Search

The article considers U.S. intelligence service agencies in China during that country's civil war from 1945 to 1949. The surveillance by those agencies of activities of services of the Koumintang government of China, the Communist People's Liberation Army (PLA) and of intelligence services of the Soviet Union is examined. The use by the intelligence services of all three countries of Japanese spies and former officials of Japan's military with whom all had been at war in 1945 is discussed.

Summary: Mollaret''s meningitis is a rare form of benign recurrent aseptic meningitis first described in 1944. We report a case of Mollaret''s meningitis due to Herpes Simplex Virus type 2 (HSV2), diagnosed with Polymerase Chain Reaction (PCR) implementation in the Cerebrospinal fluid (CSF) of the patient and treated successfully with acyclovir. To our knowledge, this is the first case of Mollaret''s meningitis reported in Greece. We reviewed the literature since PCR has become widely available. Herpes Simplex Virus type 2 has been the most commonly identified causative agent of Mollaret''s meningitis. [Copyright &y& Elsevier]

Abstract: Structural and electrical properties of ALD-grown 5 and 7nm-thick Al2O3 layers before and after implantation of Ge ions (1keV, 0.5–1×1016 cm−2) and thermal annealing at temperatures in the 700–1050°C range are reported. Transmission Electron Microscopy reveals the development of a 1nm-thick SiO2-rich layer at the Al2O3/Si substrate interface as well as the formation of Ge nanocrystals with a mean diameter of ∼5nm inside the implanted Al2O3 layers after annealing at 800°C for 20min. Electrical measurements performed on metal–insulator–semiconductor capacitors using Ge-implanted and annealed Al2O3 layers reveal charge storage at low-electric fields mainly due to location of the Ge nanocrystals at a tunnelling distance from the substrate and their spatial dispersion inside the Al2O3 layers. [Copyright &y& Elsevier]

We discuss the distribution of size and aerial density of Ge nanocrystals in a metal-oxide-semiconductor (MOS) memory structure fabricated by molecular beam epitaxy combined with rapid thermal processing; the size and aerial density of Ge nanocrystals are controlled by varying the thickness of the deposited Ge layer and the processing time. Variation of tunnel oxide thickness is demonstrated with the extension of the processing time. The effect of processing time and tunnel oxide thickness on the electrical properties of the MOS structures is investigated by high frequency capacitance-voltage measurements. [ABSTRACT FROM AUTHOR]

Abstract: The fabrication at room-temperature of thin (<8nm) silicon oxide films has been achieved, in a high-density helicon plasma source reactor using Ar/O2 mixture, exhibiting relatively low concentration of fixed oxide and interface charges, after annealing. The developed plasma oxidation process was employed in order to fabricate SOI-MOSFETs at low thermal budget with competitive operating characteristics. [Copyright &y& Elsevier]

In this work we examine the high temperature charge retention characteristics of modified SONOS (polySi–oxide–nitride–oxide–Si) capacitors. The modified SONOS structures were synthesized by Si low-energy implantation (1 keV) into oxide–nitride (ON) stacks, followed by a wet oxidation step. The retention studies were conducted within the temperature range of 50 to 180 °C after setting the capacitor to an excess electron state of ΔV FB of 3 V. The reference ONO capacitor exhibits a complex behavior which consists of an initial fast discharge rate of 0.1 V/decade which switches to a lower and temperature independent rate of 0.06 V/decade when the ΔV FB reaches 2.86 V. In turn, the implanted and wet oxidized ONO shows a temperature dependent charge decay rate which varies from 0.07 V/decade at 30 °C, to 0.18 V/decade at 180 °C. The retention data have been analytically solved using the Wang–White model. The results showed that the nitride trap distribution of the implanted and wet oxidized sample differs clearly from that of the standard stoichiometric nitride. This is attributed to ion induced structural damage in combination with the wet oxidation which increases the distribution profile at deeper trap energies. [ABSTRACT FROM AUTHOR]

• The role of a 3 nm Al 2 O 3 on top of stoichiometric LPCVD SiN x MIS RRAM cells is investigated. • The devices can be SET at different resistance levels by varying the I CC. • SET/RESET voltages are higher compared to the reference sample, corresponding to the voltage drop on the Al 2 O 3. • A compact model can be successfully used to simulate the I-V curves of our devices in circuit level designs. • Impedance spectroscopy measurements revealed that the conduction in SiN x is mostly governed by trap-to-trap tunneling mechanisms. The role of a 3 nm Al 2 O 3 layer on top of stoichiometric LPCVD SiN x MIS RRAM cells is investigated by using various electrical characterization techniques. The conductive filament formation is explained, and a compact model is used to fit the current–voltage curves and find its evolution during each operation cycle. The conduction in SiN x is also studied. [ABSTRACT FROM AUTHOR]

In this work we examine the structural and electrical properties including the memory performance of Al 2 O 3 /Si 3 N 4 /SiO 2 dielectric stacks implanted with low-energy nitrogen ions and subsequently thermal annealed at 850 or 1050 °C for 15 min. X-ray photoelectron spectroscopy reveals that the concentration and the chemical state of the nitrogen atoms within the Al 2 O 3 layer depends on the post-implantation annealing (PIA) temperature. Memory testing, performed on platinum gate capacitors, shows that charge retention of the programmed states is significantly improved for the high-temperature PIA samples as compared to the non-implanted samples. While such an improvement is not detected for the low-temperature PIA samples, the latter exhibit enhanced hole charging and thus, increased erase efficiency. Overall, our results suggest that the transport properties which control the erase and the retention characteristics of the blocking Al 2 O 3 layer can be tailored by nitrogen implantation and the PIA conditions and can be used for memory performance optimization. [ABSTRACT FROM AUTHOR]

Abstract: The dependence of the various possible physical mechanisms, governing electron flow in planar metal- molecular layer-metal junctions, upon voltage bias and temperature is investigated in this work. A molecular self-assembled monolayer containing inorganic tungsten-polyoxometalate molecules is studied. These molecules are interesting for molecular memory applications due to their redox properties. Low temperature (T <150K) conduction is attributed to indirect tunneling; the effective tunneling barrier decreases with the rise of temperature due to the variation of available electronic states in the metal electrodes. Upon further increase of temperature, hopping arises as conduction mechanism and actually dominates for temperatures higher than 200K. The hopping activation energies found are several tenths of meV. The transition between conduction mechanisms depends also on the electrode distance. [Copyright &y& Elsevier]

Abstract: We report on the development of hybrid organic/inorganic thin-film transistors using regioregular poly-3-hexylthiophene (P3HT) semiconductor material deposited by means of the solid-phase Laser Induced Forward Transfer (LIFT) technique. P3HT pixels were LIFT-printed onto Au/Ti source and drain electrodes formed on silicon dioxide/p+-type Si substrate. Deposition of the P3HT pixels was investigated as a function of the laser fluence using donor substrates with and without a dynamic release layer. Device electrical characterization reveals efficient field-effect action of the bottom gate on the organic channel. The transfer I DS–V GS characteristics exhibit well-defined sub-threshold, linear and saturation regimes designating LIFT as a promising technique for hybrid organic/inorganic transistor technology. [Copyright &y& Elsevier]

• Stoichiometric SiNx layers are doped with Si atoms. • Ultra-low energy ion implantation and annealing at different temperatures was used. • Detailed material and memory cells characterization is performed. • DC current–voltage and impedance spectroscopy measurements carried out. • The role of doping in dielectric properties of the nitride films are enlighten. • Room temperature retention of resistive states are also presented. Stoichiometric SiN x layers (x = [N]/[Si] = 1.33) are doped with Si atoms by ultra-low energy ion implantation (ULE-II) and subsequently annealed at different temperatures in inert ambient conditions. Detailed material and memory cells characterization is performed to investigate the effect of Si dopants on the switching properties and performance of the fabricated resistive memory cells. In this context extensive dc current–voltage and impedance spectroscopy measurements are carried out systematically and the role of doping in dielectric properties of the nitride films is enlightened. The dc and ac conduction mechanisms are investigated in a comprehensive way. Room temperature retention characteristics of resistive states are also presented. [ABSTRACT FROM AUTHOR]

• A crossbar array of silicon nitride resistive memories was fabricated on SOI. • Multi-Level-Cell capabilities were demonstrated through DC measurements and model. • A comprehensive analysis of the MRL gates' design and optimizations were done. In this work, the fabrication of crossbar arrays of silicon nitride resistive memories on silicon-on-insulator substrate and their utilization to realize multi-rationed logic circuits are presented. Typical electrical characterization of the memristors revealed their ability of multi-state operation by the presence of 12 well separated resistance levels. Through a dedicated modeling and fitting procedure of these levels, a reconfigurable logic based on memristor rationed logic scheme is designed and a crossbar integration methodology was proposed. Finally, several circuitry aspects were simulated in SPICE with a silicon nitride SOI crossbar array calibrated model and power optimization prospects were discussed. [ABSTRACT FROM AUTHOR]

Reference-guided DNA sequencing and alignment is an important process in computational molecular biology. The amount of DNA data grows very fast, and many new genomes are waiting to be sequenced while millions of private genomes need to be re-sequenced. Each human genome has 3.2B base pairs, and each one could be stored with 2 bits of information, so one human genome would take 6.4B bits or ∼760MB of storage (National Institute of General Medical Sciences, n.d.). Today's most powerful tensor processing units cannot handle the volume of DNA data necessitating a major leap in computing power. It is, therefore, important to investigate the usefulness of quantum computers in genomic data analysis, especially in DNA sequence alignment. Quantum computers are expected to be involved in DNA sequencing, initially as parts of classical systems, acting as quantum accelerators. The number of available qubits is increasing annually, and future quantum computers could conduct DNA sequencing, taking the place of classical computing systems. We present a novel quantum algorithm for reference-guided DNA sequence alignment modeled with gate-based quantum computing. The algorithm is scalable, can be integrated into existing classical DNA sequencing systems and is intentionally structured to limit computational errors. The quantum algorithm has been tested using the quantum processing units and simulators provided by IBM Quantum, and its correctness has been confirmed. [Display omitted] • Quantum computers are expected to be involved in DNA sequencing acting as quantum accelerators. • Novel quantum algorithm for reference-guided DNA sequence alignment modeled with gate-based quantum computing. • The quantum algorithm has been tested using the quantum processing units and simulators provided by IBM Quantum. • The quantum algorithm can align sequences very fast and be applied in any assay performed with next generation sequencing. [ABSTRACT FROM AUTHOR]

De novo DNA sequence assembly is based on finding paths in overlap graphs, which is a NP-hard problem. We developed a quantum algorithm for de novo assembly based on quantum walks in graphs. The overlap graph is partitioned repeatedly to smaller graphs that form a hierarchical structure. We use quantum walks to find paths in low rank graphs and a quantum algorithm that finds Hamiltonian paths in high hierarchical rank. We tested the partitioning quantum algorithm, as well as the quantum algorithm that finds Hamiltonian paths in high hierarchical rank and confirmed its correct operation using Qiskit. We developed a custom simulation for quantum walks to search for paths in low rank graphs. The approach described in this paper may serve as a basis for the development of efficient quantum algorithms that solve the de novo DNA assembly problem. [ABSTRACT FROM AUTHOR]

Resistive Random Access Memory (RRAM) with a structure Au/Ti/TiO 2− x /Au demonstrated a clear bipolar resistive switching behavior without the necessity of an initial electroforming process. The titanium oxide (TiO 2− x ) thin film was deposited by reactive RF magnetron sputtering at room temperature in a controlled oxygen/argon ambient. The high density of oxygen vacancies within the film (induced by the low oxygen content) is an essential component for the formation of conducting filaments and demonstration of DC or nanosecond pulsed resistance switching, but also impose limitations for the conduction behavior of the high resistance state. Conductive Atomic Force Microscopy (C-AFM) was then employed in order to investigate the nanoscale electrical properties of our device. In situ current distribution during the SET process disclosed possible formation of conducting filaments while DC sweeping bias voltage revealed an OFF/ON switching ratio of about 200. We have also demonstrated that by using C-AFM both a low resistance state and a high resistance state can be written by bipolar voltage application imaged by corresponding patterns on the TiO 2− x current image, suggesting that oxygen ions movement at the Pt-Ir coated tip/TiO 2− x interface plays a critical role in the resistive switching phenomenon and thus correlating the macroscopic characteristics of our device with its microscopic origins. Nanoscale resistance switching is also demonstrated by programming distinct patterns on the device's current image. [ABSTRACT FROM AUTHOR]

• Silicon nitride was used as resistive switching in a MIS memristor structure on heavily doped SOI substrates. • The active electrode was Cu allowing for Top electrode. • AC impedance and current-voltage sweep measurements revealed higher series on SOI compared to bulk Si substrate. • Low-frequency noise measurements at low resistance state revealed the silicon nitride traps as main noise source. • Devices on SOI exhibited self-compliance characteristics as revealed by I - V sweeps and AC impedance measurements. Several resistive memory technologies (RRAMs) are prominent, but few are fulfilling the requirements for CMOS integration and meet the commercialization standards. In this work, the fabrication and electrical characterization of a fully compatible CMOS process on SOI substrate of 1R silicon SiN-based resistance switching (RS) MIS devices is presented. The RS characteristics are compared with the same devices previously fabricated on bulk silicon. [ABSTRACT FROM AUTHOR]

Abstract: An attractive alternative for extending the scaling of Flash-type memories is to replace the conventional floating gate (a poly-Si layer) by laterally isolated floating nodes in the form of nanoparticles. This floating-gate concept has led to the emergence of the so-called nanocrystal (NC) memories which have the potential of operating at lower voltages and higher speeds compared to the conventional non-volatile memories (NVMs) without compromising the criterion of non-volatility. NC memories also offer other advantages like a better immunity to the crosstalk effect arising from the floating-gate coupling of closely spaced adjacent cells and an increased design flexibility from which quantum confinement phenomena can be judiciously exploited for enhanced memory functionality. Among the different technological routes explored in the last few years for generating nanocrystals in the gate insulator of MOS devices, two major techniques have been utilized, namely deposition in vacuum and ion beam synthesis. During the last decade, we have extensively explored the ultra-low-energy ion-beam-synthesis (ULE-IBS) technique to produce single planes of Si-NCs in very thin (5–10nm) insulator layers. This review summarizes more than 10 years of research efforts we and other groups have dedicated to the fabrication of NCs memories using this original technique. By exploiting the flexibility of the ULE-IBS route, Si-NCs single-transistor NVM cells using “classical” gate oxides (SiO2) have been successfully realized. More recently, advanced dielectric stacks employing high-κ dielectric (HfO2) as tunnel oxide and SiN as control oxide and host matrix for Si and Ge-NCs, have been fabricated and show promising performance for low-voltage operating-NVM devices. Particular emphasis is placed herein on material science issues such as anomalous and controlled oxidation processes. While much research is still needed for making reliable and competitive NC NVMs, our systematic investigations based on the ULE-IBS option demonstrate that a tight control and understanding of the properties of NCs memory cells can be achieved, provided that deep structural characterization is coupled to electrical studies. [Copyright &y& Elsevier]

Abstract: The present paper focuses on the effect of the implantation energy in the fabrication of pockets of silicon nanocrystals (Si-NCs) by stencil-masked ultra-low-energy ion-beam-synthesis (ULE-IBS). Si ion implantation was carried out into 10nm SiO2 layers using energies ranging from 1 up to 3keV and a fluence of 1×1016 Si+/cm2. After mask removal the samples are furnace annealed at 1050°C for 30min under N2 atmosphere. Control of Si-NCs characteristics was examined as a function of the stencil aperture size. Then, different patterns of Si-NCs pockets were synthesized such as squares and gratings of line arrays. Photoluminescence (PL) spectroscopy under a confocal microscope was employed to map the NCs pocket arrays. These PL images were found to perfectly mimic the mask geometry. A change in PL intensity and a blueshift of the PL energy peak were observed near the edge of the pockets. These local changes were attributed to both a smaller size and a higher density of Si-NCs in such areas, probably due to a local decrease of the implanted fluence. AFM measurements of the oxide swelling as a function of the elaboration conditions were combined to PL results to evaluate the real implanted fluence, which is lower than the nominal one. The implanted fluence was also found to decrease with decreasing aperture size, until a threshold for the absence of Si-NCs formation is reached in sub-micron patterns. [Copyright &y& Elsevier]

Abstract: In this work we examine the influence of thermal oxidation on the electrical characteristics of ultra-thin strained silicon layers grown on relaxed Si0.78Ge0.22 substrates under moderate to high thermal budget conditions in N2O ambient at 800°C. The results reveal the presence of a large density of interfacial traps which depends on the oxidation process. As long as the strained silicon layer remains between the growing oxide and the underlying Si0.78Ge0.22 layer, the density of interface traps increases with increasing oxidation time. When the oxidation process consumes the s-Si layer the interface state density undergoes a significant reduction of the order of 40%. This experimental evidence signifies that the strained silicon–Si0.78Ge0.22 interface is a major source of the measured interfacial defects. This situation can be detected only when the front SiO2-strained silicon interface and the rear strained silicon–Si0.78Ge0.22 interface are in close proximity, i.e. within a distance of 5nm or less. Finally, the influence of the material quality deterioration—as a result of the thermal treatment—to the interfacial properties of the structure is discussed. [Copyright &y& Elsevier]

We report on proton radiation effects on Si-nanocrystal (Si-NCs) MOS capacitors and nMOS transistors aiming at non-volatile memory applications. Irradiation experiments were conducted on NC MOS capacitors using protons of 1.5 MeV and 6.5 MeV and on NC nMOS transistors using protons of 1.5 MeV. The range of doses investigated was ~1 to ~100 Mrad(SiO2). A 2-D layer of Si NCs with ~ 3 nm mean diameter and 1012 cm-2 surface density was successfully achieved by low-energy (1 keV) ion-beam-synthesis in thin SiO2 layers. After irradiation, programmed capacitors are found to undergo bit flip while programmed transistors are not. Charge retention measurements at room temperature for the write and erase states of irradiated and non-irradiated samples reveal that a significant memory window exists at an extrapolated time of ten years even after an irradiation dose as high as 120 Mrad (SiO2). The flat-band decay rate of the erase state in NC MOS capacitors does not depend on the irradiation-dose while the opposite occurs for the write state which is found to be directly dependent on Dit values after irradiation. These results clearly indicate that NC Non-Volatile Memories (NVM) are promising radiation tolerant devices. [ABSTRACT FROM AUTHOR]

Abstract: We report on proton radiation tolerance of Si-nanocrystal (Si-NCs) MOS structures aiming at non-volatile memory applications. Si-NCs were formed by low-energy (1keV) ion-beam-synthesis within a 9nm thick SiO2 layer. A 2-D layer of Si-NCs with 3nm mean diameter and 1012 cm−2 surface density was successfully achieved. After fabrication of Al capacitors, samples with and without Si-NCs were 1.5 and 6.5MeV proton were irradiated at doses ranging from 1Mrad (SiO2) to 120Mrad (SiO2). Significant irradiation-dose-dependent shifts are detected in the C–V curves of the NC-MOS cells and programmed cells are found to undergo bit flip. Despite the above, the attainable memory windows after write/erase operations remain unchanged. Retention time characteristics at room temperature for the write and erase states of irradiated and non-irradiated samples reveal that even after an irradiation dose as high as 120Mrad (SiO2) the devices still exhibit long time charge storage behavior. We observe that the erase state flat-band voltage decay rate does not depend on the irradiation-dose while the opposite happens for the write state flat-band voltage decay rate which is found to be directly dependent on D it values giving insight to the physics of the discharging mechanisms. [Copyright &y& Elsevier]

An array of cross-point memristive devices has been implemented on the basis of yttria-stabilized zirconia thin film for applications in prototypes of spiking neural networks. The resistive switching phenomena and the plasticity nature of such memristive devices are studied. Reproducible bipolar resistive switching and precise tuning of resistive state are demonstrated and used to implement the plasticity rule according to STDP (spike-timing-dependent plasticity) mechanism. STDP learning is found to be dependent on the memristor's initial resistive state value and discussed in terms of the finite conductance change in studied structures. Obtained results provide the foundation for autonomous neuromorphic circuits with unsupervised learning development. Unlabelled Image • An array of YSZ based cross-point memristors has been implemented for neuromorphic applications. • Resistive switching phenomenon and plasticity nature of YSZ based memristors are studied. • Spike-timing-dependent plasticity is shown to occur in the YSZ based memristors. [ABSTRACT FROM AUTHOR]

Context: Only 55% of the athletes return to competitive sports after an anterior cruciate ligament (ACL) injury. Athletes younger than 25 years who return to sports have a second injury rate of 23%. There may be a mismatch between rehabilitation contents and the demands an athlete faces after returning to sports. Current return-to-sports (RTS) tests utilize closed and predictable motor skills; however, demands on the field are different. Neurocognitive functions are essential to manage dynamic sport situations and may fluctuate after peripheral injuries. Most RTS and rehabilitation paradigms appear to lack this aspect, which might be linked to increased risk of second injury., Objective: This systematic and scoping review aims to map existing evidence about neurocognitive and neurophysiological functions in athletes, which could be linked to ACL injury in an integrated fashion and bring an extensive perspective to assessment and rehabilitation approaches., Data Sources: PubMed and Cochrane databases were searched to identify relevant studies published between 2005 and 2020 using the keywords ACL , brain , cortical , neuroplasticity , cognitive , cognition , neurocognition , and athletes ., Study Selection: Studies investigating either neurocognitive or neurophysiological functions in athletes and linking these to ACL injury regardless of their design and technique were included., Study Design: Systematic review., Level of Evidence: Level 3., Data Extraction: The demographic, temporal, neurological, and behavioral data revealing possible injury-related aspects were extracted and summarized., Results: A total of 16 studies were included in this review. Deficits in different neurocognitive domains and changes in neurophysiological functions could be a predisposing risk factor for, or a consequence caused by, ACL injuries., Conclusion: Clinicians should view ACL injuries not only as a musculoskeletal but also as a neural lesion with neurocognitive and neurophysiological aspects. Rehabilitation and RTS paradigms should consider these changes for assessment and interventions after injury.

State-of-the-art IoT technologies request novel design solutions in edge computing, resulting in even more portable and energy-efficient hardware for in-the-field processing tasks. Vision sensors, processors, and hardware accelerators are among the most demanding IoT applications. Resistance switching (RS) two-terminal devices are suitable for resistive RAMs (RRAM), a promising technology to realize storage class memories. Furthermore, due to their memristive nature, RRAMs are appropriate candidates for in-memory computing architectures. Recently, we demonstrated a CMOS compatible silicon nitride (SiN x ) MIS RS device with memristive properties. In this paper, a report on a new photodiode-based vision sensor architecture with in-memory computing capability, relying on memristive device, is disclosed. In this context, the resistance switching dynamics of our memristive device were measured and a data-fitted behavioral model was extracted. SPICE simulations were made highlighting the in-memory computing capabilities of the proposed photodiode-one memristor pixel vision sensor. Finally, an integration and manufacturing perspective was discussed.

We investigate the effects of Line Edge Roughness (LER) of electrode lines on the uniformity of Resistive Random Access Memory (ReRAM) device areas in cross-point architectures. To this end, a modeling approach is implemented based on the generation of 2D cross-point patterns with predefined and controlled LER and pattern parameters. The aim is to evaluate the significance of LER in the variability of device areas and their performances and to pinpoint the most critical parameters and conditions. It is found that conventional LER parameters may induce >10% area variability depending on pattern dimensions and cross edge/line correlations. Increased edge correlations in lines such as those that appeared in Double Patterning and Directed Self-assembly Lithography techniques lead to reduced area variability. Finally, a theoretical formula is derived to explain the numerical dependencies of the modeling method.

By modification of the electrode-solid-electrolyte interface with graphene, transit from valence change memories (VCM) to electrochemical metallization memories (ECM) in the cell Ta(C)/Ta2 O5 /Pt is demonstrated, thus, bridging both mechanisms. The ECM operation is discussed in the light of Ta-cation mobility in TaOx . The crucial role of electrochemical processes and moisture in the resistive switching process is also highlighted., (© 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.)

In this work we report on the influence of nitrogen ambient thermal effects on the performance of Pt/Al2O3/Si3N4/SiO2/Si memory capacitors. Two post deposition annealing (PDA) furnace steps were employed, at 850 and 1050 °C both for 15 min. The alumina films were deposited by atomic layer deposition using TMA/H2O at 250 °C. The structural characteristics of the stacks were evaluated by transmission electron microscopy and x-ray reflectivity measurements. The memory performance of the stacks was evaluated by write/erase and erase/write measurements, endurance and retention testing. It was found that in as-deposited state the Al2O3 layer is defective resulting in strong leakage currents, controlled by deep defects states. Thus, this behavior inhibits the memory functionality of the stacks. PDA crystallizes and condenses the Al2O3 transforming the layer from amorphous to polycrystalline. During this transformation the Al2O3 electrical quality improves greatly indicating that a significant number of these deep defects have been removed during annealing. Physical reasoning implies that the most plausible origin of these deep defects is hydrogen. However, the polycrystalline Al2O3 films showed inferior retention characteristics which are attributed to grain boundary related shallow defects. The findings of this work could pave the way for more efficient annealing schemes, in which an important factor is the time interval for hydrogen out-diffusion from the Al2O3 layer.

The clinical use of anthracyclines in cancer therapy is limited by dose-dependent cardiotoxicity that involves cardiomyocyte injury and death. We have tested the hypothesis that anthracyclines affect protein degradation pathways in adult cardiomyocytes. To this aim, we assessed the effects of doxorubicin (Doxo) on apoptosis, autophagy and the proteasome/ubiquitin system in long-term cultured adult rat cardiomyocytes. Accumulation of poly-ubiquitinated proteins, increase of cathepsin-D-positive lysosomes and myofibrillar degradation were observed in Doxo-treated cardiomyocytes. Chymotrypsin-like activity of the proteasome was initially increased and then inhibited by Doxo over a time-course of 48 h. Proteasome 20S proteins were down-regulated by higher doses of Doxo. The expression of MURF-1, an ubiquitin-ligase specifically targeting myofibrillar proteins, was suppressed by Doxo at all concentrations measured. Microtubule-associated protein 1 light chain 3B (LC3)-positive punctae and both LC3-I and -II proteins were induced by Doxo in a dose-dependent manner, as confirmed by using lentiviral expression of green fluorescence protein bound to LC3 and live imaging. The lysosomotropic drug chloroquine led to autophagosome accumulation, which increased with concomitant Doxo treatment indicating enhanced autophagic flux. We conclude that Doxo causes a downregulation of the protein degradation machinery of cardiomyocytes with a resulting accumulation of poly-ubiquitinated proteins and autophagosomes. Although autophagy is initially stimulated as a compensatory response to cytotoxic stress, it is followed by apoptosis and necrosis at higher doses and longer exposure times. This mechanism might contribute to the late cardiotoxicity of anthracyclines by accelerated aging of the postmitotic adult cardiomyocytes and to the susceptibility of the aging heart to anthracycline cancer therapy.