Skip to main content

Development and application of ultra-high resolution nano-organized films by self-assembly of plant-based materials for next generation opto- and bio-electronics

Final Report Summary - GREENANOFILMS (Development and application of ultra-high resolution nano-organized films by self-assembly of plant-based materials for next generation opto- and bio-electronics)

Executive Summary:
The aim of the GreeNanoFilms project is to develop and apply ultra-high resolution nano-organized films by self-assembly of plant-based materials for next generation opto- and bio-electronics. (Figure 1)
Carbohydrate biomass constitutes an abundant and renewable resource that is attracting growing interest as a biomaterial. Convincingly the use of different natural “elementary bricks”, from oligosaccharides to fibers found in biomass, when mimicking self-assembly as Nature does, is a promising field towards innovative nanostructured biomaterials, leading to eco-friendly manufacturing processes of various devices. Indeed, the self-assembly at the nanoscale level of plant-based materials, via an elegant bottom-up approach, allows reaching very high-resolution patterning (sub-10nm) never attained to date by petroleum-based molecules, thus providing them with novel properties.

See Figure 1: Innovation brought by GreeNanoFilms (attached doc)

Ten research and industry partners with different backgrounds, going from materials science and engineering, to chemistry, physics, electronics and micro/nano-technologies worked in a multidisciplinary approach and developed technical films to be used in various markets, from large volume sectors, such as :
(i) high-added value transparent flexible substrate for printed electronic applications,
(ii) thin films for high-efficiency organic photovoltaics,
to growing markets, such as :
(iii) next generation nanolithography and
(iv) high-sensitivity SERS biosensors.

GreeNanoFilms main impacts are the implementation of a new generation of ultra-nanostructured carbohydrate-materials that will play a prominent role in the achievement of the sustainability improvement of various opto- and bio-electronic sectors. A network of industrial end-user leaders is integrated in the project to facilitate the innovator-to-market perspective. The prospective environmental impacts and benefits of new green processes, eco-efficient nanomaterials and nanoproducts will be quantified with Life Cycle Assessment, risk assessment and validation of the industrial feasibility, including economic evaluation of the products. The results will be disseminated to the European smart paper, printed electronic, photovoltaic, display, security and health communities.

Project Context and Objectives:
Concept - Current knowledge in modern molecular science allows for the preparation of a myriad of tailored nanomaterials, which play important, multifaceted roles in nanoscience and technology. Among the bottom-up strategies, self-assembly is an incredibly powerful concept in macromolecular engineering that offers an invaluable tool for the preparation of 2D and 3D discrete nanostructures, ranging from materials science to molecular biology, which are often not accessible by any other fabrication process.

Specifically, the huge and growing interest in high-resolution patterned films derives from the promise of manipulating matter atom-by-atom and molecule-by-molecule using the bottom-up approach, via self-assembly, to create the next generation of materials with improved performance and functionality. The novel products are expected to be smaller in size, orders-of-magnitude better and more efficient than those provided by conventional manufacturing technologies.

Attempts were made during the last few decades to attain these objectives. However some limitations in achieving high resolution patterned films were encountered. These were mainly due to a relatively weak interfacial incompatibility between the elementary building bricks (molecules and macromolecules derived mainly from petroleum) making up the films. Therefore, one key step towards novel applications with highly nanostructured materials concerns the appropriate and smart choice of the “elementary building bricks” as well as their functionalization, orientation, alignment and ultimately their self-assembly and nano-patterning that makes them ready to be implemented onto innovative nanomaterials.

To date, such materials are derived from fossil resources that are being rapidly depleted and have negative environmental impacts. In contrast, carbohydrates are abundant, renewable and constitute a sustainable source of materials. This is currently attracting much interest in various sectors and their industrial applications at the nanoscale level will have to expand quickly in response to the transition to a bio-based economy. Additionally, their self-assembly at the nanoscale level via the bottom-up approach, has allowed only recently (CERMAV’s innovation 2012) the realization of very high-resolution patterning (sub 10nm). This resolution has never been attained to date by petroleum-based molecules and provides these new new materials with novel properties. To this end and to meet the challenge of the 21st century that will hopefully open new horizons for the valorization of biosourced functional materials at the nanoscale level, the aim of this proposal is to design and develop, using environmentally friendly processes, ultra-high resolution patterned nano-organized films, obtained via the self-assembly of elementary bricks of plant-based biomaterials (glycopolymers, cellulose nanocrystals (CNC) and cellulose nanofibrils (CNF)) for innovative high-added value applications targeted in GREENANOFILMS: e.g. Transparent technical films, Flexible opto-electronic devices and High-Sensitivity SERS (Surface-Enhancement Raman scattering) Biosensors (produced from next generation nanolithography) as illustrated in Figure 1 (see attached doc).

Objectives - The project will focus on fundamental investigations of self-assembly of new bio-sourced amphiphilic copolymers (glycopolymers), on the functionalization, the surface engineering and processing of CNC, on the manufacturing of transparent technical CNF films, and will demonstrate the potential of the technology with selected innovative applications.

The scientific and technical objectives of GREENANOFILMS are:
Objective 1:
(a) Conception and synthesis of new hybrid bio-sourced amphiphilic glycopolymer systems (synthetic-boligosaccharide and semi-conducting-b-oligosaccharide) using new strategy routes: “metal-free click chemistry”.
(b) Mastering the self-assembly of those carbohydrates elementary bricks-by-bricks, particularly, at the nanoscale level to reach very high-resolution nano-patterning films never attained to date. The objective is to outperform petroleum-based self-assembly materials to reach (sub-10nm resolution). Control of the Nanodomain orientations: 2D and 3D long-range order and defect-free.
(c) Use those High-nanostructured films made from glyco-block copolymer systems (BCP) as template for next generation mask nanolithography on different hard and flexible substrates (cellulose films) and as “active films” for various opto- and bio-electronic applications (OPV and biosensors: see objectives 2).
Objective 2:
Exploiting the properties of elementary bricks of plant-based biomaterials, glycopolymers, cellulose nanocrystals (CNC) and nanofibrils (CNF) at the meso-scale and further their processing at the macro-scale towards the elaboration of:

(a) Bio-sourced transparent technical films: for smart papers and surfaces (flexible electronic functional substrates/displays) for large volume sectors: conception and roll-to-roll (R2R) realization of CNF films in pilot scale with targeted properties:
• Transparent, ultra-smooth, super-hydrophobic and gas barrier (chromatogenic technology and/or atomic layer deposition).
• Nanopatterning using nano-imprint lithography of CNF films with 10-50 nm resolution.
• Coating or printing functionalized-CNC layer on CNF films for structural orientation (nanopatterned surfaces) leading to conductivity (sheet resistance < 1kΩ/square with transmission >75% in the visible) and anti-reflective properties (rest reflection < 1% in the visible)
• Estimated costs depend on the modifications needed, simple films are in the range of 0.2-0.5 €/m2

(b) Flexible opto-electronic devices (specifically organic photovoltaic) for high added value sectors.
Conception and development of donor-acceptor nano-organized thin films (5 nm resolution) made from glyco-semiconducting block copolymers (BCP) used in OPV to favor exciton travel across the interfaces (below 10 nm), thus improving their efficiency (objective Power Conversion Efficiency > 8%): Nano-structured template using directed self-assembly (DSA) on ultra-low-cost (<1€/Wp) and low energy payback (< 6 months) bulk hetero-junction organic photovoltaic OPV on one side and on high-added value technical films (CNF & CNC) on the other side.

(c) High resolution block copolymer nanolithography for the realization of high sensitivity SERS biosensors:
Conception and development of well-organized arrays of gold-based nanostructures yielding high sensitivity biomolecule detection characterized by Enhanced Surface Plasmon Resonance (ESRP). This will be achieved by a controlled deposition of Au on nano-organized thin film templates (objective to demonstrate 5 nm-10 nm features on the area of several μm2) made from the self-assembly of oligosaccharide-based copolymer and followed by reactive ion etching (RIE).
Obviously, neither the appropriate substrate -having the right balance between dimensional stability, barrier properties, temperature resistance, transparency, roughness, conductivity and competitive price -nor the targeted applications based on the proposed approach- are available yet. Consequently, additional and important R&D efforts are needed for the valorization of bio-sourced functional materials. To reach our objectives, we will put important efforts on:

• Preparation, purification, synthesis, chemical modification and characterization of elementary plant-based bricks
(oligosaccharides, glycopolymers, CNF, CNC); architectural design of block copolymer, surface modifications of
nanocrystals and their characterizations at the molecular level with the state-of-the art techniques including surface interactions, imaging and scattering (light scattering, SAXS & GISAXS, ESRF available at Grenoble) techniques
• Self-assembly and Directed Self-assembly (DSA) of plant-based bricks as well as the characterization of the obtained films from (glycopolymers, CNC & CNF); control of nano-domain orientations (5-10 nm resolution) for BCP thin films on different substrates (hard and flexible): long-range ordering (2D & 3D), thickness optimization, solvents and thermal annealing,
• Robust and transparent technical CNF films with high O2 barrier properties and water resistance using chromatogeny and micro-scale patterning; nano-patterning of CNF films using NIL with the resolution of 10-50 nm; Realization of anti-reflection coatings with CNC as well as conductive coatings with functionalized-CNC with conductive oligomers / polymers to render the CNF films conducting for OPV applications.

The short and long term target applications of GREENANOFILMS in terms to technology readiness levels and their injection in the mass and customized markets are illustrated below (Figure 2):

Figure 2: Target Applications versus Technology Readiness Levels (TRL) and Markets - (see attached doc).

In the early stage of innovation, it is important to execute prospective technology and sustainability assessment. GREENANOFILMS will use assessment tools such as Life Cycle Assessment (LCA) and Risk Assessment (RA) to reduce the environmental impacts of new green processes, eco-efficient nanomaterials, new sustainable high added value products and the possible risks. Also potential effects of the increasing material complexity on the recyclability of the new nanomaterials will be evaluated. Safety will be ensured using reliable and generally accepted procedures developed in close cooperation with European authorities. The determined life cycle data of nanomaterials and nanoproducts will be directly supplied to the International Life Cycle Data System (ILCD). Bio-based materials are in general biodegradable and cellulose nanofibrils according to the first toxicological data obtained during the SUNPAP project do not indicate major safety concerns. Prospective costs and social aspects will be examined in the context of a socio-economic sustainability assessment.

For risk and safety related aspects, GREENANOFILMS will be integrated in the Nanosafety Cluster with existing project partners and collaborate with other relevant projects (e.g. NanoSustain, NanoValid, and MARINA). The Communication on the Second Regulatory Review on Nanomaterials recently adopted by the European Commission, points out that the 4th generation nanomaterials - with molecule-by-molecule and self-assembly capabilities - are at a research and development stage and are lacking in information for regulatory relevance. GREENANOFILMS will improve availability of information on these nanomaterials and will help to develop guidance for the future REACH modifications.
The consortium of this project consists of ten European partners (four universities, four technical centers and two industries) under Cermav's coordination (Table 1) - (see attached doc)..

1. Centre de Recherches sur les Macromolécules Végétales (Cermav), Grenoble, France. Contact: Redouane Borsali.
2. Centre Technique de l’industrie des Papiers, cartons et celluloses (CTP), Gières, France. Contact: Davy Soysouvanh
3. Teknologian Tutkimuskeskus VTT, Espoo, Finland. Contact: Ulla Forsström
4. University of Lund, Sweden. Contact: Ivan Maximov
5. Produits Chimiques Auxiliaires et de Synthèse SA (PCAS), Longjumeau, France. Contact: Pierre-Antoine Bonnardel
6. University of Bremen, Germany. Contact: Michael Steinfeldt
7. Disasolar SAS, Limoges, France. Contact: Laurence Dassas – terminated in May 2015
8. Obducat Technologies AB, Lund, Sweden. Contact: Johan Ring
9. Nederlandse Organisatie Voor Toegepast Natuurwetenschappelijk Onderzoek (TNO), Delft, Pays-Bas. Contact: Pascal Buskens
10. The International Iberian Nanotechnology Laboratory (INL), Braga, Spain. Contact: Lars Montelius
11. Centre suisse d'électronique et de microtechnique (CSEM), Basel, Switzerland. Contact: Giovanni Nisato – new partner form March 1st, 2016

Table 1. Partners involved in GREENANOFILMS (see attached doc).

1.3. Project Management
WP1 has been assigned to the coordination and the management during the whole period of the project. This WP was also devoted to ensure the achievement of the project’s objectives within the allocated “on time” budget and ensuring that the available resources are used in a controlled and structured manner. At every meeting, the coordinator has reminded all partners to exchange very often their points of views, especially within WPs, and also to disseminate about the project.
A Consortium Agreement (CA) has been established between the coordinator and all partners in order to regulate reciprocal relationship and especially to reconcile conflicts or default cases within the project. As mentioned in the DoW, a financial follow up on a 6 monthly basis has been inserted in the Interim reports. All partners have filled in the Form C and download it in the European Commission Platform.
A Quality Management Plan (QMP) has been defined to describe the internal rules and quality procedures. The IPR management procedure, based on the Grant Agreement and Consortium Agreement, has been also written and each partner approved it. An achievements and exploitation table is attached on this document and it defines owners or co-owners of foreground/background (see Achievement table and exploitation plan section).
The website of the project has been regularly updated with the newsletters and the different events of the project and is available at As planned, the GreeNanoFilms videos (part 1 & part 2) describing the objectives and the obtained results of the project are also available on the “greenanofilms”’s website.
The extranet allows all partners and the EC to exchange all documents and information related to the project. Such documents are available using the following link:

Project Results:
1.4. Main S&T results/foregrounds
1.4.1. WP2: Self-Assembly
This "Self-Assembly" WP is focus on the design of diblock copolymers (BCPs) to be used as active thin film for the target applications. It includes the preparation and characterization of functionalized oligosaccharides, synthetic and semi-conducting polymers and amphiphilic BCP systems to master self-assembly with sub-10 nm resolution. The subsequent preparation of thin films is conducted in controlled conditions. Directed Self Assembly (DSA) technique is applied to provide long-range nano-ordering and thickness optimized films. The developed BCP have been supplied to WP4 for the realization of patterned CNF/CNC films. Donor-acceptor BCP have been evaluated for OPV application (WP5). Amphiphilic BCP will also be used for next generation nanolithography and high-sensitive Biosensors in WP6.

Task 2.1: Preparation & characterization of oligosaccharides
In the perspective to obtain size-homogeneous domains in oligosaccharide based-BCPs thin films, we hypothesize that the access to oligosaccharides with well-defined structures and of controlled molecular weight is of great significance. For such purpose, we have identified three readily accessible oligosaccharides: maltoheptaose (MH), xylo-glucooligosaccharides (XGO) and β-cyclodextrin (βCD).
“Clickable” oligosaccharides were obtained at few tens of grams by regio- and stereo-selective functionalization of the anomeric position (reducing-end) with alkyne groups without the fastidious requirement of the protection/deprotection strategy (Figure 3). (See attached doc)

Figure 3: N-acetyl propargyl glycosylamides of XGOs, MH and propargyl-βCD (see attached doc).

Task 2.2: Conducting polymers & PCBM compounds
The synthesis of end functionalized synthetic polymers (Polystyrene (PS), Poly(3-hexyl)thiophene (P3HT), Polycarbazole (PCBz) and fullerene (PCBM)) have been initially proposed in this task.
The azide-terminated PS was obtained at a few tens of grams from a two-step procedure consisting in a tosylation of the mono hydroxyl end functionalized PS-OH (Mw = 4500 g.mol-1; PDI = 1.10) and a subsequent nucleophilic substitution by using sodium azide in excess (Figure 4). (See attached doc)
Figure 4: Synthesis of PS-N3 from commercially available PS-OH (see attached doc).

The synthesis of the end-functionalized P3HT (initially planned to be done by PCAS) has not been achieved despite the many efforts provided by PCAS. Therefore, during a steering committee, it was decided that CERMAV takes in charge the synthesis of 10 g of clickable-end functionalized P3HT according to a validated procedure developed in this Laboratory (Figure 5) 8.6 g of P3HT-alkyne have been successfully obtained.
Figure 5: Synthesis of Alkyne-Terminated P3HT (Alkyne-P3HT) by Modified Grignard Metathesis Polymerization (see attached doc).

The synthesis of the hydroxyl end-functionalized PCBz (in spite of many experimental conditions) has not been successfully achieved by PCAS because of the presence of a large range of complex by-products. This semi-conducting polymer has not been used for coupling reaction and OPV applications.
PCAS supplied to CERMAV in March 2015 around 20 g pure batches of each (C60)PCBM and (C60)PCBA. (Figure 6) - (see attached doc).

Figure 6: Structures of the (C60)PCBM and the (C60)PCBA (see attached doc).

Task 2.3: Synthesis of amphiphilic glycopolymer systems
We have been able to prepare in a single batch almost 10g of purified PS-b-oligosaccharides (MH, XGO & βCD) thanks to traditional click chemistry using Copper nanoparticles (CuNPs) as catalyst (Figure 7). The Copper is totally removed by using a specific commercially available Cuprisorb® resin. The analytical data shows that we have almost totally removed copper, which was detected only at ppb level by atomic absorption spectroscopy. This purification was done for all hybrid sugar-based BCPs including P3HT-b-MH.

Figure 7: CuAAC click coupling between reducing-end alkyne-containing oligosaccharides and azide-terminated synthetic
(see attached doc)

Alternatively, a metal-free click chemistry approach has been developed which combines cleavable feature useful for the elaboration of nanoporous thin-films. (Figure 8)
Figure 8: Thiol-maleimide addition as metal-free click chemistry approach for the synthesis of sugar-based BCPs

The synthesis of PCBz-b-oligosaccharides has not been performed since PCAS was not able to produce with a sufficient purity the end-functionalized PCBz.

Task 2.4: Thin film: domain orientation & annealing process
CERMAV has studied the surface treatment of silicon (100) as “hard” substrates for spin-coating Polystyrene-b-Maltoheptaose (PS-b-MH), which will be used for next generation nanolithography and high-sensitive biosensors in WP6. CERMAV group has been used the so-called Piranha solution to clean organic residues off the silicon substrate surface and make the surface hydrophilic to increase the affinity between the substrate and PS-b-MH. The Piranha solution treatment is generally used for spin-coating of block copolymer (BCP) and gave nice thin films of PS-b-MH system. However, strong acidic condition (dipping substrates into a mixture of H2SO4 and H2O2 at 80 °C for 1 hour) of this treatment could damage pre-patterned substrates for directed self-assembly (DSA) which will be used for long-range ordered self-organization in the WP2, WP5 and WP6. Thus, oxygen plasma treatment was investigated as an alternative method for the substrate treatment. We have shown that 30 seconds of oxygen plasma treatment is good enough to have hydrophilic surface of silicon (100) substrate instead of long piranha treatment.
Figure 9 (see attached doc) shows AFM topological images of DSA line patterned substrates prepared by LUND group before and after oxygen plasma treatment. The maximum height of the line patterning in the images before and after plasma treatment was almost the same (ca. 35 nm) indicating that there was no damage on the line patterning by the oxygen plasma treatment. From these results, we decided to use oxygen plasma method for the surface treatment of silicon (100) substrates.
Figure 9: AFM topological images of DSA line patterned substrates before and after oxygen plasma treatment
(see attached doc).

CERMAV’s group already reported that the film thickness of PS-b-MH can be controlled from 20 nm to 280 nm by changing the concentration of spin-coating solution.1 As the height of the line patterning on the DSA substrate prepared by LUND group is about 13 nm, we focused on controlling the film thickness less than 13 nm. Thus, we prepared 4 different concentrations of PS-b-MH in anisole (solvent) and spun-coat on the silicon (100) substrates treated with oxygen plasma. The ellipsometry measurements showed that the film thickness was almost proportional to the concentration and we could successfully control the film thickness less than 15 nm.
The obtained PS-b-MH thin film was self-organized by annealing under vapor from a mixture of THF and water (THF/water = 4/1 w/w). A hexagonally close-packed cylinder phase was successfully self-organized as illustrated in the AFM phase image (Figure 10). Therefore, oxygen plasma treatment of the substrate did not change the annealing condition for the self-assembly of PS-b-MH thin film.

Figure 10: AFM phase image of PS-b-MH thin film after solvent vapour annealing with THF/water = 4/1 (w/w). Insets are 2D Fourier-transformed image. (see attached doc)

Task 2.5: Directed self-assembly (DSA) & plasma etching
The CERMAV reported the PS-b-MH self-organization into vertical or horizontal hexagonally close-packed cylinder phase by solvent annealing1. Surface neutralization is a common way used to induce vertical orientation of the nano-domains formed in self-assembled block copolymer (BCP) films. This approach consists in the modification of the substrate surface energy in order to obtain a surface which is not more favorable for one polymer block than for the other. In this way, lamellar and cylindrical organization can be obtained with a vertical orientation.
A way to modify the substrate surface energy is to graft one of the homopolymer forming the BCP or directly the BCP. The PS, MH or PS-b-MH grafting has been carried out by condensation between the hydroxy groups of plasma treated Si wafers (O2 plasma at 75 W for 10 min) and the hydroxy groups of a monohydroxy terminated polystyrene (PS-OH), the MH or the PS-b-MH, respectively. The chemical modification of the Si wafers leads to a change to the substrate surface energy and hydrophobicity. And as expected, the substrate surface energies, calculated by advancing/receding water contact angles6, increases while the hydrophilicity increases.
The MH-b-PS, synthesized using Cu nanopowder, has been spin-coated onto the four different substrates (40 nm thick MH-b-PS films) in order to study the effect of the surface energy on the BCP self-organization and the nano-domain orientation. After THF annealing, it seems that, independently of the surface nature, the BCP is nano-organized in hexagonal close-packed vertical cylinders but this organization is not as clear as for the 15/1 THF/H2O annealing.
Newt, we investigated the effect of solvent annealing temperature on the self-assembly of MH-b-PS. We demonstrated that the self-assembly of MH-b-PS becomes faster when the annealing temperature is increased from 23 °C to 30 °C.
The effect of solvent annealing temperature on the self-assembly of MH-b-PS from solvent mixtures of different THF/Water ratios was also studied. With THF/H2O=5/1 the orientation of some MH-b-PS cylinders changes from horizontal to vertical when the annealing temperature is increased from 23 °C to 30 °C. The orientation of the vertical MH-b-PS cylinders changed back to horizontal for THF/H2O=5/1 at 40 °C.

1.4.2. WP3: CNF Film Processing
The main objectives of the WP3 were the production of cellulose nanofibrils (CNF) using fibrillation procedures involving chemical and mechanical treatments as well as the quality control of all the materials involved (pulp, fibrils and films). The objective in the WP was also to manufacture technical CNF films and functionalize those via chromatogeny grafting, atomic layer deposition (ALD) and patterning with different length scales. Macroscale patternings were constructed in WP4 using roll-to-roll nanoimprinting lithography (NIL) whereas in this WP the nanoscale patterns were generated by self-assembling sugar-based glycopolymers on CNF films. The overall goal of the functionalizations was to control the film water susceptibility, to enhance barrier performance and to modify the surface architecture in order meet the demands of selected GNF final applications (Optical photovoltaics and High-sensitive Biosensors).

Task 3.1: CNF production and film manufacturing
Several batches of CNF have been produced during the project including mechano-enzymatic CNF production, chemically pre-modified CNF production (TEMPO-oxidised CNF) and chemically unmodified CNF production (VTTref CNF). Fibrillation procedures involving chemical (TEMPO oxidation) or mechano-enzymatic pulp pretreatments followed by several passes into the homogenizer were carried out at CTP and VTTref production which involved several passes through the fluidizer was carried out at VTT. Despite of the increasing number of passes through the homogenizer, the mechano-enzymatic CNF (endoglucanase pulp treatment followed by a drastic refining then homogenization), it was not possible to reach the film manufacturing requirements i.e. high gel strength and high amount of nanosized materials. Therefore, the focus was set on VTTref CNF (native CNF grade from bleached birch pulp) and TEMPO CNF (TEMPO-oxidised CNF). The protocol for TEMPO oxidation of the pulp was adapted to take into account both the pulp grade (Bleached birch kraft pulp), the requested amounts (20 liters of CNF suspension at 1% concentration minimum) and VTT’s film manufacturing process requirements (high gel strength, high amount of nanosized materials). Several productions were done during the first 18 months and in the end, the protocol for the production of TEMPO oxidized CNF suspensions was stabilized and good enough for CNF film production. VTT also used the protocol developed by Saito and Isogai, University of Tokyo to produce TEMPO-oxidised CNF from bleached spruce pulp. In addition, several batches of microfibrillated celluloses (MFC) were produced at CTP to make model films in lab scale, which were used to gain better understanding of the phenomena happening when turning CNF films hydrophobic by chromatogeny grafting.
The CNF film production method used in the project is patented and is based on cast coating of the gel-like CNF suspension on the supporting substrate [T. Tammelin, A. Salminen, U.Hippi Patent WO2013/060934 (2013).]. CNF film production is based on precise control of adhesion, spreading and drying of CNF on a plastic substrate. A defined amount of CNF suspension with softener (sorbitol) and/or strength additive (polyvinyl alcohol (PVA): Mowiol or Alta Aesar) is cast coated on a pre-treated plastic substrate in semi-pilot scale using a commercial converting machine (Coatema line). After drying (after evaporation of water) the film is separated from the plastic substrate as a self-standing structure. The mechanical performance and barrier performance (oxygen transmission rates and water vapour transmission rates) of the CNF films used in the project are tabulated in Table 1. TEMPO CNF film passed the milestone threshold roughness (<100nm) with the measured average roughness of 82 nm (~mm scale area) and the VTTref CNF film was very close to the milestone threshold with an average roughness of 122 nm. Micrometer scale roughness values for TEMPO CNF film and VTTref CNF film were 3 nm and 45 nm, respectively.
Table 2. Mechanical performance and barrier performance of the CNF films (see attached doc)

The strain values of the TEMPO oxidised CNF films are relatively low but can be slightly improved by using sorbitol. PVA addition gives improved tensile strength values and simultaneously the oxygen barrier performance is on the good level. For example oxygen transmission rate for synthetic LDPE is 218 mL mm m-2 day-1. However the water vapour transmission rates are high; if compared to e.g. LDPE which gives values as low as 2.5 g m-2 day-1. Based on these findings it can be concluded that the efficient utilisation of TEMPO oxidised CNF films with respect to barrier and mechanical performance can be foreseen in multilayer structures not necessarily as self-standing films, and this approach was fully in the scope of the GreeNanoFilms project.

Task 3.2: CNF film surface functionalization
Several strategies were utilized in order to improve the water susceptibility and barrier performance of the CNF films. In order to turn CNF films hydrophobic, a chemical modification, named chromatogeny, was used. Chromatogeny is a solvent free chemical grafting process that involves the reaction between the hydroxyl groups present at the surface of a given material and long chain fatty acid chloride (Figure 11). (see attached doc)

Figure 11: Reaction between a fatty acid chloride and a hydrophilic substrate (see attached doc)
This chemistry has been demonstrated at industrial scale on porous materials (papers) and non-porous materials (polyvinyl alcohol coatings). However this technology had never been assessed with high density cellulosic materials such as CNF films. During the project, several laboratory studies and pilot trials were carried out to understand the real potential of the chromatogeny to develop the CNF film hydrophobicity.
It was demonstrated that the chromatogeny process was efficient to turn CNF films hydrophobic since contact angle with water of the grafted CNF films were raised up to 120° as soon as grafting occurred; even with very low grafting density (Figure 12) (see attached doc). Unfortunately, at low grafting density, the CNF films kept a moderately high water absorption capacity showing thus their sensitivity to water It demonstrated that the increase of the grafting density decreased the water absorption capacity. This had however no further effect on the water contact angle that remained in the range 120°-125° whatever the pulp origin (amongst unbleached and bleached kraft birch and bleached kraft softwood) or the naturally occurring roughness (in the range of Ra 0.3µm to 1.5µm). Successive graftings were proved to be the efficient way to increase the grafting density. In the end, it was possible to modify a 50 g/m² CNF film in order to get low water absorption capacity (Cobb60s of 4g/m² after grafting, compared to 45g/m² before grafting), high water contact angle (120° after grafting, compared to 50° before grafting) and low surface energy (23mJ/m² after grafting, compared to 50mJ/m² before grafting). These performances are remarkable and quite sufficient for many applications (Figure 13) (see attached doc).

Figure 12: Water absorption capacity of pilot and laboratory grafted MFC films (see attached doc)

Figure 13: Water contact angle of pilot and laboratory grafted MFC films (see attached doc)
The concept of using successive iterative grafting was also validated at pilot scale. Despite the very short time for grafting (4 seconds at pilot scale compared to 600 s at laboratory scale), the pilot grafting appeared to be efficient to graft relevant amounts of fatty acids to get high water contact angle and a noticeable decrease of the Cobb60s water absorption capacity. The performances of pilot grafted MFC films remained however lower than the ones gained at laboratory scale and higher number of passes on the machine was needed to reach the same level of performance.
It has thus been demonstrated for the first time ever that the chromatogeny technology, operated several times, is a viable technology to turn CNF films hydrophobic on a reel-to-reel machine.
A simple surface silylation procedure was utilized as a reference for the CNF film surface hydrophobisation in order to show the effect of surface pattern on hydrophobisation (lotus leaf effect). Surface hydrophobization of CNF films by roll-to-roll HMDSO plasma deposition and nanoimprinting gave evidence that NIL increases the hydrophobicity level of the CNF film. Static water contact angles raised from 103 degrees up to 130 degrees due to the NIL pattern.
Atomic Layer Deposition (ALD) surface treatment was used as a method to efficiently improve especially the oxygen barrier performance of the CNF films. During the project, the nanoscaled multilayer concept was developed in order to further enhance the oxygen and water vapor barrier performance, and simultaneously limit the influence of water molecules on the barrier performance. This was carried out by constructing nanoscaled SiO2/Al2O3 with individual layer thickness values of 3.7 nm and 2.6 nm, respectively, combined with thin 5 nm protective SiO2 layer as a top layer with the aim to create highly tortuous pathway for the molecules to passage through the inorganic layer.
Significant improvements related to the ALD process itself were achieved: Thermal ALD processes based on AP-LTO® 330 (a silicon precursor) and ozone was utilised at low temperatures which has been limited due to the low reactivity of ozone at the temperatures below 300 °C in cross-flow ALD mode. This was overcome by exploiting the ALD reactor in stop-flow mode with increased precursor residence time which enabled SiO2 deposition even at 80 °C with a deposition rate of 1.5 Å(cycle)-1. Lower temperatures are beneficial when considering biomaterial functionalisations in order to avoid the damaging of the biomaterials at high temperatures.
It was shown that by using a multilayer concept of Al2O3 and SiO2 nanolayers, drastically lowered OTR and WVTR values were achieved: OTR at 80RH% varied between 0.15-0.5 mL mm m-2 day-1 and WVTR at 100/50 RH% varied between 65-91 g m-2 day-1. These values are measured using rather harsh conditions for the films prepared using biomaterials. However, the transmission rates must be further improved when considering real OPV applications and further work is still needed in order to combine all individual film treatment concepts (chemical modification + patterning + ALD). On the whole, all the treatments can be successfully performed with desired performance, and therefore, these positive achievements warrant the further work to meet the demands derived from the applications.

Task 3.3: Glycopolymer self-assembly on CNF film
Here the systematic studies on glycopolymer self-assembly have been conducted by two different approaches.
1) Molecular level self-assembly and glycopolymer interactions with nanocellulosic materials have been investigated by means of model film approach using surface sensitive methods (Quartz crystal microbalance with dissipation monitoring (QCM-D) and AFM.
2) Patterns of self-assembled glycopolymer structures have been fabricated on solid CNF substrates and on flexible CNF films by spincoating.
To demonstrate the feasibility of the concept exploiting the naturally sourced nanomaterials, two different grades of CNF were utilized, namely, mechanically disintegrated cellulose nanofibrils (VTTref CNF) and TEMPO-oxidized cellulose nanofibrils (TEMPO CNF). It has been shown that the highest amounts (~9nm) of adsorbed maltoheptaose based block copolymer (BCP) on CNF surfaces were achieved using water:THF ratios which are close to the cloud point. The focal driving force for the attractive interactions seems to be related to the solubility of the block copolymer i.e. polymer adsorbs when polymer-surface and polymer-polymer contacts dominate over the polymer-solvent contacts. Weak interactions favor the self-assembly to take place due to the fact that certain amount of chain mobility is needed for the polymer blocks to undergo spontaneous rearrangement needed to build up patterns.
Self-assembled structures of sugar based block copolymers can be fabricated on both solid and rigid CNF films (fibrils spincoated on smooth silica substrate) and on flexible, self-standing CNF films (TEMPO CNF films and CNF films from VTTref). Patterned structures can be formed despite of the CNF surface roughness. Indeed the horizontal glycopolymer orientation seems to follow the nanoscaled topography of the CNF surface, (see Figure 14). (see attached doc)

Figure 14: Horizontal orientation of CBCP on a freestanding CNF film can be observed in the AFM 3D image of a sample and in the cross sectional TEM image of the structure. (see attached doc)
Vertical orientation is more challenging and the self-assembly was successful only on ultrathin films of CNF, both TEMPO CNF and VTTref CNF. Pattern features with low block distance (5-15nm) have been achieved. Nanoscale level self-assembly of plant-based materials i.e. cellulose nanofibrils (CNF) and glycopolymers, with an elegant bottom up approach allows the introduction of eco-friendly manufacturing processes which unprecedentedly enables the fabrication of continuous high-resolution patterns with green, simple and upscalable processes to be utilized in next generation devices such as biosensor, security and optoelectonic applications.
1.4.3. WP4: Transparent Nanocellulose Technical Films
The primary goals for WP4 was to process and analyse roll-to-roll nanoimprinted (R2R NIL) CNF films and implement Cellulose Nanocrystal (CNC) production and coat functionalized-CNC layer on CNF films leading to conductivity and anti-reflective properties.
Task 4.1: CNF film NIL nanopatterning
Primary goal in this roll-to-roll nanoimprinting lithography (R2RNIL) task 4.1 was to develop roll-to-roll process for micro and nanoscale patterning of native nanocellulose films. Pattern features from micron scale down to ca. 50 nm scale was targeted. Film was selected and developed in WP3 to fill demands on nanoimprinting process. This was the first time when these films have been used in this kind of processing. In the future this method opens a way to use nanocellulose film in various applications as recyclable and biodegradable solutions for use in, for example, printed optics, decorative packages, optoelectronics and disposable medical devices. During to project we developed a relatively simple method to modify biobased nanocellulosic films with thermal R2RNIL. We demonstrated micro- and nanostructured patterning on films by manufacturing micropillar and optical grating structures. By itself these patterns can already be used as decorative purposed e.g. in recyclable packages.
In the NIL process a patterned roll and an elastic backing roll are pressed against each other at elevated temperatures, and the pattern is replicated onto the film structure. We demonstrated the fabrication of patterned nanocellulosic films prepared using mechanically disintegrated cellulose nanofibrils (CNF) and TEMPO-oxidised cellulose nanofibrils (TEMPO-CNF). Although both CNF films are brittle and nanocellulosic material itself is not thermoplastic with a clear softening point, it was shown that the mechanical patterning of both CNF films was possible. Distinct temperature dependency was observed with both films above 70 oC. The height of the pillars increased almost linearly up to 155 oC with a constant pressure of 8 MPa. The temperature increase seems to improve the replication. Simultaneously the nanoimprinting process smoothed the film surface roughness. AFM images show that edges of the printed pillars are quite blunt. This is typical behavior for materials which are not close to their melting point during the process.
Thermal R2RNIL needs a flexible mould and therefore silicon or quartz mold is typically copied to flexible Nickel replica molds produced by nickel electroplating. Nickel replica molds are widely used in R2RNIL replication. We find the optimal roll-to-roll NIL conditions with respect to temperature and pressure, with a view to pattern structures that generate an optical effect. The NIL-process temperature was varied from room temperature up to 160 oC, with two pressure levels used (3.7 MPa and 8.3 MPa). Other conditions like printing speed (0.2 m/min) were kept constant.
Feature heights up to 1500 nm can be imprinted when the mould depth is 2200 nm. Furthermore, pressure and temperature are essential parameters in the roll-to-roll process for CNF. Printing speed (imprint time) is also critical when thermoplastic polymer replication of the Ni-stamp is needed. The compressibility of CNF films due to high moisture was the main reason for successful patterning. Modifying CNF films, for example, by increasing the amount of plasticizers, might improve the imprinting properties due to the ability of different plasticizers to either bind water or increase softening due to their own thermoplastic character. Reversion of CNF and TEMPO-CNF films were measured one week after imprinting after the CNF film samples had been stored at room conditions and had again absorbed moisture. Results indicate that reversion in the TEMPO-CNF structures was slightly smaller than in the case of CNF.
In this project roll-to-roll imprinted structures in CNF films were demonstrated for the first time using a custom-made printing device (Figure 15). (see attached doc)

Figure 15: Photograph of the R2RNIL patterned pillars on the TEMPO-CNF film. The structures are the first evidence of pigment less colours for using e.g. in decorative recyclable packages. (see attached doc)
Pillars with a diameter of 7.5 µm and 90 nm wide nanogratings were printed in the experiments. The height measurements show a linear increase in pillar height when temperature was increased up to 155 oC. Both CNF and TEMPO-CNF show similar behavior in the experiments. R2RNIL for CNF and TEMPO-CNF films show a huge potential for future applications in optics and electronics. The results on R2R nanoimprinted patterned CNF films (Figure 16) have been described in following publication:
T. Mäkelä; M. Kainlauri, P. Willberg-Keyriläinen; T. Tammelin, U. Forsström; Fabrication of micropillars on nanocellulose films using a roll-to-roll nanoimprinting method; Microelectronic Engineering. Elsevier. Vol. 163 (2016), 1-6; doi:10.1016/j.mee.2016.05.023

Figure 16: R2RNIL printed micropillars on TEMPO-CNF film. (see attached doc)
Task 4.2: Anti-reflective & functionalized electro-conductive CNC coatings on CNF films
CNC production and characterization
Dispersion-stable Cellulose Nanocrystals, CNC nanoparticles have been reproducibly extracted at lab and small pilot scale from commercial and consortium lignine-free cellulose sources (Figure 17). These have been characterized using state of the art techniques such as dynamic light scattering (DLS), electrophoretic mobility measurements (zeta potential), transmission electron microscopy (TEM), Cryo-TEM and X-ray diffraction (XRD). Specifications for CNC suited for use in high-tech coatings were set, and batches produced on both size scales were well within these specifications and are therefore considered to be suitable nano-sized building blocks of sufficient quality for use in high-tech coatings.

Figure 17. Transmission Electron Micrograph of CNC anisotropic nanoparticles extracted from lignin-free cellulose pulp.CNC-silver building blocks development - (see attached doc).
Figure 18. Reaction scheme of CNC coverage with silver and Transmission Electron Micrograph of CNC-silver composite particles (see attached doc).

The GreeNanoFilms partners successfully developed electrically conductive CNCs through selective deposition of Ag on CNCs (Figure 18) (see attached doc). The first synthesis route that yields CNCs with a high degree of metal coverage on the surface was reported.

CNC-electrically conductive coating product demonstrator
The CNC-silver nanobuilding blocks were exploited for the preparation of coatings with an unprecedented high conductivity of 2.9·104 S·cm-1. This exceeds the highest conductivity reported to date for cellulose composites by a factor of 30. After photonic curing, typical coatings displayed a sheet resistance in the range of 40-50 Ω/sq, demonstrating that the sheet resistance target can be realized (Figure 19). The work on closed-shell metallization of nanocellulose whiskers and the characteristics of the coatings prepared are novel to the scientific community and a manuscript of a scientific research paper has been accepted for peer-reviewed publication:
• N Meulendijks, M Burghoorn, R van Ee, M Mourad, D Mann, H Keul, G Bex, E van Veldhoven, M Verheijen, P Buskens Electrically conductive coatings consisting of Ag-decorated cellulose nanocrystals, accepted for publication in Cellulose
Figure 19. High resolution Scanning Electron Micrographs of photonically-sintered CNC-Ag coatings. CNC-silica building blocks development (see attached doc)
Figure 20. Transmission Electron Micrograph of CNC-silica core-shell particles (see attached doc)
The GreeNanoFilms partners developed silica-coated CNC particles for application quarter-wave anti-reflection coatings. Needle-shaped CNCs with an aspect ratio of 20 were extracted and subsequently covered with a silica layer (Figure 20) (see attached doc).

CNC-Optical coating product demonstrator
In one single dip coating step, highly porous coatings of CNC-silica core-shell particles were deposited on glass slides and silicon wafers. The lowest refractive index achieved was 1.03 which corresponds to an extraordinarily high porosity of 94%; the thickness of these coatings ranged from 101 nm to 239 nm. Experiments were performed to add binders to the coatings and to remove cellulose through the silica particle surfaces by means of pyrolysis, which resulted in 100% inorganic porous coatings. The ultra-low refractive index coatings realized in this study (Figure 21) has formed the basis for a series of high tech coatings with advanced functionalities such as quarter-wave antireflective coatings, multi-layer interference stacks, coatings for optical fibers with a high numerical aperture and optical adhesives. The results on CNC extraction and CNC-silica optical coatings has been described in following publications:
• P Buskens, M Mourad, N Meulendijks, R van Ee, M Burghoorn, M Verheijen, E van Veldhoven, Highly porous, ultra-low refractive index coatings produced through random packing of silicated cellulose nanocrystals, Colloids and Surfaces A: Physicochemical and Engineering Aspects, Volume 487, 20 December 2015, Pages 1-8, ISSN 0927-7757, DOI: 10.1016/j.colsurfa.2015.09.041
• P Buskens, M Burghoorn, M Mourad, and Z Vroon, Antireflective Coatings for Glass and Transparent Polymers, Langmuir 2016 32 (27), 6781-6793, DOI: 10.1021/acs.langmuir.6b00428

Figure 21. High resolution Scanning Electron Micrographs and photograph of CNC-silica antireflective coating (see attached doc)

1.4.4. WP5: Opto-electronic device
WP5 is originally focused on the development of flexible opto-electronic device using films made from BCP (WP2) deposited on CNF (in WP4) and nano-ordered BCP system developed in WP2. Such device is expected to be an OPV on flexible substrate.
Task 5.1: Bulk hetero-junction solar cell on transparent CNF film
Cellulose-based transparent conductive electrode
VTT has manufactured transparent CNF films with ~30µm thickness and >80% transparency in the visible range (380-750 nm). The rougher side of CNF films has been measured by white-light interferometry with roughness Ra ~400 nm. CTP has demonstrated that a thin transparent polymer layer with ~4-5 µm thickness can be used for both the passivation and the planarization of the CNF substrate. Conductive Ag electrode has been printed on an intermediate PET donor film by flexography and this electrode has then been transfer-laminated onto the CNF films by using a cross-linkable UV resin. The embedded grid permits to have a smooth surface, reflecting the donor substrate and the silver ink characteristics. The roughness of printed areas has been measured to ~30 nm and is due to the size distribution of ink silver particles, while the non-printed areas show a roughness of ~16 nm, very close to roughness of the donor PET substrate. The transition between printed and non-printed areas has been measured to be less than the measurement limit of our white-light interferometer (i.e. <10 nm). The smooth and transparent conductive electrode with embedded silver grid has a 0.09 Ω/ surface resistivity along the conductor direction and optical transmittance in the visible spectrum >80% (Figure 22).

Figure 22: Transparent conductive cellulose electrode (a) Embedded Ag-grid onto CNF films (b) Optical transmittance of the CNF/grid substrate and PET/ITO reference substrate. Dark-blue & green: total transmittance, Purple & red: scattered transmittance(haze). (see attached doc).

Realization of the disordered BHJ solar cell
For the disordered BHJ solar cell we have use the high efficiency, low bandgap (Eg~1.6 eV) PTB7-Th donor and PC71BM acceptor in a normal architecture over the new cellulose substrate (Figure 23). The final device is constituted by CNF / transparent polymer / embedded Ag-grid / high conductivity hc-PEDOT:PSS / PEDOT:PSS / PTB7-Th:PC71BM / ZnO / Ca / Al. The silver grids were blade-coated with 130nm highly conductive hc-PEDOT by CSEM to support the extraction of the current towards the Ag grid. Cells were fabricated by blade coating the HTL (PEDOT:PSS 75nm), the active layer (PTB7-Th mixed with PC71BM 120nm), and the ETL (ZnO 20nm), followed by evaporation of Ca/Al through a shadow mask. Defined by the overlap of the hc-PEDOT coated grid and the top electrode, 8 individual cells of 21 mm2 were obtained.
Figure 23: Stack of the disordered BHJ test device used for testing the new cellulose substrate CNF (see attached doc)

These cells were characterized by measuring the EQE (External Quantum Efficiency) and the J-V curves, in dark and at ~0.7 sun illumination, through a mask with 2x2mm2 openings. The J-V curves show a significant leakage current which are most likely caused by microscopic shunts created by inhomogeneity in the coating of the CNF substrate. The best cell parameters measured: Voc=0.6V Jsc_eqe=9.6mA/cm2 FF=42%, resulting in power conversion efficiency (PCE) of 2.5% (Figure 24) (see attached doc).. For comparison, the reference device on a commercial ITO-coated PET substrate yielded a PCE of 5% (Voc=0.7V Jsc_eqe=12.2mA/cm2 FF=58%). The difference in Jsc cannot be explained by the lower transmission of the CNF/grid substrate compared to the PET/ITO substrate because the transmission is very comparable to that of the reference (Figure 22) (see attached doc).. The difference in FF is attributed to the micro-shunts caused by defects in the CNF substrate.
Figure 24. GreeNanoFilms disordered BHJ solar cell. a) Design device top view section and b) photo of device realized
(for scale: the grid width is 6 mm) (see attached doc).

Task 5.2: Nano-ordered bulk heterojunction solar cell on transparent CNF film
Realization of the nano-disordered BHJ solar cell
The objective was to realize of a nano-ordered polymer/fullerene bulk hetero-junction solar cell (Figure 25 and Figure 26) on transparent cellulose nano-fibre film, at the lab-scale (few mm² up to 5x5cm²), by using the self-assembly of hybrid block co-polymers, constituted by plant-sourced hydrophilic oligosaccharides and petroleum-sourced hydrophobic semiconductor polymers.

Figure 25. : Illustration of the ideal nano-ordered BHJ solar cell with vertical D/A phase separated domains, having 5-20 nm spacing for optimum efficiency. (see attached doc).

Figure 26: Cross-section of the lab scale device used for testing the CNF substrate with embedded grid. (see attached doc).

Due to the delay in synthesis of end-functionalized PCbz, only P3HT-b-AcMH block copolymer has been investigated in WP5. Because of the strong π-π interactions between P3HT chains, P3HT has a high natural tendency for crystallinity and in order to force P3HT-b-AcMH to self-order, CERMAV has been forced to use low molecular weight P3HT grades (Mn ~ 5 kDa). Unfortunately, P3HT grades typically used in photovoltaics have by far larger molecular weights (Mn ~30-90 kDa).
For the P3HT-b-AcMH block copolymer system, CERMAV has carried out extensive trials of solvent vapor annealing, by changing solvent composition (THF/H2O, anisole, CHCl3/MeOH), co-solvent ratio and annealing time, but all attempts failed to achieve nano-ordering. As an alternative, CERMAV has shown that the lamellar periodic structure appears after thermal annealing above a threshold temperature ~180°C and dramatically improves with further temperature increase: a 220°C thermal annealing under vacuum during 10 min has been retained as the standard procedure for P3HT-b-AcMH nano-ordering
For most organic semiconductors used in OPV devices, the photo-active layer must have a thickness around 200-250 nm to ensure the full absorption of incoming photons. To evaluate nano-ordering in the depth direction, CERMAV has realized thick P3HT-b-AcMH films (~10µm) onto PTFE substrates. Analysis shows that, at the air-film interface, P3HT-b-AcMH organize into perpendicularly oriented lamellar structure, while at the substrate/film interface and in the depth of the film, the domain orientation is locally random, without large scale preferential orientation (Figure 27) (see attached doc). This can be absolutely detrimental for the targeted OPV application because this results in poor charge extraction and can form isolating layers.
Figure 27: Lamellar phase orientation measured by TEM imaging on microtome sections of thick P3HT-b-AcMH films prepared on PTFE substrate: (a) preferential perpendicular orientation at the air/film interface, (b) and (c) random lamellar domain orientation in the depth of the film (Illustrations obtained from PhD Thesis of Y. Otsuka, 2017). (see attached doc).

The charge transport is the second essential parameter impacting solar cell efficiency: the low mobility of charges - especially holes - induces high carrier recombination loss. In order to evaluate the charge transport in the new P3HT-b-AcMH block copolymer material, CSEM has realized back-gated FET (field effect transistors) and compared them to reference transistors based on commercial P3HT. Overall better FET transistors were obtained on Si substrate with silanization, after annealing at 210°C for 10 min in N2 (Figure 28). The best carrier mobility was determined around 10-5 cm2 V-1 s-1, but this is still ~3 orders of magnitude less than reference P3HT based transistors. For comparison, best transistor mobility measured at CSEM in the same configuration for various commercially available grades of P3HT ranges between 10-3 and 10-1 cm2 V-1 s-1.

Figure 28: (a) Output characteristics of the best performing P3HT-b-AcMH transistor, (b) Transfer (see attached doc).

The initial plan to realize the nano-ordered active layer was to remove the oligosaccharide phase (AcMH) and to fill the obtained nano-porous/nano-structured template with small molecules of acceptor (PC61BM). Due to the delay of the synthesis of P3HT-b-AcMH and PCz-b-sugar block co-polymers, only the PS-b-MH system has been extensively studied by CERMAV and Lund University (Deliverable 2.6). Nevertheless, CERMAV has carried out some trials of Ar/O2 plasma etching on P3HT and MH films, but developments have been stopped because no significant selectivity has been found and because the plasma irradiation is damaging the electrical conductivity of P3HT. CERMAV also tested selective chemical etching of nano-ordered P3HT-b-AcMH films on Si substrates with concentrated sulfuric acid. Etching selectivity seems satisfactory, but BCP film is delaminated from the substrate during rinsing. Moreover, the bare CNF substrate is damaged by concentrated sulfuric acid, and acrylic planarization coating is only compatible with sulfuric acid having concentration < 40%.
New photoactive materials P3HT-b-AcMH and C60-b-AcMH were tested in OPV devices. The following experiments were performed:
• Effect of solvent, layer thickness and post-baking conditions
• Comparison to commercial P3HT, effect of annealing
• Variation annealing conditions
• Effect of active layer thickness on FF
• Standard versus inverted devices
• Inverted devices – evaluation of annealing conditions
• Test of P3HT-b-AcMH:C60-b-AcMH mixture in inverted device
• Test of P3HT-b-AcMH:C60-b-AcMH mixture with additional PCBM in inverted device

As a result the best solvent, baking condition, device geometry, layer thickness, and donor-acceptor ratio was identified for the P3HT-b-AcMH material. In comparison to blends of the commercial P3HT with PCBM, the device parameters Voc, Jsc, and FF are all lower, which indicates that despite the potential for nano-ordering of the neat material, in devices a morphology favorable for charge extraction was not obtained.

Nano-imprinting lithography of PEDOT:PSS
Before going to the final CNF film substrate, we attempted to NIL print PEDOT:PSS onto a less brittle cellulose substrate. For this purpose, CTP has coated tracing paper with the transparent acrylic coating and has printed series of plain square areas of PEDOT:PSS by flexography. CTP has characterized these samples by SEM cross-section imaging for estimating the layer thickness (ca. 1-2 µm for PEDOT:PSS) and by white-light interferometry for measuring the surface roughness (ca. 200-400 nm Ra for PEDOT:PSS) (Figure 29a) (see attached doc). Samples have been sent to VTT for roll-to-roll NIL printing. According to the specifications discussed with CERMAV concerning the DSA (directed self-assembly) Obducat company has supplied a mold with regularly spaced lines in the 100-nm range and having 1:1.7 aspect ratio (grating lines 90 nm width, 90 nm space, rectangular profile with ~155 nm depth). VTT has performed a series of experiments by varying the roll-to-roll NIL printing parameters: temperature, roll pressure and linear speed.
Unfortunately NIL results are not good: it seems that roughness of PEDOT:PSS prints and acrylic layer are very high and PEDOT:PSS prints is not uniform enough. Optical microscopy has been realized by VTT and show that the surface of the sample is very rough and no patterns can be seen (Figure 29b) (see attached doc).

Figure 29: PEDOT:PSS printed by flexography on acrylic-coated tracing paper : (a) SEM cross-sectional view, (b) roughness of PEDOT:PSS measured ca. Ra~200-400 nm by white-light interferometry (see attached doc).

1.4.5. WP6: Next-generation high-resolution nanolithography, high-sensitivity SERS biosensors
The WP6 is dedicated to the development of next generation nano-lithograhy for high sensitive biosensors using thin films made from BCP as template (WP2) for nanolithography and high sensitive biosensors.
Task 6.1: Nanopatterning design
During this initial period, we have developed a reliable technique of controlled deposition of polystyrene-block-maltoheptaose (PS-b-MH) block-copolymers (BCP) by optimizing the cleaning of Si substrates, investigation of different solvents and studying the effects of O2-plasma treatment of the Si surface on spin-coating process of PS-b-MH. We also started to develop the solvent vapor annealing (SVA) procedure to initiate a self-organization in the PS-b-MH films on Si surface.
Further, we established a standard method for producing 15x15 mm2 Si samples by cleaning of 2” Si wafers, spin-coating of anisole-based PS-b-MH and cleavage of the wafer to make many samples with uniform polymer film thickness. For accurate determination of polymer film thickness, spectroscopic ellipsometry measurement technique has been developed using a double-layer model (PS-b-MH on 1.5 nm thick native oxide SiO2 on Si) in the 400-900 nm wavelength range.
During the 2nd year period, we focused on experiments to study the directed self-assembly (DSA) in PS-b-MH BCP with a help of Si substrates patterned using high resolution electron beam lithography (EBL) and lift-off. High resolution EBL was used to make a variety of narrow (30-40 nm) lines with separation between 25 and 150 nm and arrays of dots made of 10 nm thick Au or Al films. During the EBL process optimisation it was found that the line edge roughness (LER) can be minimised by using Al instead of Au. After the EBL and lift-off patterning, the Si wafers were cleaved into 15x15 mm2 samples and 8.4 nm thick PS-b-MH polymer film was deposited by spin coating. Solvent annealing was performed using a mixture of THF: H2O with a weight ratio of 1:1 and 4:1, which results in formation of arrays of horizontal or vertical cylinders. AFM at CERMAV has been used to study the directed self-assembly of the PS-b-MH polymers as a function of size and shape of the constrictions defined by the metal lines and dots. For the first time, we found that the parallel lines with distances of 130, 100, 70, 50 and 25 nm result in well-defined hexagonal arrays of vertical cylinders in the PS-b-MH system with somewhat larger irregularities for 25 and 50 nm distances. It can be explained by stronger effects of LER compared to larger distances. Other types of constrictions (arrays of dots) did not result in any observable directed self-assembly under the identical annealing conditions. Annealing of the polymers at THF:H2O (1:1) ratio resulted in a mixture of horizontal and vertical cylinders at line distances of 30, 60 and 70 nm, while the line constrictions with larger distances lead to cylinders mostly with horizontal orientation.
In order to find the optimum SVA parameters for formation of well-ordered periodic PS-b-MH BCP, we studied different annealing times for both 1:1 and 4:1 THF:H2O ratios. The 1:1 ratio could be of interest due to formation of horizontal MH-cylinders as found by CERMAV, while the 4:1 ratio leads to vertical structures. The horizontal cylinders can be used in SERS and other applications. As an alternative to the THF:H2O annealing, we also made systematic studies of SVA in a mixture of Methanol:THF (MeOH:THF) within ratio range of 20:80 and 40:60 to form vertical MH-cylinders. The main conclusion from those experiments is the wide range of MeOH:THF, which results in well-defined vertical cylinders with good ordering without de-wetting of the BCP film as compared to the standard THF:H2O SVA.
Task 6.2: Removal process control
To address the task of Removal Process Control, we performed experiments with deposition of PS-b-MH BCP and separate polymers, PS and MH on the Si surface. Deposition of MH, which does not dissolve in anisole, unlike PS and PS-b-MH, was made using deionised water as a solvent. Concentration of 3% of MH in water gives polymer thickness of 30-50 nm after a double-step spin-coating process at 1500-3000 rpm. We obtained very good quality of MH films on Si - sufficient to be used in O2 plasma etching tests. The reactive ion etching (RIE) tests of MH, PS and PS-b-MH in oxygen and O2+CF4 plasma had a focus to find etching conditions with high selectivity to remove the MH-component from the BCP. Experiments to etch MH and PS separately indicated very high etch selectivity of 10-14 times (MH etches 10-14 times faster than PS in O2+CF4 plasma). However, it was difficult to measure the height variation in the PS-b-MH BCP film with vertical cylinders by AFM due to their extremely small lateral sizes (<5 nm), although we could obtain some profile data after RIE.
In our efforts to make stable etch masks and to keep well-ordered periodic structures of PS-b-MH, we made a detailed investigation of the SIS processes and Selective Area Atomic Layer Deposition (SA-ALD) using TMAl and water as precursors. Previously we made a breakthrough by demonstrating that the SIS process can be used successfully to visualize MH-cylinders in a conventional SEM. Unlike the SIS, the SA-ALD process is based on a short exposure of MH-component to TMAl, which as we found experimentally, leads to a much better separated AlOx dots as seen in SEM. However, there are also indications that the SA-ALD may lead to change of orientation of the cylinders, which limit the usefulness of this approach. In another set of experiments, we combined SIS and SE-ALD by treatment of a 14 nm-thick PS-b-MH film to long (SIS) and short exposures (SA-ALD), using 2 cycles for SIS and 13 to 23 cycles for SA-ALD. Our data show that the interval of cycles 2+18 and 2+23 gives best results, later we also demonstrated the advantage of longer pumping times in the infiltration process.
Some of the best SIS PS-b-MH samples with vertical cylinders were etched using a reactive ion etching process optimised for a high-resolution Si etch in order to use the AlOx as the etch mask and transfer the pattern from the BCP periodic structures into the substrate. Unfortunately, those experiments were not very successful, the alumina masks were not stable enough – EDX meaurements indicated that after 10 min of reactive ion etching, Al-signal disappears from the surface of Si. One possible explanation could be a limited thickness of the mask, the AFM measurement indicate a thickness of about 3 nm, which could be insufficient for a reliable etching process. Besides, the lateral size of the masks is 5-6 nm, which makes it very challenging to etch Si with such resolution.
Task 6.3: NIL stamp family process
In a further development of the DSA approach, we made guiding lines and dots using EBL-exposure of negative tone e-beam resist silsesquioxane (HSQ), which demonstrated very small line edge roughness (LER) <5 nm. The previously made lift-off based DSA patterns had typical LER of about 5-10 nm, which is comparable with size of the PS-b-MH BCP domains. Both the CERMAV and the Lund groups have demonstrated successful DSA of 15 nm-thick PS-b-MH films after a standard 4:1 THF:H2O SVA using the HSQ resist. As the characterisation methods, both AFM (CERMAV) and SEM (Lund) after SIS treatment (below) were successfully used. As an alternative method, Lund University together with Obducat Technologies AB have developed a technique to define the DSA pattern by nanoimprint lithography, which has a much higher throughput compared to EBL.
Task 6.4: SERS-based biosensors
Other critical issue was the BCP pattern transfer, which, according to the initial plan, should be used to fabricate sub-10 nm features in a hard substrate (Si). The original idea was based on the assumption that by using highly selective reactive etching process one can make sufficient (15-20 nm) step height in the PS-b-MH BCP. Due to the difficulties to characterize the height difference in the etched PS-b-MH BCP, we have applied and developed a novel approach for this polymer system, sequential infiltration synthesis (SIS), based on selective reaction of a metal precursor (e.g. tri-methyl aluminum – TMA) with one of the block (MH) to form Al oxide dots for the BCP with vertical cylinders. The SIS method allowed us to easily visualize the BCP domains by SEM and AlOx served as an etch mask in pattern transfer. Another developed approach called metal precursor loading, allowed us direct formation of metals dots (e.g. Au) in one domain from liquid solution by a selective chemical reaction, used later in SERS device fabrication.
The last period of activity had a strong focus on studies of various methods of dry etching for BCP pattern transfer, including reactive ion etching (RIE) and Ar+ ion beam etching (IBE) and different etch chemistries in RIE, including Cl- and F-based approaches. The motivation of this work was the need to develop a PS-b-MH BCP-based technology to make NIL stamps for bio-sensors and to demonstrate nanoimprint replication. The SIS-treated ordered PS-b-MH BCPs were used as masks in etching of Si, using both vertical cylinders and lamella structures. We also addressed the issue of mobility of infiltrated masks of AlOx after removal of polymers by an oxygen plasma etching, ozone cleaning or rapid thermal annealing. One of the conclusions was observed limited stability of masks formed from vertical cylinders as opposed to vertical lamella structures, another conclusion was about good etch performance of SF6/C4F8 etch chemistry in a RIE process with or without Cl2/Ar pre-etch step to etch Si with a SIS-BCP mask. Still a reliable RIE process with BCP as an etch mask remains an issue, as sub-10 nm features are very easy to move during the etch process, that results in at least partial loss of BCP ordering. Thin, 3-4 nm thick AlOx layers formed after the SIS process, may not always be sufficiently stable in a typical RIE process, which requires etch process adjustment (pressure, ion energy, chemistry). However, we managed to demonstrate a pattern transfer from the PS-b-MH BCP in a Si substrate, the Si substrate was used later as a stamp in a thermal nanoimprint process.
As part of our activity towards the realization of the SERS-device, we used test PS-b-P4VP samples treated by Au precursor loading process to make arrays of Au particles for SERS measurements at INL, Portugal. The SERS bio-sensor samples were not produced by a nanoimprint technique as planned earlier, but the ordered BCP and Au-precursor loading provide an alternative way of a large-scale nanofabrication to nanoimprint. The SERS measurements at INL showed an improved Raman signal.

1.4.6. WP7: Demonstration
WP7 aims to demonstrate the industrially relevant of Technical CNF films for :
• electronic application developed in WP3 and WP4 using R2R fabrication
• OPV device using films made from BCP deposited on CNF (WP2, WP4, WP5) using inkjet
• NIL stamp based on BCP (WP2, WP6)
Task 7.1: Technical films for printed electronics
The target of the work was to demonstrate the use of technical CNF films for electronic application. The CNF films were prepared by using mechanically disintegrated cellulose nanofibrils (CNF) in pilot scale.
Fully-printed top-gate-bottom-contact organic Thin Film Transistors (TFT) were fabricated on Cellulose NanoFibril (CNF) substrate using commercial electronics inks (Figure 30) (see attached doc).. Gravure printing was first used to coat the CNF film with a polymer resist to smooth and close the surface. InkJet (IJ) and gravure printing was then used to fabricate the conducting, dielectric and semiconducting device layers. Transistor performance was characterized for mobility, threshold voltage, on/off ratio, and bias stress stability.
Figure 30: (left) Top-gate-bottom-contact TFT device structure. (right) Optical image of the fabricated TFTs, where the continuous vertical lines are the top-gate conductors while the rectangular vertical electrodes are the source and drain (see attached doc)

At the second phase, great enhancement in electrical behavior was achieved by a simple chemical treatment step on the metal electrodes. The chemical treatment was done by drop casting the M001® (Merck) solution on the patterned substrate. The transistors on cellulose nanofibril substrate with printed silver electrodes show performance close to that of transistors with gold electrodes on plastic substrate. These results are compared with other electrode and substrate materials (Table 3) (see attached doc).
Table 3. Comparison of transistors with the same printed active materials but with different electrode processing methods
and different substrates. Channel length L is stated next to the electrodes (see attached doc)

Task 7.2: Flexible opto-electronic device
Following the successful validation of the CNF film with embedded silver grid in WP5, a demonstrator in the form of an OPV module was fabricated using commercial materials with the CNF film as substrate. A functional demo was obtained (Figure 31) (see attached doc). This shows the potential for possible application of the substrate for OPV. During the project the quality of the substrate was improved and the number of defects reduced significantly. Still, further improvements in the quality of the printed Ag grid are required before the substrate is ready for large area application for OPV. Especially the sensitivity to mechanical stress makes the film hard to handle and up-scalability in R2R coating needs still to be investigated.

Figure 31: Demonstrator of the CNF substrate with commercial photoactive material (see attached doc)

Task 7.3: Large-scale high resolution nanolithography for fabrication of biosensors
The initial technological approach in Task 7.3 of WP7 was to use PS-b-MH BCPs as templates for making nanoimprint stamps for large-scale patterning of a hard surface to fabricate SERS biosensors. Thus this approach is closely related to Task 6.3 of WP6 and Deliverable D6.4: Development of NIL stamp family and includes capability to make stamps from the ordered BCP structures. The key process step to realize those Tasks 6.3 and 7.3 and Deliverable D6.4 is to find and demonstrate a suitable pattern transfer method to “replicate” well-defined structures of vertical cylinders in the PS-b-MH BCP into a hard substrate, which is required for NIL stamps.
Thus, the last period of activity had a strong focus on studies of various methods of dry etching for BCP pattern transfer, including reactive ion etching (RIE) and Ar+ ion beam etching (IBE) and different etch chemistries in RIE, including Cl- and F-based approaches (WP6). As mentioned earlier, the motivation of this work was the need to develop a PS-b-MH BCP-based technology to make NIL stamps for bio-sensors and to demonstrate nanoimprint replication (Deliverable 7.3). The SIS-treated ordered PS-b-MH BCPs were used as masks in etching of Si, using both vertical cylinders and lamella structures. We also addressed the issue of mobility of infiltrated masks of AlOx after removal of polymers by an oxygen plasma etching, ozone cleaning or rapid thermal annealing. One of conclusions was limited stability of masks formed from vertical cylinders as opposed to vertical lamella structures, another conclusion was about good etch performance of SF6/C4F8 etch chemistry in a RIE process with or without Cl2/Ar pre-etch step to etch Si with a SIS-BCP mask. Still a reliable RIE process with BCP as an etch mask remains an issue, as sub-10 nm features are very easy to move during the etch process, that results in at least partial loss of BCP ordering. Thin, 3-4 nm thick AlOx layers formed after the SIS process, may not always be sufficiently stable in a typical RIE process, which requires etch process adjustment (pressure, ion energy, chemistry). However, we managed to demonstrate a pattern transfer from the PS-b-MH BCP in a Si substrate, the Si substrate was used later as a stamp for thermal nanoimprint (Deliverable D6.4).
As part of our activity towards the Deliverables D6.5 and D7.3 we used test PS-b-P4VP samples treated by Au precursor loading process to make arrays of Au particles for SERS measurements at INL, Portugal (Figure 32) (see attached doc).

Figure 32: Schematic of metal precursor loading process (left) and SEM image of deposited Au dots using PS (see attached doc)

The SERS bio-sensor samples were not produced by a nanoimprint technique as planned earlier, but the ordered BCP and Au-precursor loading provide an alternative way of a large-scale nanofabrication to nanoimprint.

1.4.7. WP8: Sustainability, LCA & risk assessment
The work package 8 Sustainability, Life Cycle Assessment (LCA) and Risk Assessment had the objective to investigate the environmental impacts of the selected nanomaterials and the associated (prospective) applications.
The project GreeNanoFilms has selected and investigated the following specific nanomaterials and possible associated products in case studies (CS). First focus was “cradle-to-gate” LCA of three nanomaterials. Cradle-to-gate is an assessment of a partial product life cycle from the preproduction of raw materials to manufacture of the nanomaterials to the factory gate:
• Case study 1: Glycopolymers (PS-b-MH)
• Case study 2: Cellulose nanofibrils (CNF)
• Case study 3: Cellulose nanocrystals (CNC)
Second focus was the LCA of the production of three selected applications:
• Case study 4: Nanocellulose Technical Films
• Case study 5: Organic Photovoltaic
• Case study 6: Biosensors
Based on a comprehensive literature search and a questionnaire sent to all manufacturing partners, specific process models for the application, use and end-of-life phases (recycling, treatment and disposal) have been developed in a first step.
This was the basis for the next steps of the LCA: the Life Cycle Inventory (LCI) and Life Cycle Impact Assessment (LCIA), and interpretation. The Life Cycle Inventory includes at data collection and calculation procedures in order to quantify from “cradle-to-gate” (with focus on production) all relevant inputs (e.g. material inputs) and outputs (e.g. emissions to air) of the product system. The next step was the Life Cycle Impact Assessment and the interpretation of the results. The life cycle impact assessment (LCIA) based on the recommended ILCD impact categories and on ReCiPe midpoints combined with shadow prices. The modeling, calculation, visualization, evaluation and analysis according to applicable environmental impacts are carried out by means of the LCA software Umberto and the Ecoinvent database.
The variety of investigated nanomaterials and applications are very different and so also the results of the LCA case studies. In some case studies, the environmental impacts from nanomaterials are low, in other case studies, however, significant. The results of the life cycle impact assessment and the calculated shadow prices for the production / synthesis of the selected nanomaterials are shown in the table 4. (see attached doc)
Table 4. Summarized LCA data with shadow prices of the nanomaterials CS1-CS3 (functional unit: 1 kg) (see attached doc)

The lowest environmental impacts and also the lowest shadow prices have the two types of CNF in comparison to the investigated synthesis of CNC and Glycopolymer PS-b-MH.
The complexity of the application case studies was very high.
In case study 4 different technical nanocellulose films are investigated. The production of the CNF film is only dependent on the energy usage and the impact of the preproduction of the bleached sulfate pulp. The influence of the chromatogeny technique to create a superhydrophobic CNF film of all environmental impact categories in comparison to the CNF film production is almost negligible. This process is very efficient and sustainable. On the other side the following ALD process to produce an Al2O3 and SiO2 layer has high impact in all categories with a range factor of 2.3-2.4 in comparison to the other variants. The highest shadow price was in the scenario superhydrophobic ALD CNF film with 0.24€/m² film. The summarized shadow price of the other scenarios variant was on the level of 0.10-0.11€/m² film, and thus substantially lower. The result of these LCA can be used also for internal process optimization.
The application of the case study 5 considers prospective organic photovoltaics (OPV) that belong to the 3rd generation of solar cells. The goal of this application is to produce electricity from the sun radiation by harvesting light molecules. The solar cell generally consists of different layers, which are linked together to enable the exciton flow. Compared to conventional OPV, the bio-based OPV that is developed within this project differs in layer arrangement and in used materials as well as applied processes.
In GreeNanoFilms following different OPV solar cells are investigated:
- Conventional BHJ-OPV solar cell with ITO
- New 1. Level: conventional BHJ-OPV solar cell with CNF / Ag-grid substrate
- New 2. Level: Nano-structured BHJ solar cell
Unfortunately, due to the delay of the synthesis of P3HT-b-AcMH and PCz-b-sugar block polymers, only the PS-b-MH has been extensively studied by CERMAV and Lund University.
For this reason we had to make a simplification for the modeling. Now PS-b-MH was used as BCP instead of possible P3HT-b-AcMH or PCz-b-sugar block polymers. Same efficiencies for parameters (product lifespan, power conversion efficiency, same end-of-life procedure) have been adopted for the comparison of the conventional OPV with the investigated prospective OPV level I and II. So, the focus was in the comparison of the prospective production of OPV. Nevertheless realistic models could be created and realistic calculations are performed.
The results of the life cycle impact assessment and the calculated shadow prices for application case studies 5 are shown in table 5. (see attached doc)
Table 5. Summarized LCA data with shadow prices of the CS5 Prospective OPV (functional unit: 1 m²) (see attached doc)

If the entire OPV production is considered, the influence of the substrate is very low. This effect is illustrated by the comparison of OPV conventional with GNF OPV level I. The difference of the summarized shadow prices is only 0.38€/m² OPV by total prices of 35.79€/m² OPV by the highest scenario OPV conventional.
On the other side the results has illustrated that the OPV substrate biobased CNF is better as the OPV substrate conventional in most environmental impact categories. The highest summarized shadow price has also the OPV substrate conventional with 1.15€/m² substrate in comparison to 0.77€/m² substrate for the OPV substrate biobased CNF.
The prospective change of the structure and architecture in the scenario GNF OPV level II has a much larger potential for optimization of the OPV production. For the GNF OPV level II it is possible for many environmental impact categories of improvements to 2/3 in comparison to the OPV conventional. Further research in new architectures of OPV seems very useful from an environmental point of view.
The LCA of prospective Biosensor in case study 6 was the first LCA in this detail. The objective of GNF was to improve the sensor chip by having a higher detecting resolution, which is related to the pattern of the sensor chip. Nevertheless realistic models could be created and realistic calculations are performed. The results of the life cycle impact assessment and the calculated shadow prices for application case study are shown in table 6. (see attached doc)
Table 6. Summarized LCA data with shadow prices of the CS6 Prospective Biosensors (functional unit: 1000 SERS-Chips)
(see attached doc)

All environmental impact categories are dependent on the metallization process and the spin-coating process. The influence of the other processes is low. Through the high influence of the impact categories particle matter formulation (spin-coating and metallization) the summarized shadow price is 9.49€/1000 SERS-Chips. Also the categories climate change and water depletion are significant. The present results also give approaches for further optimizations for the different processes from an environmental standpoint.

Potential Impact:
1.5. Potential impact
One of the potential impacts is to “evaluate the up-scalability and cost structure of the opto-electronic device manufacturing process. CSEM has used its experience with polymer solar cell production to determine the cost structure of a solar cell made on CNF film with a nanostructured surface. The cost can be measured in €/Wp, and the special advantages of easy integration with polymer solar cells on CNF film has also been considered. The advantages of a cellulose-based production and the environmental advantages, such as highly decreased energy payback time is highlighted. For the purpose of this report, the layout shown in Figure 33 (see attached doc) is used a starting point of the discussion. This is representative the status of the OPV demonstrated in the GreeNanoFilms project.

Figure 33 GreeNanofilm OPV, experimental, technological carrier device cross section. (see attached doc)

The project team realized lab scale proof of concept devices based on the cross section in Figure 33 (see attached doc). The process is described in D5.2.

Figure 34 GreeNanoFilms OPV. a) Design device top view section and b) photo of device realized (for scale: the grid width is 6 mm) (see attached doc).

The overall power conversion efficiency of reference devices fabricated (2.5% for the best device) is lower compared to OPV processed on conventional substrates (typically 5%) using similar classes of active materials, but with further improvements in the quality of the silver grid, a comparable efficiency may be expected.
The working assumption for an industrial upscaling scenario involves roll to roll atmospheric (controlled environment) coating and drying. Alternatively, roll-to-roll ink-jet printing could be deployed.
The experimental work realized so far indicates that solution processed OPV cells realization is possible. To gain more insights, a demonstration of fully printed especially for inverted OPV structures and encapsulation would be highly valuable. At this stage, the core process steps for production are deemed feasible, but it would be necessary to continue the development to ascertain scalability to industrial scale OPV processes. It would be advisable to address at least the following points:
o Ability to produce rolls with > 30cm web; 120 cm webs and larger could be required in high volume.
o Length: several hundred meter rolls for pilot runs
o Solvent sensitivity, brittleness (NB: strict control of hygroscopic, mechanical process conditions)
o Uneven and stiff planarization layer with acrylic resin complicates processes like NIL structuring
o Slot die whole process in sheet to sheet – needs to be tested
o NB: some aspects of process (e.g. handling) would be easier to deploy in R2R
o NIL steps: tried in discrete / sheet – theoretically possible in process but there are substrate limitations
o Grid: should be provided to all substrate – possible issue with alignment and registration of process in R2R coatings – unless using “live” IJP process system for whole stack. Ultimately, this point will lead to OPV design, dead areas and production yield
o Encapsulation: more systematic tests are required to assess scalabilty. OPV encapsulation process also include:
• UV curing in near UV: should be possible
• Alternatively: Thermal curing also could be an option
o Singulation: from roll to single devices, different cutting techniques need to be tested (including impact on crack formation, given potential brittleness)
o Contacts
In conclusion, the current status indicates that there is potential for upscalable OPV production processes, but a significant amount of work is still needed to assess the industrial scalability of the concept in practice.

Cost structure
In absence of a validated, upscalable industrial process, a quantitative cost structure is highly speculative. Given the relatively early stages of GreeNanoFilms technology developments it is at least relevant to give some qualitative indications.
The cost structure of an OPV is deeply linked to several factors. These include the device structure, the volumes or materials used, the production process yield, the cost structure and business models of the value chain partners which plays a key role. For this section, a standard OPV stack is assumed, which is largely similar to the one illustrated in Figure 33 . It is further assumed that the stacks would be produced under similar conditions yielding similar performances. The sources of cost estimates for different layers are taken from exploitation plans of the Sunflower EU project (), thought for project confidentiality reasons specific details cannot be disclosed. Cost estimates for the CNF substrates are provided by the GreeNanofilm consortium.
For large production volumes, processing costs and amortization of equipment, while not negligible, play a marginal role in the cost structure. The latter is largely dominated by materials, especially the active materials (absorbers and conductors) and the barrier layer. One of the current situations is that active layers are currently premium specialty chemicals materials and marketed as such, while base polymer films are usually commoditized materials, or expected to be such. As the field matures, it is expected that the overall costs of OPV per square marketed will decrease and the relative portion of the base polymer may increase, though not its cost per square meter.
An OPV substrate typically comprises of: a base film, a Barrier layer and a transparent conducting layer. The latter is integrated in the form of an embedded silver grid plus an homogenously coated hole conductor (PEDOT:PSS) in the GreeNanoFilms project.
Details on the productions estimated productions costs of cellulose-base films for OPV mass production are difficult to give at this stage, nevertheless, very significant cost reduction per square meter can be exacted with respect to the current standard PET (up to a factor ten, indicatively). In terms of production costs of OPV, the base polymer film is expected to play a relatively minor, but still significant, role overall. The cost contribution per square meter produced is estimated in the order of a few percent’s of the overall bill of materials and less than a tenth. This may change slightly in the mid-long term, once the active materials reach higher volumes of production and the value chain becomes more varied and established.
In terms of barrier deposition, it is currently assumed that the cellulose-based foils would not offer a (process) cost advantage with respect to the reference ( PET) material, hence no significant cost saving is expected. Furthermore, it is not clear how easy it would be to adapt barrier production processes to cellulose based films, so there is a risk. It is to be noted that the necessity of barrier layers is linked to production, application case and to business model. Namely, if the applications considered require very limited operational and shelf lifetimes are extremely limited (in the order of weeks to months and not more than 1 year in ambient conditions) barrier-free substrates may be sufficient and could be deployed. Alternatively, if OPV were to be laminated in a secondary encapsulation (e.g. glass/glass) system, the costs of barrier layer may be saved.
In terms of the conducting layers, the cost advantage arises from the use of “silver grid plus PEDOT/HTL” concept instead of ITO; but again similar concepts are being developed for more conventional PET–based foils.
In summary, the base polymer film substrate plays a relatively minor part of the expected bill of materials for OPV per square meter. If sufficient production yield for barrier process and for OPV production can be achieved, adoption of cellulose film could contribute to lowering the costs of comparable (ITO-free) OPV, but these would not be a major contribution (perhaps leading to a reduction in the order of 10% or less). In case of comparison with ITO-based OPV, the cost reduction could be more substantial.
Environmental impact
An extensive life-cycle assessment of nanomaterials and OPV devices based on cellulose-derived substrates and materials is provided in Deliverable D8.2. (Steinfeldt, 2016). (see Table 7) below:

Table 7: Summary LCA data from “Shadow prices” summarize and quantify the equivalent environmental impact
from reference (Steinfeldt, 2016). (see attached doc)
As seen in Table 7, the cellulose-based substrate has a shadow price of 0.77 Eur, versus a shadow price of 1.15 Eur for a conventional OPV substrate. It is however important to notice that ITO-free variants of OPV substrates are also in the research and development pipeline, at similar readiness level as those developed in the GreeNanoFilms project.

Figure 35 Environmental footprint kg CO2-Eq/m2 OPV devices ; figure 53 from reference (Steinfeldt, 2016). (see attached doc)

In practical production implementation conditions, it should be noted that a different OPV stack and technology would be deployed, namely an all printed inverted structure is more likely to be production compatible than one with evaporated electrodes. In terms of comparison, it is also important to notice that there are all printed ITO alternatives with similar (or more advanced) technology readiness level as the silver grid demonstrated in the GreeNanoFilms project. Hence the comparison used in (Steinfeldt, 2016) is partial and one should also consider the case of ITO-free, PET based substrates.
When the PET base film is replaced by cellulose, all else being equal or comparable, one expects a limited environmental impact change in terms of CO2 equivalent since PET contributes a small fraction (percents) to the overall footprint. As caveat, we note that detail investigation is warranted, which would require precise process input and is premature at this stage given no stable R2R process is in place).

What is clear from Figure 35 is that a larger impact expected from e.g. the silver layer. Further, one would need to assess the impact of barrier layers produced on cellulose nano films.

Concerning the energy pay-back time, the lab results indicate that a similar efficiency performance may be obtained from standard OPV stacks processed on CNF substrates, which is a very nice technical result. Since the base-film (PET as reference) is not a major contributor to the overall energy footprint of OPV device productions, there is no expectation of major reduction of the energy-pay-back time with respect to comparable OPV realized on PET.
Conclusion and recommendations
The report assesses the status and expected impact of the adoption of cellulose films for the processing and production of OPV. Three aspects were considered: up-scalability, cost structure and environmental impact.
• Up-scalability: process development is in the early stages TRL 2-3. The first key steps indicated a general feasibility of solution-processing of OPV. One would need to proceed with larger scale trials, including deposition of barrier films on CNF substrates.
• Cost Structure: a positive, but not major impact, of cellulose-base substrates could be expected when comparing with equivalent, ITO-free, PET-based substrates
• Environmental impact: some positive impact expected from several eco-indicators. Concerning the energy pay-back time, the lab results indicate that a similar efficiency performance may be obtained from standard OPV stacks processed on CNF substrates. Since the base-film (PET as reference) is not a major contributor to the overall energy footprint of OPV device productions, there is no expectation of major reduction of the energy-pay back time with respect to comparable OPV realized on PET. A degradation of the power conversion efficiency with respect to the reference would lead to an increase of the energy pay-back time.

2.1 Project objectives for the period
WP9: WorkPackage 9 was devoted to the communication, dissemination and exploitation activities. During the 36 months, it has been done through the:
o Development of tools and a dedicated toolbox: Project website, brochure and logo.
o Communication of the different GreeNanoFilms Newsletters.
o Dissemination activities to the scientific community and end customers.
o Organisation of two Workshops.
o Presentation of GreeNanoFilms projects in different international conferences.
o Definition of the GreeNanoFilms applications possibilities for identified markets.

2.2 Work progress and achievements during the period
Summary of progress towards objectives and details for each task
The consortium agreement has been signed before the start of the project in order to regulate the critical aspects of governance and the management of intellectual property and access rights to results.
Task 9.1: Project dissemination kit
CTP and CERMAV, with the help of all the partners, have developed tools and environment for efficient communications and disseminations of the project. They have been updated all along the project. Among these:
A Toolbox including:
o A logo, a graphical chart and six templates (poster, reports, powerpoint presentation...) for an impactful communication,
o Goodies, like mugs of the GreeNanoFilms project, a brochure, a poster are available to present the project,
o Video film: The first part presents the partners and the objectives of this project. The second part is highlighting the main results made within these 3 years. The second part of the video has been delivered in January 2017.
A public website ( updated with Newsletters and events and an extranet have been set up. The extranet site is shared with the partners and allows easy documents storage and exchange.
Communication and dissemination actions:
o 4 Newsletters have been made and posted in the public website. A “Newsflash” has also been created in 2015 for New Year greetings and some updates of the project,
o Two Workshops have been organized during the project: The first one in VTT, Finland on June 9th, 2016 (D9.7) and the second one in CERMAV, France on January 26th, 2017 (D9.10).
o Publications:
• TNO: 3 paper
• VTT: 2 papers
• CERMAV: 1 paper
• Ongoing: 1 paper from University of Lund & Cermav and 2 papers from VTT

The table summarizing the different communications, which have been done all along the 3 years is illustrated in 4.2 section

Task 9.2: Exploitation
GreeNanoFilms, with the help of the European Commission, organized an ESS (Exploitation Strategy Seminar) meeting on December 17th, 2015 during Bremen meeting. Two reports has been made by Dario Mazzella: ESS report and a Plan for Using and Disseminating Foreground.
In parallel, OBDUCAT and the partners have worked on the exploitation plan of the project results (D9.6). This report defines the main potentialities of GreeNanoFilms results in terms of applications on identified markets, and the plan for their use and dissemination.
Table 9: Repartition of accepted communication versus type (see attached doc)

Table 10: Repartition of achieved communications versus partners (see attached doc)

List of Websites: