Self-Heating Aware Design of ICs in Deep Sub-Micron FDSOI and Bulk Technologies

Bulk CMOS technologies left the semiconductor market to the novel device geometries such as FDSOI and FinFET below 30 nm, mainly due to their insufficient electrical characteristics arising from different physical limitations. These innovative solutions enabled the ongoing device scaling to continue. However, the threshold voltage and the power supply values did not shrink with the device sizes, which caused an excessive amount of heat generation in very small dimensions. With the high thermal resistivity materials used in FDSOI and FinFET, the generated heat cannot leave the device easily, which is not the case in bulk. With all of these, modern geometries brought a major problem, which is the self-heating.

Due to self-heating effects (SHE), the temperature of a device rises significantly compared to its surroundings. Having very large local temperature brings important reliability issues. Moreover, the electrical behaviour of a device also changes dramatically when its temperature is very large. These facts bring the need of considering SHE and the temperature of each device separately. Nevertheless, in many of today's CAD tools, a single global temperature is applied to all of the devices. Even if some advanced simulation options are used, estimating the temperature of a device is not a simple task as it depends on many parameters.

The focus of this thesis is to show the significance of SHE in the design of ICs and provide self-heating aware design guidelines. In order to achieve this, different circuit implementations are studied by considering the SHE. The study consists of two main parts, which are the reliability of the high-speed digital circuits and the performance of analog blocks where noise is critical. Moreover, detailed device-level electro-thermal simulations are performed to explain the self-heating phenomena more in detail and to perform a comparison between bulk and FDSOI.

The digital part of the self-heating study is performed on two very high-speed full-custom 64-bit Kogge-Stone adders in 40 nm and 28 nm technologies. Thermal simulations are performed on these blocks to compare SHE in bulk and FDSOI geometries. The comparison of two implementations also provides the increasing significance of SHE with scaling. Extensive heating analyses are performed to find the most critical devices that are the primary heat generators. Design guidelines and solutions are proposed to flatten the temperature profiles in precharged and static logic implementations and to decrease the probability of electromigration.

The analog study of the work focuses on the thermal noise performance of LNAs and SHE on the flicker noise. Since thermal noise of a device linearly depends on the temperature, it is directly affected by SHE. To show the amount of SHE on the noise figure, three common gate cascode LNAs operating at 2 GHz with different device lengths are implemented in 28 nm FDSOI. The measurements show that the self-heating effects are clearly observed on the noise figure and the performance of the blocks deviate importantly from the simulations. Moreover, the self-heating effects are significantly more in short channel devices due to their large heat density. Similar experiments are also performed on different test structures in FDSOI at lower frequencies to observe SHE on flicker noise. The experiments show that flicker noise degrades at larger temperatures and more in short channel implementations.

Fast temperature cycling of the catalytic CO oxidation using microstructure reactors

Martin Luther

The possibility to increase the performance (productivity or selectivity) of a chemical reactor by using periodic variations of reaction parameters (e.g. reactant concentration or temperature) has been theoretically envisaged since the beginning of the 70th. The experimental validation of the predicted positive effects was successful in the case of concentration variation but failed for temperature variation. This was mainly due to the high thermal inertia of the conventional chemical reactors used for the measurements which prevented to create variations having a sufficiently high frequency. Microstructure reactors own, at the contrary to conventional reactors, a very low thermal inertia and allow to generate temperature oscillations with an amplitude of about ten to hundred Kelvin at a frequency in the order of magnitude of 10-1 to 4 Hz. These properties, coupled with the possibility to introduce a catalytic active material within the devices, seem to make them well suited for the study of the effects of fast periodic temperature variations of a catalytic reaction. The objective of this work was to demonstrate that non-stationary temperature conditions may increase the reaction rate of a heterogeneously catalyzed reaction up to values not predicted by the classical Arrhenius dependency towards the temperature. Two different types of microstructure reactors have been used. The reaction was taking place in the first one (FTC-type 2) in microstructured channels on whose walls a catalytic layer was deposed. In the second one (FTC-type 3), the reaction was taking place on a piece of sintered metal fibres (SMF) plate placed in a reaction chamber. The catalytic active material was deposed on the SMF plate filaments. Both devices where permanently heated by electrical resistors and periodically cooled down with water flowing through cooling channels incorporated in the devices. The test reaction chosen for the experimental measurements was the CO oxidation reaction, heterogeneously catalyzed by platinum supported on alumina (Pt/Al2O3). The reaction behaviour under stationary thermal conditions was consistent with the one predicted by the Arrhenius law. The dependency of the reaction rate with the temperature was exponential with an apparent activation energy of 104 kJ·mol-1, a negative partial reaction order for CO and a positive one for O2. The measurements effectuated under quasi-stationary thermal conditions (slow temperature ramps) have shown that a temperature change rate between 7 and 14 K·min-1 was not sufficient to observe any non-trivial effect of the temperature. The reaction behaviour was always predicted by the Arrhenius law. The surface coverage of the reactive species is, in this case, always able to follow the slow temperature changes and the reaction behaves as being always under steady-state conditions. The experiments realized under non-stationary thermal conditions using the FTC-type 2 reactor have also failed to demonstrate any non-trivial effects of the temperature oscillations. This device allows indeed to generate temperature oscillations with an amplitude of up to 120 K with a frequency of 0.1 Hz but unfortunately correlated with a very high thermal inhomogeneity. A temperature difference of up to 80 K was measured between a cold spot and a hot spot inside the device. Due to their very low local reaction rate the colder areas of the reactor have attenuated the product concentration (CO2) oscillations which should have been created by the temperature variations. This attenuation prevented the temperature oscillations to have any positive effect compared to the stationary thermal conditions. The FTC-type 3 reactor allowed temperature changes of lower amplitude but the thermal homogeneity was much better. The maximal temperature difference measured between two points within the reactor was only 15 K. Under temperature oscillation conditions, the measured instantaneous CO2 concentration was higher for any temperature within the oscillation range compared to the one recorded under stationary thermal conditions. The increase obtained in the mean CO2 concentration was ranging from 34% for a frequency of the oscillations of 0.035 Hz to 85% for a frequency of 0.052 Hz. The amplitude of the oscillations was kept constant at approximately 40 K and the mean temperature value at 437 K. The simulations effectuated using a theoretical model for the catalytic CO oxidation including a feed-back step in the form of the oxidation-reduction of the catalyst have shown that the experimentally observed increase may be qualitatively explained. Above a certain frequency, the temperature oscillations are fast enough compared to the characteristic time of the feedback step and are able to perturb the stationary established reactive species surface coverage. The formation of a transient surface coverage more favourable for the reaction than the surface coverage at high temperature allows during the transient period to reach an instantaneous surface reaction rate values higher than the one at high temperature. This reaction rate peak is then responsible for the increase of the mean reaction rate under non-stationary thermal conditions and, thus, for an increased yield.

EPFL2008

Development of a Millikelvin Scanning Tunneling Microscope for Applications in Ultra High Vacuum and High Magnetic Fields

Maximilian Assig

In this thesis we discuss the realization of a scanning tunneling microscope (STM) operating at temperatures below 20 mK. This is accomplished by attaching the STM to a dilution refrigerator. We can apply high magnetic fields up to 14 T perpendicular and 0.5 T parallel to the sample surface. A full ultra high vacuum (UHV) preparation chamber is attached to the cryostat allowing in situ preparation of the samples to investigate. This setup allows us to investigate low temperature phase transitions in solids and offers the possibility for ultimate energy resolution tunneling spectroscopy measurements on the atomic scale. The concept of realizing the mK-STM is given. A main point of concern was the design of the STM unit. The choice of materials should be well adapted for the use at very low temperatures and in high magnetic fields. A special concern are thermal expansion coefficients and thermal conductivity. In addition, we analyzed the variation of the tip sample distance for a given external excitation for our STM design and two variations of the Besoke-STM by using a one dimensional mass spring model. This gives us the possibility to extract an optimization criterion for our design to limit the response of the tip sample distance to vibrations that are transmitted to the STM. The external damping mechanism, which ensures a low mutual vibration level of the STM as well as the grounding concept for the electronics is explained. To characterize the STM at operating temperatures of 1.5 K, 800 mK and 15 mK we measured topography images of Au(111) and Cu(111) surfaces. This allows us to extract the stability between tip and sample. A statistical analysis of the noise in the tunneling current without activated feedback loop yields a stability of 2.7±0.1 pm in a frequency range f ≈ 0.03 – 25 Hz at 800 mK. To verify if tip and sample are cooled efficiently, we measured the temperature at their positions with additional thermometers. We achieved 17 mK at the tip and 20 mK at the sample position, respectively. The energy resolution of the STM setup was verified by taking tunneling spectra between a superconductor and a normal conducting metal. These spectra could be fitted using the BCS theory of superconductivity. From those ts we can extract the upper limit of the effective temperature of the electrons which is Teff = 87 ± 3 mK for our best measurement. This corresponds to an energy resolution of 3.5 kbTeff = 26 ± 1 µeV. Further, the tunneling conductance between two superconductors was investigated. A sub-gap structure, which can be related to Andreev reflections in asymmetric tunneling junction was observed. In addition, at zero bias voltage a peak due to the Josephson effect can be measured. Side peaks with intensities that scale with this zero bias conductance peak appear most probably due to the interaction of Cooper pairs with the electromagnetical environment in which the junction is placed. Investigating the dependence of the conductance between a superconducting tip and a normal conducting sample on externally applied magnetic fields leads to the appearance of shoulders in the coherence peaks. These shoulders shift linearly with the external magnetic field. Such an effect has been measured for planar tunnel junctions and can be explained with the lifting of the spin degeneracy of the quasi-particle density of states in a superconductor. It was observed for the first time by Meservey, Tedrow and Fulde. The fact, that we are now able to prepare STM tips showing the Meservey-Tedrow-Fulde effect enables us to measure the absolute spin polarization at EF with atomic resolution.

EPFL2011

Submonolayer growth of cobalt on metallic and insulating surfaces studied by scanning tunneling microscopy and kinetic Monte-Carlo simulations

Philipp Buluschek

Submonolayer epitaxial growth is obtained by the deposition of less than a complete layer of atoms on a single crystal surface. It is of fundamental interest for industrial applications (e.g. in the semiconductor industry) as well as from the point of view of basic research. For example, it is known that nanometer-sized atomic structures (nanostructures) exhibit remarkable physical and chemical properties which can differ greatly from those of bulk matter. In order to investigate these properties, it is often necessary to create large quantities of well defined and possibly spatially ordered nanostructures. One way to achieve this result is self-organized growth where one deposits atoms on a clean crystalline surface and lets the growth process evolve freely. Here, nanostructures result from the atomic diffusion and aggregation processes taking place at the surface. Understanding the exact nature of these processes is of ongoing interest in the field of nanostructure growth. In this thesis we report about submonolayer growth experiments of the ferromagnetic transition metal cobalt. Cobalt was chosen because it exhibits remarkable magnetic properties. These experiments were performed in an ultra-high vacuum chamber and molecular beam epitaxy was used to grow the nanostructures. Observations of the surface were made using a variable-temperature scanning tunneling microscope (STM). In order to get a better understanding of the atomic processes happening at the surface, we developed two adapted computational simulation methods. Kinetic Monte-Carlo (KMC) simulations were used to get an atomistic picture of the surface while mean field rate equations were integrated numerically to yield cluster densities. We study the growth of cobalt on three different surfaces. By depositing Co on a clean Pt(111) crystal surface, we observe that the platinum surface reconstructs by forming star-shaped partial dislocations for sample temperatures above 180 K (-93°C). We also observe that island densities deviate from predictions of all known models towards higher values for these same temperatures. By simulation we are able to show that insertion of Co into the topmost platinum layer creates a repulsive network of dislocations. We show that these dislocations act as diffusion inhibiting barriers and thus influence the island density by constraining the free movement of atoms at the surface. We also show that the Co dimer and trimer diffuse on Pt(111) before dissociating and are able to extract corresponding activation barriers. We also study the deposition of Co on a Ru(0001) surface which shows a more conventional temperature dependence. By comparing simulation and experiment we extract all relevant diffusion activation barriers and show that at high temperatures the Co dimer and trimer dissociate. Finally, we investigate the growth of Co on the hexagonal boron nitride superlattice which forms on a Rh(111) surface. On this surface Co clusters form three-dimensional hemispherical dots. We show how the substrate corrugation influences diffusion of Co atoms and that clusters up to the size of the pentamer can diffuse. By simulation, we also extract the relevant migrations and desorption barriers.

EPFL2008

Submonolayer growth of cobalt on metallic and insulating surfaces studied by scanning tunneling microscopy and kinetic Monte-Carlo simulations

Philipp Buluschek

EPFL2008

Development of a Millikelvin Scanning Tunneling Microscope for Applications in Ultra High Vacuum and High Magnetic Fields

Maximilian Assig

EPFL2011

Fast temperature cycling of the catalytic CO oxidation using microstructure reactors

Martin Luther

EPFL2008