Computational Design of Bio-Hybrid Wood Materials

Computational Design of Bio-Hybrid Wood Materials (Dr. Mark Schubert)

Nature has provided and will provide society and academics with a vast collection of evolutionary optimized processes. Enzymatic lignin biodegradation, for example, has undergone a long evolutionary adaptation and is an integral process in ecosystem functioning. The inves-tigation and understanding of such functional mechanisms allow for the development of new strategies for material modification. Biomole-cules such as proteins offer a huge variety of interesting properties. They are provided by nature, mostly cheap and can often be used as a greener alternative to conventional chemistry. In our group we combine these biomolecular functionalities with the structure of wood and wood-based materials. This includes isolated and immobilized proteins, as well as microbial cells (living cells or entire biofilms). As a re-sult, wood-bio hybrid materials are developed that shwo interesting new properties and offer fascinating functions. As a special feature, we incor-porate machine learning techniques particularly biologically inspired deep neural networks in our research. This technique is a helpful tool for the development of wood-bio hybrids and enables us to analyze and optimize (bio-)processes more efficiently.

 

 


 

Research topic: Efficient and eco-friendly protection of wood against microorganisms by enzymatic iodination

/documents/55954/0/laccase+iodide+treatment/917bcba0-3012-4c86-a17b-946fcd768d0f?t=1463493873913

Traditionally wood durability has been improved by using preservatives containing toxic chemicals. Apart from the health risks involved in using toxic substances for wood treatment, there is increasing concern about the problems arising from the disposal of the timber at the end of their commercial lifetime. We developed a more eco-friendly method for preserving wood and wood-derived materials. The approach uses and enzymatic modification to preserve these materials against several microorganisms including bacteria, yeast, blue stain fungi and wood decay fungi. In the process iodide(I-) is oxidized to iodine (12) by laccase in the presence of wood leading to an enhanced resistance of the wood surface against all microorganisms even after exposure to leaching. 

In collaboration with our industrial partner we improve the weathering durability of wood and wooden facades on basis of this laccase-catalyzed wood iodination.

Research topic: Deep neural network for analysing and prediction of laccase-catalysed oxidation of different substrates

/documents/55954/79760/Deep+believe+neural+network/c19694ac-d4ef-4e6c-adba-a29809718ec6?t=1540045786357

In spite of the versatility of the enzyme laccase (E.C. 1.10.3.2) for different industrial applications the full potential has not been exploited yet, due to several reasons. The substrates of interest may not be oxidized by the laccase directly if it is too large to penetrate into the enzyme's active site or exceeding redox potential. Thus an intermediate substrate (mediator), whose oxidized radical forms interacht with the target compound is need. Furthermore, laccases from different origins differ in terms of the architecture of the active site and oxidation potential. Hence, the capability of laccases to oxidize a specific substrate is hard to predict and the search for suitable laccase-substrate (mediators) couples for certain biotechnological processes still requires time-consuming and cost-intensive experimental screenings.

To overcome this problem, we developed a robust and reliable model based on deep learning techniques in combination with sensitivity analysis allowing the prediction of laccase-catalyzed oxidation of different substrates and the determination of the underlying reaction mechanisms. Such predictions help to prioritize the experiments will substantially reduce the experimental work that needs to be carried out and allow us to apply laccases more efficiently.

 

 

 

Research topic: Application of machine learning in the manufacturing processes of wood fibre insulation boards

Wood fiber insulation boards are an important product of the Swiss wood industry. Like in other manufacturing industries, the quality control of the products is assessed time-delayed after the production process. Consequently, there is the risk to produce fiberboards with low or even insufficient quality. Our aim is to generate a model on basis of deep neural networks which analyses the machine- and process data on-line and allows the prediction of the product quality in real-time. Hence, adjustments can be made either by the plant operator or automa-tically, before the quality of the product decreases. The advantages would be more uptime, less downtime, optimized material consumption, less waste and consistent product quality. 

 

 

/documents/55954/79760/Use+of+deep+neural+networks+in+the+production+process/cbcdff2b-d782-4831-b217-645922f42eba?t=1540049534535

Research topic: Catalytically active wood-bio-hybrids

Catalysis is needed for chemical processes in almost every industrial field. Catalysts are mostly made from noble metals what makes them expensive and unsustainable. Enzymes are a greener alternative to the conventional catalysts. These biocatalysts are usually more efficient, faster, have higher yields, and are more specific than the artificial catalysts. However, enzymes have a low stability due to their fragile 3D-structure. In order to stabilize them and make them more suitable for industrial applications, they can be immobilized to solid scaffolds.

Wood with its hierarchical porous composition is an interesting material for enzyme immobilization. It offers a high inner surface area as well as anisotropic flow through. Hence, in our work we create a flow-reactor-like material from wood and enzymes. The variety of different enzy-mes and different structures of different wood species makes the approach promising for numerous applications.

 

 

Research topic: Thermo-responsive gating membranes

Wood is a promising material for the development of new functional membranes. It offers dimensional stability, anisotropic flow-through and a high permeability for liquids. In this project we aim to develop a thermo-responsive gating-membrane by the modification of wood with thermo-responsive biomolecules. Our approach is simple to execute and purely water-based. Hence, the final product is cheap and sustaina-ble. Furthermore, the membrane can easily be chemically modified to add further functionalities. This makes the material promising for various applications.