transistors reveal that device-to-device fluctuations in the threshold voltage (Vth; the turn-on voltage of the device) are less for those structures with ordered dopant arrays than for those with conventional random doping. We also find that the devices with ordered dopant arrays exhibit a shift in Vth, relative to the undoped semiconductor, that is twice that for a random dopant distribution (- 0.4 V versus -0.2 V); we attribute this to the uniformity of electrostatic potential in the conducting channel region due to the ordered distribution of dopant atoms. Our results therefore serve to highlight the improvements in device performance that can be achieved through atomic-scale control of the doping process. Furthermore, ordered dopant arrays of this type may enhance the prospects for realizing silicon-based solid-state quantum computers.

High-Resolution Numerical Study of Conductance and Noise Imaging of Mesoscopic Devices

We present a numerical approach, based on an optimized recursive Green’s function procedure, for the simulation of the imaging process performed by scanning a device with a negatively biased probe while monitoring its conductance, or, as we propose in this contribution, its shot noise. We discuss a few examples, for an adiabatic quantum dot and for mesoscopic cavities, studied over a 200 × 300 points discretization mesh. The effect of disorder associated with the random distribution of dopants is included in some of the simulations by means of a semi-analytical procedure for the evaluation of the screening from the 2DEG.

‘Atomistic’ simulation of ultra-submicron MESFETs

The effects of a discrete random dopant distribution on the MESFET characteristics were investigated using device simulation. The first three moments of the Boltzmann equation coupled to the Poisson equation are solved self-consistently. The model accounts for hot-electron effects, degeneracy and surface-states effects. The simulation indicates a nonuniform potential distribution in the active layer, which creates sharp variations in the electron density in that layer. It was also found that the drain electric current depends on the actual distribution of impurities. A comparison is made between the solutions of the transport equations for both dopant models (discrete and uniform) for MESFETs.

Short-range and long-range Coulomb interactions for 3D Monte Carlo device simulation with discrete impurity distribution

A 3D Monte Carlo (MC) device simulation model is developed for the treatment of discrete random dopant distribution in sub-100 nm MOSFET. The electron–ion (e–i) interaction model is based on a suitable description of long-range Coulomb interaction included by a particle–mesh (PM) calculation method and short-range interaction taken into account by a scattering mechanism. Attention is given to the correct definition of the short-range domain and scattering model. This new approach is validated by computing the low-field electron drift mobility in Si by means of MC simulation of 3D resistors in the doping concentration range of 10 15–6.4×10 19 cm −3. A good agreement is found between calculation and experimental data at 300 K.

Dopant Pairing in a Molecular Semiconductor

AbstractRecent doping experiments in n-type perylene diimide (PPEEB) semiconducting thin films showed an unexpected quadratic dependence of electrical conductivity upon dopant molecule concentration. We propose that singly-ionized dopant pairs outnumber ionized unpaired dopants and dominate conductivity. Random association into dopant pairs during spin coating then explains the quadratic dependence. Classical calculations confirm that dopant pairing reduces the binding energy of the easiest-to-ionize electron. Our model agrees with the measured conductivity activation energy and magnitude, assuming typical electron mobility in the crystal. The random distribution of dopants implies their distribution cannot equilibrate during the spincoating process.

A 25-nm MOSFET with an Electrically Induced Source/Drain

As MOSFET devices are scaled down to gate lengths of 50 nm and beyond, the requirements for the bodydoping concentration, the gate-oxide thickness, and source/drain doping profiles to control short-channel effects become increasingly difficult to meet [1]. A low channel impurity concentration is desirable since it gives a large carrier mobility and a small threshold voltage fluctuation for a random dopant distribution. However, the threshold voltage of this kind of device depends mainly on the work function of the gate material [2]. This paper reports on a new 25-nm MOSFET structure utilizing n polysilicon floating gate spacers (FGS). In the 25-nm MOSFET with a thin gate oxide, control of the threshold voltage by increasing depletion charges has many disadvantages, like a carrier mobility degradation, a large threshold voltage fluctuation due to a random dopant distribution and a large subthreshold slope. So the work function of the gate material is adjusted to control the threshold voltage. The work function of the main gate is assumed to be 4.5 eV to obtain the designed threshold voltage value of 0.3 V with a low channel doping concentration. Also, the main gate is made of materials whose work functions are close to

Discrete random dopant distribution effects in nanometer-scale MOSFETs

In this paper, discrete random dopant distribution effects in nanometer-scale MOSFETs were studied using three-dimensional, drift-diffusion “atomistic” simulations. Effects due to the random fluctuation of the number of dopants in the MOSFET channel and the microscopic random distribution of dopant atoms in the MOSFET channel were investigated. Using a simplified model for the threshold voltage fluctuation due to dopant number fluctuation, we examine the standard deviations of the threshold voltage that can be expected for a highly integrated chip.

CMOS scaling into the 21st century: 0.1 µm and beyond

This paper describes the design, fabrication, and characterization of 0.1-µm-channel CMOS devices with dual n+/p+ polysilicon gates on 35-A gate oxide. A 2× performance gain over 2.5-V, 0.25-µm CMOS technology is achieved at a power supply voltage of 1.5 V. In addition, a 20× reduction in active power/circuit is obtained at a supply voltage of < 1 V with the same delay as the 0.25-micron CMOS. These results demonstrate the feasibility of high-performance and low-power room-temperature 0.1-µm CMOS technology. Beyond 0.1 µm, a number of fundamental device and technology issues must be examined: oxide and silicon tunneling, random dopant distribution, threshold voltage nonscaling, and interconnect delays. Several alternative device structures (in particular, low-temperature CMOS and double-gate MOSFET) for exploring the outermost limit of silicon scaling are discussed.

Monte Carlo treatment of the nonradiative energy transfer process for nonrandom placements of dopants in solids

The nonradiative energy transfer process from donor-to-acceptor ions is simulated for the garnet lattice using the Monte Carlo (MC) method. The probabilities of the events which occur after a donor is excited are calculated, i.e., the donor and acceptor emission transients. Two different simulation results are reported. One is obtained under the Forster and Dexter (FD) assumptions—dopants are randomly distributed in the crystal, no donor-to-donor and no acceptor-to-donor transfers occur, and the transfer is proportional to (1/R)s with s a unique integer. The second is obtained by replacing the random spatial distribution of dopants in the FD model by a nonrandom distribution. The nonrandom placements result from a short-range interaction between donors and acceptors which may be attractive or repulsive. For both distributions, the FD assumptions that s is a unique integer is relieved and transients are obtained for an arbitrary multipolar expression. The FD model was found to give a rather good approximation to the donor emission transient determinated by the MC simulations for the FD assumptions. The donor luminescence decay is faster for an attractive interaction between donors and acceptors than for the random distribution. It is slower for a repulsive interaction. Using the arbitrary multipolar expression and using random and nonrandom spatial distributions of dopants give distinguishably different decay transients. However, better discrimination among causes for some particular given transient is afforded by using different dopant levels.

Scattering from ionized dopants in Schottky barriers

We calculate the true electrostatic potential that forms a Schottky barrier, arising from a random distribution of dopants in the depletion region. It is customary to model the electrostatic potential in a Schottky barrier by a one-dimensional quadratic potential, as would obtain from a jellium of charge in the depletion region. The difference between the true potential and the ‘‘ideal’’ one acts as a perturbation to electrons traversing the depletion region. This perturbation is found to be small, and effectively one dimensional. Its effect on the transmission probability of an electron tunneling through the depletion region is shown to be negligible. It is shown, however, that because of the large quantum dipoles near the metal–semiconductor interface, the image-force lowering correction does not in general vary as N1/4d, as is widely thought, at least for heavily doped materials. Local-density calculations of the Schottky barrier height were made for several metal/GaAs interfaces, to obtain the magnitude and range of the interfacial dipoles. They are found to be large and persist sufficiently far from the interface that they dominate the slowly varying potential from the ionized dopants. Thus, the interplay between the image-force potential and the interfacial dipoles determines the maximum in Schottky barrier height; moreover the quantum dipoles add an additional correction of the same magnitude as the image-force correction in the heavily doped case.

Minimization of dopant-induced random potential fluctuations in sawtooth doping superlattices

Potential fluctuations due to random dopant distribution are estimated in a doping superlattice for different doping profiles. It is shown that statistical potential fluctuations are minimized by employing a doping profile consisting of a train of δ functions, which result in sawtooth-shaped band edges. Clearly resolved quantum-confined optical absorption and luminescence transitions are observed in this improved doping superlattice structure. The sawtooth superlattice provides the basis for a novel GaAs technology which is suited to operate in the minimum dispersion region of silica fibers at a wavelength of λ=1.3 μm.

Effects of Doping Induced Disorder in Polyacetylene

Abstract In this paper, the effects of disorder on the electronic structure of doped trans-polyacetylene are evaluated. In effect, at the high doping levels currently obtained on (CH)x, it is not realistic to consider the dopant molecules as isolated impurities. Therefore we have examinated in this work how a random distribution of dopant molecules broadens the impurity levels lying within the band gap. We consider a model where the molecules form covalent bonds with the neighbouring n orbitals. Making use of the SSH model for (CH)X chain, the electronic density of states is obtained by averaging the Green's function over the random distribution of dopants along the lines of a cluster CPA. From our results, we discuss the formation of impurity bands in the gap and their dependence on both the doping concentration and the chemical nature of the doping molecules.