Summary of the work done in the first year

Summary of the work done in the first year

Attention: open in a new window. PDFPrintE-mail

 Summary 

The main objective of the project is the realization of a new, highly innovative technology for the production of functional molecular assemblies that can perform advanced tasks of information processing, involving learning and decision making, and that can be tailored down to the nanoscale. The underlying concept is the assumption that the paradigms of learning (in particular, Hebbian-type learning rules including spike time dependent plasticity, and other forms of learning in biological cognitive systems) can also apply to complex assemblies of electrochemically controlled, conducting polymeric  non linear elements. The success of this endeavour is expected to open up a totally new, bottom-up technology for the fabrication of information processing hardware capable of adaptation and decision making on the basis of previous supervised and unsupervised learning. More specifically, our work has been divided into four workpackages, for each one of which specific objectives are defined:

WP1   Fabrication and optimization of electrochemically controlled simple networks

·Identification of element and network composition and synthesis of all necessary compounds;

· Optimization of discrete element construction and realization of simple 2-D network based of the interconnections of these elements;

·         Formation of fibre  matrix composed by PANI-PEO-gold  nanoparticles composite on 2D supports.

WP2    Training of the network, and successful application to a biological system

  • Development of the training protocols of the networks
  • Synthetic reconstruction of the snail nervous system (CPG) exhibiting  several rhythmic activities in response to different inputs.

WP3   Fabrication of the complex statistical matrix and bio-inspired tailoring

  • Fabrication of a hybrid polymeric-gold nanoparticles matrix capable of “intelligent “ handling of information and its optimization to mimic important functional areas of the brain, such as the CC.

WP4   Discrimination and learning in  the biologically inspired supramolecular device.

 ·  Development of a new, bottom-up technology for the fabrication of sophisticated functional supramolecular structures which can be miniaturised in principle down to the nanoscale, which are capable of decision making and complex signal analysis;

·  Fabrication of a synthetic system which by mimic biological sensory and cognitive systems can be used as a new, revolutionary instrument for neuroscience. Following our Gantt table, during the first year work was performed on Wps 1, 2, 3, with a prevalence of time devoted to Wp1; this is reflected in the space  which will be devoted to our activities in this report.   

WP1

 The overall objective of WP1 was the optimization of the electrochemically controlled polymeric heterojuction of the conducting polymer polyaniline (PANI) and the solid electrolyte Li-doped Polyetheneoxyde (PEO) which is at the heart of the proposed adaptive network we aim to build. The two critical parameters here are the efficiency of the control and the stability in the functioning of the device. In both cases these are crucial issues to address in molecular electronics in general and in our molecular systems, particularly PANI. From this point of view we must specify that by stability we not only mean the obvious static stability of element and circuit characteristics, but also the dynamic reproducibility in adaptive behaviour over as long a time (or number of learning cycles) as possible.Concerning the first parameter, the most important benchmark is the conductivity ratio between the electrochemically controlled “conducting” and “insulating” phases of PANI. In our first versions of the device such ratio was about 100 (1). We set out to increase this to the target ratio of 1000. In order to do this we used differently synthetized PANI and PEO with varying chain length and polydispersity, and we explored different dopants and doping methods for both polymers to increase the  conductivity of PANI and the conductivity ratio. More specifically:PolymersPolyaniline (PANI) doped with DBSA (dodecyl benzene sulfonic acid) have been synthesized using cobalt sulphate as agent to increase the polymer chain length and regioregularity and hence the conductivity. The use of the DBSA is important, not only for filmability, but also for the mechanical stability of the film in the device.Polyethylene oxide (PEO) samples of  different  molecular weight (12,000, 35,000 and 8,000,000 Da) were used as solid electrolytes.DopantsLithium tetrafluoroborate (LiBF4, 98 %), Lithium trifluoromethanesulfonate (LiCF3SO4, 99.99 %) were investigated as ionic dopants of PEO with respect to LiClO4 to improve the device redox cycling without affecting the mobility of Li ions between the two conductive layers in the device. These different materials were then tested by assembling them to yield our device, using different fabrication protocols: In particular: two methods were used for the PANI channel formation, namely, modified Langmuir-Schaefer (LS) technique and polyelectrolyte self-assembling, known also as layer-by-layer (LbL) method. Solid-electrolyte area was formed by gel casting with successive dryingWe found that the best results were obtained with LS technique; however we still plan to use LbL especially for the layer formation on non-planar surfaces.To improve the device stability after repeated cycling we explored several assembly configurations, using different combinations of the component materials. In particular: different acids were used for the PANI channel formation. Best results were obtained when dodecyl benzene sulfonic acid (DBSA) was used as doping agent.Different salts were tested for the solid electrolyte formation. It was shown the variation of the device cyclic voltage-current characteristics. The most promising results were obtained when the electrolyte was formed from PEO doped with the mixture of lithium perchlorate and DBSA. We found that the most stable configuration is the sealed one. With this device we improved the stability in functioning after repeated cycles by at least two orders of magnitude of both the single device (node) and a simple network relative to our first devices and simple networks.  This improvement is impressive and guarantees that there will be sufficient time for the device and the networks to learn and adapt during long training protocols.Therefore the first two objectives of WP1 have been fully reached. All necessary chemical compounds were identified and synthesized. In particular, active channel must be composed from high molecular weight PANI, using DBSA as a dopant, allowing high conductivity ratio in reduced and oxidized states as well as significant stability. The solid electrolyte matrix is high molecular weight PEO. In the case of complex, especially statistical, networks, the mixture of the lithium perchlotate with DBSA must be used as the PEO dopant. The sealed configuration of the discrete element was found to be the optimal one. Finally, we fabricated different types of “statistical” matrices, i.e. a hybrid polymeric networks of PANI and PEO, and also functionalized gold nanoparticles, in which we expected to have a significant number of intersection points  which would function as nodes just like our deterministic device. In particular, most work was carried out for the fabrication of networks containing statistically distributed PEO and PANI fibers. However, we also started the work related to WP3, dealing with functionalised gold nanoparticles (FGNP): the gold surface modification, necessary for the interaction with polymer components to assemble 2D and 3D structures, was obtained by the synthesis of thiol capped Au nanoparticles and Au NPs stabilised by aniline capping. The two types of nanoparticles are characterized by ionic and hydrophobic surfaces so that they can be dispersed in different phases. The best results so far were obtained with the following matrix. The system contained one input and two output electrodes. The statistical matrix, containing PEO and PANI fibers, formed by vacuum treatment, provided multiple signal pathways between the input and both output electrodes. This matrix  has shown elementary  adaptative behavior similar to supervised learning.  Summarizing, the device composition and construction were identified and optimised, including the synthesis of all necessary compounds. Simple deterministic and statistical fibrillar structures were realized, tested and showed adaptation capabilities, allowing the study of their response to simple training procedures performed in WP2.  

 Publications related to WP1: [1] V. Erokhin and M.P. Fontana, “Electrochemically controlled polymeric device: A memristor (and more) found two years ago”, ArXiv: 0807.033 (2008).[2] A. Smerieri, T. Berzina, V. Erokhin, and M.P. Fontana, “Polymeric electrochemical element for adaptive networks: Pulse mode”, J. Appl. Phys., 104, 114513 (2008).[3] T. Berzina, A. Smerieri, M. Bernabò, A. Pucci, G. Ruggeri, V. Erokhin and M.P. Fontana, “Optimization of an organic memristor as an adaptive memory element”, J. Appl. Phys., accepted (2009).[4] T. Berzina, A. Smerieri, G. Ruggeri, M. Bernabo’, V. Erokhin, and M.P. Fontana, “Role of the solid electrolyte composition on the performance of organic memristor”, Mater. Sci. Engineer. C, submitted (2009). 

WP2 

The networks and matrices fabricated in WP1 were  tested for simple adaptive behaviour. We started with an eight element deterministic network. This network has demonstrated reproducibly for many cycles  supervised learning, i.e. it changed its output for the same input according to a training voltage procedure.We also made the first step towards a “statistical” adaptive  matrix. Here the bet is that many of the statistically formed junctions between the conducting polymer and the ionic polymer will function as the single device, thereby changing the probability (or weight) of the different signal paths.  Applying the improved materials and fabrication procedures developed in WP1, we fabricated several types of statistical matrices.  The best stability and reproducibility was obtained on circuits composed from deterministic elements. Statistical networks, formed by successive vacuum treatment of PEO and PANI solutions, have demonstrated adaptive behaviour. However, the stability of their structure and properties was not high, and more work needs be done on this, using the optimized materials and fabrication procedures developed in WP1.Whereas the work in WP1 was totally concerned with materials science aspects of the project, in WP2 we have the first concrete meeting between materials science and neuroscience. In particular, the Warwick unit carried out work on:

1.      Biophysical mechanisms underlying learning in the pond snail (Limnea Stagnalis) (Conditioned Response). In one of our recent publications [5], we have worked out, in collaboration with the experimental team at Sussex University, that persistent sodium and potassium channels are the key mechanisms for nonsynaptic learning observed in Conditioned Response. Our results should be valuable for Milestone 1 (Fabrication of matrix nodes; fabrication of fiber matrix; first application to biological system) due in the 18th month.

 2.      In collaboration with an experimental team in Babraham Institute, Cambridge, we have analysed data obtained with multi-electrode arrays in the left- and right-hemisphere in inferotemporal cortex in sheep. We observed that learning results in a significant increasing of theta power, but not gamma power. A model was then developed to explain the phenomenon and explore the functional consequence of increasing theta power. The paper is current under reviewing for Nature (Neuroscience) [6]. The finding could be useful to benchmark the adaptive behavior of the fully developed statistical matrix of WP3.      3.      Developing causality analysis tools to tackle experimental data. With the development of recording techniques in neuroscience, we often face high-throughput data such as the data we mentioned in 2 above. How to understand the data via reverse engineering approaches is a key issue in computational biology and could be useful for our project as well (to assess how a network is changed before and after learning). We have recently worked on developing such tools [7,8,9].

In conclusion the beginning work on WP2 is satisfying: in addition to the very productive period of the first year, as demonstrated in the publications below, one PhD student from University of Parma is visiting Warwick for a year  and will be a very useful bridge between the Parma and Warwick teams., particularly for the work in WP2 leading to the first milestone of the project.

 Publications related to WP2 [5] ES Nikitin, DV Vavoulis,Feng J.F.,M O’Shea, PR Benjamin & G Kemenes(2008) Persistent sodium current is a non-synaptic substrate for long-term memory Current Biology vol. 18: 1221-1226 [6] Kendrick KM, Y Zhan, H Fischer, A U Nicol, XJ Zhang, Feng J.F. (2009) Learning alters theta-nested gamma oscillations in inferotemporal cortex Nature Precedings hdl: 10101/ npre. 2009.3151.1.[7] Zou CL, Feng J.F. (2009) Granger causality vs. Dynamic Bayesian network inference: A Comparative Study BMC Bioinformatics vol. 10:122 doi:10.1186/1471-2105-10-122.[8] Feng J.F. Yi DY, Krishna R, Guo SX, Buchanan-Wollaston V.(2009) Listen to genes: dealing with microarray data in the frequency domain PLoS One 4(4): e5098. doi: 10.1371 / journal.pone. 0005098. [9] Guo SX, Wu J.F.H.,  Ding MZ, Feng J.F. (2008) Uncovering interactions in the frequency domain PLoS Comput Biol 4(5): e1000087. doi:10.1371/journal.pcbi.1000087.[10] Rossoni E., Feng J.F.,   Tirozzi B., Brown D., Leng G., and Moos F. (2008) Synchronous bursting of oxytocin neurons; emergent behaviour of a model, PLoS Comp. Biol. 4(7): e1000123. doi:10.1371/journal.pcbi.1000123 (widely reported in worldwide news)[11] Zhang XJ,You GQ, Chen TP, Feng J.F. (2009) Readout of spike waves in a microcolumn Neural Computation (accepted, IF=2.2) [12] V. Erokhin, A. Schüz, and M. P. Fontana, “Organic memristor and bio-inspired information processing”, Int. J. Unconventional Computing, submitted (2009). 

WP3 

This work package started at the 8th month, so only preliminary activity can be reported.

 

This concerns the synthesis of functionalized gold nanoparticles according to two different protocols, characterization of the materials by SEM imaging and microRaman spectroscopy, and correlations of activities to table 1 of Annex 1 of the CA. In particular:1. The Au NPs modified at the surface by aniline ligands have been used as reactive monomers in the aniline polymerization to obtain the hybrid Pani-Gold nanoparticles matrix in which the macromolecular chains are linked to the surface of nanoparticles allowing the electronic interaction through a Schottky barrier.2. Alkyl thiol capped Au nanoparticles, prepared with tailored diameter distribution and purity, have been used to obtain components for self assembling by modification with cavitand derivatives.3. The first steps of the synthesis of PEO derivatives for self assembling materials and synthesis of polymeric materials or 2-3D  network have been carried out. In particular we prepared  an end-groups functionalized PEO (Me-pyridine-PEO-pyridine-Me) and a /ethylene oxide/styrene block copolymer (PEO-b-PS).4. Gold nanoparticles (3nm diameter) functionalized on the surface with phosphonate cavitands have been prepared and characterized. These hybrid materials exhibited precise molecular recognition properties toward methylalkylammonium and methylpyridinium guests. Methylpyridinium terminated PEO was connected to the nanoparticles to form reversible networks. The formation/dissociation of these networks, driven by molecular recognition events, was proven. In particular displacement of PEO by a competitive guest (methylbutyl ammonium chloride) led to the disassembly of the polymer-nanoparticle network. Subsequent base treatment restored the free cavitands on the nanoparticles, which immediately reformed the network. The overall self-assembly cycle was demonstrated. In parallel to the synthesis, fabrication and characterization activities  in the Pisa and Parma units,  brain studies were carried out at the MPI in Tübingen. Since the aim of BION is to imitate features of the nervous system which are crucial for information processing and learning, the task of the MPI was to contribute the relevant knowledge on this topic. Basic knowledge on the nervous system, relevant already from WP1 on, was transferred during various meetings (see also [15]) and the features of the proposed molecular network were adapted accordingly. In addition, optimization of neurological data has been carried out  by quantitative studies on the internal  connectivity of the mammalian cerebral cortex [16,17]. These are highly relevant for the formation of the cortex-like matrix envisaged in WP4. Publications related to WP3:  [13]  B. Gadenne, M. Semeraro, R. M. Yebeutchou, F. Tancini, L. Pirondini, E. Dalcanale, A. Credi:
Electrochemically Controlled Formation/Dissociation of Phosphonate Cavitand-Methylpyridinium Complexes
Chem.Eur. J. 2008, 14, 8964-8971.[14]  R. M. Yebeutchou, F. Tancini, N. Demitri, S. Geremia, R. Mendichi, E. Dalcanale;
Host-Guest Driven Self-Assembly of Linear and Star Supramolecular Polymers
Angew. Chem. Int. Ed. Engl. 2008, 47, 4504-4508.[15] Schüz, A.: Neuroanatomy. Scholarpedia 3(3), 3158, Scholarpedia.org, San Diego, CA, USA (2008) [Note: Online-Resource][16] Voges, N., A. Schüz, A. Aertsen and S. Rotter: A modeler‘s view on the spatial structure of intrinsic horizontal connectivity in the neocortex. Progress in Neurobiology (submitted) [17] Schüz, A. and F. Sultan: Brain connectivity and brain size. New Encyclopedia of Neuroscience 2, 317-326. (Eds.) Squire, L. R., T. Albright, F. Bloom, F. Gage, N. Spitzer, Elsevier, Amsterdam, Netherlands ( 2009)    More information of the project, including references to our published work, news and highlights can be found at our Project internet site http://www.fp7-bion.eu