Search

Present computer processing capabilities are becoming a restriction to meet modern technological needs. Therefore, approaches beyond the von Neumann computational architecture are imperative and the brain operation and structure are truly attractive models. Memristors are characterized by a nonlinear relationship between current history and voltage and were shown to present properties resembling those of biological synapses. Here, the use of metal-insulator-metal-based memristive devices in neural networks capable of simulating the learning and adaptation features present in mammal brains is discussed.

The recent realization of memristors, nanodevices exhibiting non-volatile resistive switching, has sparked tremendous interest for applications in fields such as nonvolatile memories. Here we report unipolar resistive switching in Pt/MgO/Ta/Ru structures, with an oxide barrier thickness of only 15 nm. No electroforming process was required to achieve resistive switching and an ohmic conduction mechanism is associated with the ON state. We observed an inverse dependence of the ON state resistance on the SET current compliance and average values of 1.61 V and 1.38 V for the SET and RESET voltages, respectively. We show the stability of the switching for over 40 cycles and a clear separation of the ON (10¹ Ω) and OFF (10² Ω) states during at least 10⁴ s.

Resistive switching in metal–insulator–metal nanosctructures is being intensively studied for nonvolatile memory applications. Here, we report unipolar resistive switching in Pt/MgO/Ta/Ru structures, with a 30 nm oxide barrier. A forming process was needed to initiate the resistive switching, which was then observed for all Set and Reset voltage polarity combinations. We studied the influence of the voltage polarity on the variability of the Set/Reset voltages and ON/OFF resistances and revealed the importance of a thin TaOx layer working as an oxygen revervoir for resistive switching. The mechanism behind this phenomenon can be understood in terms of conductive filaments formation/rupture with a contribution from Joule heating. Resistance change is thus caused by a voltage-driven oxygen vacancy motion in the MgO layer and a filament model was proposed for each polarity mode. A OFF/ON resistance ratio of at least 2 orders of magnitude was obtained with resistive states stable up to 104 s. Our results open the prospect to improve switching performance in other resistive switching systems, by proving a better understanding of the differences between operation modes.

In this work we studied the electrical behavior of Linde type A zeolite (K+) in the presence of two polymers, poly(vinyl alcohol) (PVA) and poly(vinyl pyrrolidone) (PVP), with excellent film forming properties. Homogeneous composite thin films of PVA/LTA-zeolite and PVP/LTA-zeolite were prepared with different zeolite concentrations. The current−voltage (I−V) characteristics of the composites were measured at different applied voltages. The results show that the conductivity properties are composition-ratio-dependent and are also related to the type of polymers. Moreover, a well-defined step-like change was detected in the I−V curve of PVP/LTA-zeolite at very high applied voltage. © 2013 Society of Chemical Industry

11 pages, 13 figures, 3 tables., Targeted metabolite analysis of aqueous extract of Rumex induratus leaves, in terms of phenolic compounds and organic acids, and the study of its antioxidant activity against the DPPH• radical, a reactive oxygen species, hypochlorous acid, and a reactive nitrogen species, nitric oxide (•NO), were performed. The samples were collected in several locations, spontaneously occurring or from greenhouse culture, at different stages of development and seasons. The phenolic composition was achieved by high-performance liquid chromatography (HPLC)−diode array detection, and four hydroxycinnamic acid derivatives and 10 flavonoid glycosides (C- and O-heterosides) were determined. Organic acids composition was established by HPLC-UV, revealing five compounds. The total amount of phenolic compounds and organic acids were affected by growing conditions and developmental phase. The aqueous extract exhibited a dose-related activity against all tested reactive species.

Phenolic profile of 10 different varieties of red "Vinho Verde" grapes (Azal Tinto, Borraçal, Brancelho, Doçal, Espadeiro, Padeiro de Basto, Pedral, Rabo de ovelha, Verdelho and Vinhão), from Minho (Portugal) were studied. Nine Flavonols, four phenolic acids, three flavan-3-ols, one stilben and eight anthocyanins were determined. Malvidin-3-O-glucoside was the most abundant anthocyanin while the main non-coloured compound was much more heterogeneous: catechin, epicatechin, myricetin-3-O-glucoside, quercetin-3-O-glucoside or syringetin-3-O-glucoside. Anthocyanin contents ranged from 42 to 97%. Principal component analysis (PCA) was applied to analyse the date and study the relations between the samples and their phenolic profiles. Anthocyanin profile proved to be a good marker to characterize the varieties even considering different origin and harvest. "Vinhão" grapes showed anthocyanins levels until twenty four times higher than the rest of the samples, with 97% of these compounds.

Strong increase in electrical conductivity of graphene oxide (GO) (I a parts per thousand 10(-9) A) is found by addition of Ni nanoparticles (Ni-NPs) preliminarily solved by HCl (Ni-sol) (I a parts per thousand 10(-4) A) or powder (Ni-pow) obtained from this solution (I a parts per thousand 10(-6) A), while simply mixing GO with Ni-NPs an insulator similar to pure GO is obtained. Thus, Ni-sol and Ni-pow can be used to transform GO from insulator to semiconductor. One of the transformation mechanisms is Ni as spillover. At the same time, different kinds of the magnetic response are obtained on GO and reduced GO (rGO) samples with and without Ni. Weak paramagnetic response is detected in pure GO. Stronger paramagnetic behavior is observed for GO and rGO mixed with Ni-sol or Ni-pow. Pure rGO sample shows weak ferromagnetism represented by slim but visible hysteresis with remnant magnetization M (r) of 0.05 emu/g. GO with Ni-NPs presents clear hysteresis with M (r) of 2.8 emu/g, while sample prepared by addition of Ni-NPs to rGO presents the largest hysteresis with M (r) as high as 11.8 emu/g. Thus, the optimal procedure to obtain the magnetic response requested for particular application can be chosen.

Electrical properties and magnetic properties of graphene oxide and reduced graphene oxide films with addition of Ni nanoparticles were investigated. It is found that reduction of GO by Hydrazine hydrate (HH) and subsequent mixing with Ni NPs yields a higher current (10(-5)A) compared to the samples where GO is first mixed with Ni NPs and further reduced (10(-6)A). At the same time, GO sample prepared by addition of Ni NPs to GO with subsequent reduction by HH presents the largest hysteresis with Mr as high as 11.8emu/g. Moreover, mixing of Ni NPs with HH before addition to GO or addition of Ni NPs to already reduced GO yields smaller values of the magnetization. Thus, the optimal procedure to obtain the magnetic response requested for particular application can be chosen. (C) 2015 Elsevier Ltd. All rights reserved.

The recent realization of memristors, nanodevices remarkably similar to biological synapses, opened the possibility to fabricate highly scalable artificial neural networks. While the physical implementation of such networks is still emerging, it is useful to perform simulations to determine the impact of non-ideal devices or device faults in the performance of memory networks. Here, we numerically evaluate a memristor-based Willshaw associative memory network, studying its capacity and robustness to noise as a function of defects probability and device parameter variations. Two types of defective memristors are addressed (stuck-at-0 and stuck-at-1) and Gaussian distributions are imposed to their threshold voltages, ON and OFF resistances. We conclude that the type and number of defects strongly determine how the network should be operated. The reading current threshold also plays a key role in determining the network's capacity and robustness to noise. Furthermore, there is a maximum defect percentage above which the network can no longer reliably store information. We also found that the memristor-based Willshaw network is more sensitive to resistance variance than to threshold voltage variance.

The use of matrix solid-phase dispersion (MSPD) was tested to, separately, extract phenolic compounds and organic acids from white grapes. This method was compared with a more conventional analytical method previously developed that combines solid liquid extraction (SL) to simultaneously extract phenolic compounds and organic acids followed by a solid-phase extraction (SPE) to separate the two types of compounds. Although the results were qualitatively similar for both techniques, the levels of extracted compounds were in general quite lower on using MSPD, especially for organic acids. Therefore, SL-SPE method was preferred to analyse white “ Vinho Verde ” grapes. Twenty samples of 10 different varieties (Alvarinho, Avesso, Asal-Branco, Batoca, Douradinha, Esganoso de Castelo Paiva, Loureiro, Pederna, Rabigato and Trajadura) from four different locations in Minho (Portugal) were analysed in order to study the effects of variety and origin on the profile of the above mentioned compounds. Principal component analysis (PCA) was applied separately to establish the main sources of variability present in the data sets for phenolic compounds, organic acids and for the global data. PCA of phenolic compounds accounted for the highest variability (77.9%) with two PCs, enabling characterization of the varieties of samples according to their higher content in flavonol derivatives or epicatechin. Additionally, a strong effect of sample origin was observed. Stepwise linear discriminant analysis (SLDA) was used for differentiation of grapes according to the origin and variety, resulting in a correct classification of 100 and 70%, respectively.

The electrical properties of polymer composites based on polycarbonate (PC) and panipol CXM (CX), filled with reduced graphene oxide (rGO), were investigated. The composite preparation conditions allowed good dispersion of rGO in the polymer matrix. We show here that when used as a nanofiller in polymers, rGO offers an appreciable improvement of the electrical current in 3 orders of magnitude (from 10−10 A to 10−7 A at 10 V), as observed in current-voltage (I-V) data for both PC and CX polymers with rGO. The suggested mechanism for the observed switching effects is the migration of oxygen groups aided by both the electrical field and Joule heating. Moreover, some reset- and set- like changes similar to resistive switching were observed in the I-V data of PC and CX-based films upon the addition of rGO. Clockwise (resembling a memristive system type II) and counter-clockwise (resembling a memristive system type I) directions were detected in the I-V data of the analyzed films. According to the obtained results, rGO can be a good filler for PC and CX polymer-based films for application in electronic and photonic areas, due to the significant improvement of the electrical conductivity of these polymers.