General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

Users may download and print one copy of any publication from the public portal for the purpose of private study or research.

You may not further distribute the material or use it for any profit-making activity or commercial gain

You may freely distribute the URL identifying the publication in the public portal If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from orbit.dtu.dk on: Aug 22, 2020

Scalable fabrication of organic solar cells based on non-fullerene acceptors

Gertsen, Anders Skovbo; Fernández Castro, Marcial; Søndergaard, Roar R.; Andreasen, Jens Wenzel

Published in:Flexible and Printed Electronics

Link to article, DOI:10.1088/2058-8585/ab5f57

Publication date:2020

Document VersionPublisher's PDF, also known as Version of record

Link back to DTU Orbit

Citation (APA):Gertsen, A. S., Fernández Castro, M., Søndergaard, R. R., & Andreasen, J. W. (2020). Scalable fabrication oforganic solar cells based on non-fullerene acceptors. Flexible and Printed Electronics, 5, [014004].https://doi.org/10.1088/2058-8585/ab5f57

Flexible and Printed Electronics     

PAPER • OPEN ACCESS

Scalable fabrication of organic solar cells based on non-fullereneacceptorsTo cite this article: Anders S Gertsen et al 2020 Flex. Print. Electron. 5 014004

 

View the article online for updates and enhancements.

This content was downloaded from IP address 192.38.90.17 on 15/01/2020 at 06:58

Flex. Print. Electron. 5 (2020) 014004 https://doi.org/10.1088/2058-8585/ab5f57

PAPER

Scalable fabrication of organic solar cells based on non-fullereneacceptors

Anders SGertsen1 ,Marcial FernándezCastro1 , RoarR Søndergaard and JensWAndreasenDepartment of Energy Conversion and Storage, Technical University ofDenmark, Frederiksborgvej 399, DK-4000Roskilde, Denmark1 These authors contributed equally to the work.

E-mail: jewa@dtu.dk

Keywords: organic photovoltaics, scalable fabrication, printing and coating, non-fullerene acceptors, polymer solar cells

AbstractOrganic solar cells have recently experienced a substantial leap inpower conversion efficiency, in partdrivenby formulationswithnewnon-fullerene acceptors. This has brought the technologypast thepsychologically importantmarkof 15%efficiency for unscaled laboratory devices, and the results arestimulating another burst of research activity.Whether thiswill propel the technology into a viablecommercial contender has yet to bedetermined, but to realize thepotential of organic solar cells for utilityscale application, fabricationusing scalable processing techniqueshas to bedemonstrated—otherwise, thepassingof the 15%markwill eventually leavenomore lasting impact thanwhat the passingof the 10%markdid.Thus, addressing the scaling lag between the 15%cell efficiencies of lab-scale devices on rigidglass substrates fabricatedusingnon-scalable techniques and the 7%efficiencies of scalably fabricateddevices onflexible substrates is key.Here,wediscuss the concept of scalability and give an account of theliterature onnon-fullerene acceptordevices fabricatedwith scalablemethods andmaterials.On thebasis ofthis,we identify three crucial focuspoints for overcoming the lab-to-fab challenge: (i)dual temperaturecontrol, i.e.simultaneous control of the ink and substrate temperatures duringdeposition, (ii) systematicin situmorphology studies of active layer inkswithnew, green solvent formulations during continuousdeposition, and (iii)development of protocols for continuous solutionprocessing of smooth, transparentinterfacial layerswith efficient charge transfer to the active layer.Combining these efforts and in generalaccompanying such studieswith stability analyses and fabricationof large-area, scalably processeddevicesare believed to accelerate the relevanceof organic solar cells for large-scale energy supply.

1. Broader context

Climate change is arguably one of the biggest chal-lenges currently faced by human kind. Honouring theParis Agreement and thus keeping the average globaltemperature rise in this century below 2 °C relative topre-industrial levels demands an ambitious effort toreplace fossil fuels with sustainable energy sources inour electricity production. Silicon solar cells areexperiencing a rapid increase in worldwide installedcapacity, but also new generations of solar celltechnologies have the potential to reach maturity as asustainable technology in the near future and thus toaid this transition. The key to the sustainability interms of energy and materials use of these emergingtechnologies is scalability. Although silicon solar celltechnologies have proven that upscaling fabrication also

leads to significant cost reductions, their fabricationremains very energy consuming. Organic solar cellscould prove to be a viable alternative with projectedenergy payback times of only fractions of those of siliconmodules. Already now, organic solar cells are used forniche applications owing to their semi-transparency,flexibility, low weight, and possibilities of customdesigns in terms of colors and shapes. In addition, utilityscale competitiveness of organic solar cells with maturethin-film technologies is edging closer in current yearswith researchers pushing laboratory cell efficienciesbeyond 15%using novel non-fullerene acceptormateri-als and several companies continuously improvinglarge-scale fabrication; bridging these efforts and thusaddressing the lab-to-fab challenge remains the mostsignificant hurdle for the sustainable scalability oforganic solar cells.

OPEN ACCESS

RECEIVED

3October 2019

ACCEPTED FOR PUBLICATION

31October 2019

PUBLISHED

9 January 2020

Original content from thisworkmay be used underthe terms of the CreativeCommonsAttribution 3.0licence.

Any further distribution ofthis workmustmaintainattribution to theauthor(s) and the title ofthework, journal citationandDOI.

© 2020TheAuthor(s). Published by IOPPublishing Ltd

2. Introduction

Organic photovoltaics (OPVs) are often cited as one ofthemost promising third generation solar cell technol-ogies because of their compatibility with solutionprocessed roll-to-roll fabrication, enabling a fast andcontinuous fabrication [1–3]. Whereas the projectedenergy payback times of roll-to-roll fabricated OPVsare as short as weeks [4], at least an order ofmagnitudebetter than those of silicon technologies [5], the large-scale, grid-connected installations of OPVs continueto be halted by economical inferiority. In order toovercome this, improvements of especially stabilities,but also efficiencies, of flexible OPV modules areneeded [6, 7]. However, properties that are beneficialfor building integration such as low weight and partialtransparency have given OPVs an advantage, and anumber of companies are focusing on these alternativeapplications.

With the surpassing of the psychologically impor-tant 15% power conversion efficiency (PCE)mark forsingle junction cells earlier this year [8], it is clear thatthe field of OPVs is experiencing a revitalization whichcanmainly be attributed to the emergence of non-full-erene acceptor materials [9, 10]. The advantages ofthese over fullerene-based acceptor materials do notonly comprise higher efficiencies, but also improvedoptical absorption and tunability as well as superiorcell stabilities [11–16]. Combined with the recentlyreported low dependence of the PCE on active layerthicknesses and areas for high-efficiency systems[8, 17, 18], many prerequisites for the upscaling oforganic solar cells are nearly fulfilled.

In accordance with previous endeavors related tofullerene-based OPVs [1, 19–23], we believe that it isparamount for the continued growth of the researchfield as well as a sustainable scaleup of the technologythat the current focus on high PCEs and materialdevelopment within fullerene-free OPVs is accom-panied by:

(a) the use of roll-to-roll compatible or other scalabledepositionmethods in addition to spin coating,

(b) efficiencies of larger cells (�1 cm2)or evenmodulesalongside the small scale champion devices whichare often only on the order of mm2, and

(c) stability analyses.

If these parameters were to be consistently reported, itwould enable a more concentrated effort towardsaddressing the lab-to-fab challenges (visualized infigure 1) and meeting the 10-10 targets for flexibleorganic solar cell modules of 10% efficiency and 10years stability [1]. Held up against extensive econom-ical analyses based on flexible OPV modules with 7%PCE and stabilities of 5–10 years that predict superiorlevelized costs of electricity compared to mature solar

technologies such as e.g. crystalline silicon [7, 24], the10-10 targets almost seem like a conservative estimatefor sustainable scalability of organic solar cells. Thisfurther motivates overcoming the lab-to-fab chal-lenges: in recent years, efficiencies well above 10%have consistently been reported for fullerene-free,spin-coated, small-area laboratory devices on glasssubstrates [17, 25–30], and 10% has also been reachedusing partly or fully scalable active layer depositiontechniques (see section 3 for a discussion of scalability)[31, 32]. Even flexible devices with scalably depositedfullerene-free active layers are exhibiting efficienciesabove 7% [33, 34], which, coupled with recent reportsof 10 year lifetimes (extrapolated from 200 h stabili-ties) in fullerene-free, laboratory-scale devices by Du,Brabec et al [16], indicates that the 10-10 targets arewithin immediate reach.

The current limitations in upscaling of OPVs arecomplex and involve a series of challenges, includingmaterials’ compatibility, choice of non-toxic solvents,choice of compatible interface materials and, mostimportantly, stability and costs. Evaluating and seek-ing to overcome these limitations in devices processedusing non-scalable deposition techniques such asspin-coating is inherently problematic, and theyshould instead be evaluated in the framework of fullyscalably deposited OPVs. In this perspective, wereview state-of-the-art fullerene-free, single junctionOPV devices and the extent to which scalable techni-ques and materials are used in the fabrication. Aidedby a discussion of the terminology related to the con-cept of scalability as it is used in connection with fabri-cation of organic solar cells, we aim to assess thepotential of organic solar cells for sustainable scal-ability, to evaluate whichmaterial systems are themostpromising for upscaling, and to suggest focus pointsfor overcoming the lab-to-fab challenges currentlyfaced. We note that a similar deserved attention isgiven to the closely related field of upscaling of per-ovskite solar cells in a recent publication, underliningthe relevance of this challenge for third generationsolar cells in general [35].

3. Scalability: a note on semantics

The word ‘scalable’ can generally be interpreted as thecapability of a process to handle a larger workloadwithout significantly compromising functionality andcost. However, formulating a clear cut definition ofscalability in the context of organic solar cell fabrica-tion is at best a very difficult task and not the purposeof this section. Instead, we seek a discussion andeventually a community-wide consensus on a termi-nology, allowing for a higher degree of transparency inthe reporting of OPV devices. As a first iteration, andfor categorizing fabrication processes reviewed in the

2

Flex. Print. Electron. 5 (2020) 014004 A SGertsen et al

present article, we suggest the following three classifi-cations (illustrated infigure 2).

(a) Fully scalable: high throughput deposition tech-niques that are directly compatible with contin-uous roll-to-roll setups and are linked to nomaterial waste.

(b) Partly scalable: deposition techniques that can bemade compatible with continuous roll-to-rollsetups with somemodifications and/or are linkedto somematerial waste.

(c) Non-scalable: low throughput deposition techni-ques that are incompatible with continuous roll-to-roll setups and/or are linked to a high materialwaste.

These can be used to classify deposition techniques ofboth active layers and electrodes as well as interfacial

layers such as hole- and electron-transport layers. Thearguments for the placement of specific depositiontechniques in these categories will be given in section 4alongside descriptions of these.

Defining scalability of the actual active layers basedon non-fullerene acceptors is yet more complex. Amaterial can have properties that allow for the use ofscalable deposition methods without being scalableitself, simply because the material synthesis or manu-facture is too elaborate and thus too expensive to usein an upscaling process. In this context, we would liketo highlight the recent work by Li et al on an industrialfigure of merit for the cost potential of fullerene-freeOPVs [36], an extension of previous work byMin et alfrom 2017 [23], which in turn is inspired by the workof Bundgaard et al from 2015 [37]. By taking the synth-etic complexity of the donor and acceptor materialsinto account alongside the PCE and the photostabilityof a device, this industrial figure of merit, i-FoM,

Figure 1.The lab-to-fab challenge: upscaling fabrication of organic solar cells from (a) lab-scale devices fabricated using non-scalablemethods andmaterials on rigid substrates through (b) lab-scale devices fabricated using scalablemethods andmaterials on flexiblesubstrates to (c) large-scale, roll-to-roll fabrication on flexible substrates using fully scalablemethods andmaterials.

Figure 2.Classifications of depositionmethods.

3

Flex. Print. Electron. 5 (2020) 014004 A SGertsen et al

allows for a quantitative comparison of viability forupscaling across different active layers and the result-ing devices. The synthetic complexity reflects thenumber of synthetic steps, the yield, the isolation/pur-ification process, and the number of hazardous chemi-cals used, and it is therefore indirectly a qualitativeestimation of both cost and sustainability of the donorand acceptor materials, making it a strong indicativemeasure of the scalability of the active layer itself. Tak-ing the current rapid development of increasinglycomplex donor polymers and non-fullerene acceptorsinto account [14, 38–40], we believe that this i-FoMvalue could serve as an important tool in the evalua-tion of their scalability going forward. This, however,should not stand alone when discussing the potentialfor upscaling fabrication, since it does not incorporatethe scalability of the fabrication as discussed above.

It is important to underline that scalability encom-passes more than what is related to deposition techni-ques and active layers. Especially broader economicalconsiderations regarding materials, processing condi-tions, and solvents are important for sustainable scal-ability of organic solar cell modules, but alsoenvironmental concerns should be taken into account.A number of significant contributions to the discus-sion of OPV scalability in terms of these parametershave been published throughout the years in the formof economical analyses [6, 7], life-cycle assessments [6,41–43] and analyses of energy payback times [4, 5].Although indeed interesting, these analyses are outsideof the scope of this work, and we thus refer the inter-ested reader to the cited articles.

4.Deposition techniques

In order to discuss scalability from a device fabricationpoint of view, it is important to understand the generalworking principles of how the different layers aredeposited. Traditionally, the focus in the field has beenon varying the active layer deposition, but here we willalso try to incorporate interfacial layer deposition aswell as electrode deposition. Comprehensive reviewsof the different deposition techniques themselves havebeen published elsewhere [1, 44, 45], andwewill hencerefrain from extensive descriptions in the presentpaper and instead emphasize the discussion of theirindividual applicability to large-scale fabricationsetups.

4.1. Coating and printingWewill here distinguish between coating and printingtechniques—the former are used to deposit contin-uous layers along the translational direction of thesubstrate without direct contact to the surface of this,whereas the latter often are associated with thepossibility to perform complex patterning throughdirect contact with the surface of the substrate, e.g. viaa stamp, through the use of masks, or through control

of the flow as is the case for inkjet printing. Because ofthe ability of printing techniques to deposit welldefined patterns, they are highly applicable for elec-trode deposition in semi-transparent devices, whereascoating techniques are most often used for active andinterfacial layer deposition given their continuousnature and possibilities to control film thicknesses byvarying flow rates and/or web speed. All scalabletechniques mentioned in the below paragraphs areillustrated infigure 3.

Spin-coating Spin-coating is a thin-filmdeposition technique relying on the dispensing of asolution onto a rotating substrate. The centrifugal‘force’will distribute the dispensed solution across thesubstrate surface, and combined with simultaneousevaporation of the solvent(s), a uniform thin-film ofthe solute(s) is obtained. This technique allows for easycontrol of the thin-film thickness from tens of nan-ometers to several micrometers by varying the angularspin-speed (the thickness, d, is proportional to theinverse of the square root of the angular velocity, ω:µ

wd 1 ) [44], which, coupled with the possibility of

spin-coating on very small areas, provides a powerfullab-scale technique for testing wide ranges of proces-sing parameters. However, there are significant draw-backs for large-scale implementations such as a highmaterial waste (most of the dispensedmaterial is slungoff of the substrate and onto the walls of the spin-coater) and the lack of possibilities to engineer a con-tinuous version of this. Although commercial imple-mentations of spin-coating in e.g. the LED industryenables deposition on areas of up to 1m2, the inherentbatch process nature of spin-coating combined withthe high material waste has made us label it non-scal-able (see section 3 and figure 2). Spin-coating is thusone of the only deposition techniques that can not beused in scalable fabrication of OPVs, however, itremains to be themost widely used technique in litera-ture for active layer and interfacial layer depositiondue to the low equipment requirements and easyoperation. The many years of experience with spin-coating deposition and historically the record deviceefficiencies achieved are obviously also significant dri-vers for thewidespread use of this technique.

Doctor-blading/blade coating Because of itsminimal equipment requirements and easy transfer-ability to a roll-to-roll setup (as knife coating, seebelow), doctor-blading is often employed as the firststep towards amore scalable deposition of active layerscompared to spin-coating. By depositing an inkdirectly onto a substrate and subsequently dragging asharp knife or a blade across it at a fixed distance, a wetthin-filmwith a well-defined thickness is obtained (seefigure 3(a)). The empirical relationship defining thedry thickness, d, of this film is given by =

rd g c1

2,

where g is the distance between the blade and the sub-strate, c the ink concentration, and ρ the dry film den-sity [44]. This technique is accompanied by some

4

Flex. Print. Electron. 5 (2020) 014004 A SGertsen et al

material waste, but with the possibility of obtainingcoatings using only small amounts of material, it is astrong technique for laboratory scale testing. As themain limitation, formation of the thin-film using

doctor-blading is slow compared to spin-coating, andvolatile solvents combined with highly concentratedinks can thus lead to non-uniform films if aggregationhas time to occur. In general, however, the longer

Figure 3.Deposition techniques that are partly or fully scalable (excluding vacuum techniques).

5

Flex. Print. Electron. 5 (2020) 014004 A SGertsen et al

solvent evaporation times for doctor-blading aremorecomparable to the ones of other scalable depositiontechniques.

A variation of the doctor-blading technique thatdeserves mentioning is the fluid-enhanced crystalengineering (FLUENCE) technique developed byDiaoet al in 2013 [46]. By patterning the ink contact side ofthe blade with micropillar arrays, a flow-inducedextensional strain facilitates increased crystallinitywhile simultaneously reducing domain sizes in all-polymer solar cells, in turn improving the device char-acteristics [47]. This principle could readily be appliedto other deposition techniques described herein.

Knife coating Knife coating (or knife-over-edge coating) can be regarded as the continuous, roll-to-roll compatible analog to doctor-blading: an inksupplier gradually adds excess ink to a bath down-stream of a knife, which controls the thickness of thewet film through its proximity to the substrate (seefigure 3(b)). It is in essence a zero-dimensional techni-que, but the inherent lack of control of the width of thedeposited layer can in part be solved by adding barriersto the ink bath and thus prohibit material waste,whereas also viscous inks allow for well-definedwidths. The somewhat unknown width can further-more complicate the calculation of the dry thickness(which is defined in a similar way to the one of doctor-blading, see above). In spite of knife coating being afully scalable technique, the higher degree of pattern-ing control in e.g. slot-die coating for active layerdeposition will probably limit the applicability of knifecoating in commercial setups.

Slot-die coating Slot-die coating enables a con-tinuous, roll-to-roll compatible deposition of manyvarieties of inks and not least a one-dimensional con-trol of coating patterns in the form of one or morestripes with well-defined widths. In slot-die coating,the ink is supplied via a pump to a slot-die coatinghead through which the ink is deposited onto a mov-ing substrate (illustrated in figure 3(c)). There is prac-tically almost no material waste, and the thickness of

the dry film can thus easily be calculated as =r

d c f

vw,

where f is the ink flow rate, v the coating velocity (i.e.the speed of the substrate or of the coating head), andw thewidth of the deposited ink [44].

The limitations of this technique are mainly rela-ted to the fluid dynamics defining the so-called coatingwindow, i.e.the range of parameters for which a stablemeniscus can be obtained, which relies on a range ofproperties, including flow rate, ink viscosity, distanceof the coating head to the substrate, and coating velo-city [48]. Slot-die coating is, however, a forgiving tech-nique in the sense that a broad range of ink viscositiescan be used and that its continuous nature allows forvisual feedback while fine-tuning the above men-tioned mechanical parameters until a stable meniscusis formed.

One of the drawbacks of conventional slot-diecoating is the lack of temperature control in the tubingand the slot-die head. Compared to spin-coating,where the time from removal of the ink from a heat-bath to the deposition onto a heated substrate can bevery short, the longer time needed for slot-die coatingcan cause problems for inks that undergo gelationbelow certain temperatures. This can in part be solvedby hot slot-die coating in which the slot-die head isheated and connected to a thermo-couple, providingan extra temperature control in addition to substratehot-plates—this depositionmethod has also proven tobe very beneficial in terms of device efficiency asdemonstrated in [33, 49].

Flat-bed screen printing Especially applicablefor electrode deposition owing to its full two-dimen-sional patterning control, flat-bed screen printingemploys a mask through which the ink is pushed intocontact with the substrate using a moving squeegee(see figure 3(d)). Because of the nature of this process,there are significant limitations in terms of the inkproperties: it should have a high viscosity to avoiddeviations from the patterning and the solvent shouldhave a low volatility to avoid evaporation causing con-centration gradients and thus differences in dry filmthickness along the squeegee translational direction.As the thickness of the deposited layer is defined by themask thickness, screen printing techniques are mostlyapplicable when thick layers (wet layer thicknesses of10–500 μm) are needed; [1] the dry thickness can beestimated by =

rd k Vp

cscreen , where Vscreen is the paste

volume of the screen (theoretical volume of wet inkdeposited per area of mask hole) and kp is the ratio ofthis wet ink that practically remains on the substrate[44]. There is potentially only a low material wasteconnected to this technique, and with it being a sheet-to-sheet process easily applicable for large areas, it islabeled partly scalable. This, however, makes it astrong technique for laboratory scale testing, and withthe possibility to adapt it to fully continuous roll-to-roll setups through rotary screen printing (see below),observations and results from flat-bed screen printingare readily transferable to fully scalable fabrication.

Rotary screen printing Developed as a roll-to-roll compatible version of screen printing, rotaryscreen printing makes use of a stationary squeegeearound which a mask rotates (see figure 3(e)). An inkbath supplies material, which in a similar fashion toflat-bed screen printing is pushed through the holes inthe mask onto the substrate to reproduce the patternof themask. The limitations related to the ink viscosityare the same as in flat-bed screen printing, but thevolatility of the solvent can be higher, as the ink issomewhat protected from the surroundings inside thescreen.

Inkjet printing Inkjet printing is a digitallycontrolled patterning technique known from standardprinters. Using a nozzle with a piezoelectric stage or athermal unit to eject ink droplets that are then

6

Flex. Print. Electron. 5 (2020) 014004 A SGertsen et al

electrostatically charged and accelerated towards thesubstrate by an electric field, a digital pattern can bereproduced with high resolution and no material loss(illustrated in figure 3(f)). Although indeed attractivefor niche applications where complex or varying pat-terning as well as aesthetics are necessary, the slowspeeds, relative to slot-die coating, with which inkjetprinting can coat large areas are a potential limitationfor its use in large-scale fabrication of organic solarcells. We have hence, despite its apparent compat-ibility with a continuous roll-to-roll setup, labeled itpartly scalable. Furthermore, restrictions on the ink tohave a low viscosity to be able to form droplets can beprohibitive for inkjet printing of some layers. Thethickness of a dry film deposited by inkjet printing canbe calculated as =

rd N Vd d

c , where Nd is the number

of droplets with volume Vd deposited per unitarea [44].

Spray coating Like inkjet printing, spray coat-ing relies on droplet formation of the ink. However,the requirements to the ink are more lenient com-pared to inkjet printing, facilitating the use of inkswith a wide variety of rheologies and viscosities [50].Themost common spray coating technique is airbrushspray coating, where an aerosol is formed by forcingthe ink out of the nozzle using a gas flow, usually N2

(see figure 3(g)). Whereas inkjet printing is a two-dimensional patterning technique, spray coating isessentially zero-dimensional, although with the possi-bility of some one-dimensional control if variations inthe stripe edges can be accepted. This sacrifice ofpatterning control, however, enables a significantspeedup of the deposition, allowing for a more mean-ingful roll-to-roll implementation, but the relativelyhigh surface roughness and thus the need for thickerlayers to prevent pinholes can in practice lead to low-efficiency or even defect cells and in turn to materialwaste. We have therefore categorized this techniqueas only partly scalable despite its roll-to-rollcompatibility.

Gravure printing Known from commercialprinting, gravure printing is based on a gravure rollertransferring ink from a bath via its engraved cavities tothe substrate when pressed into contact (visualized infigure 3(h)). This allows for high speed processing andtwo-dimensional patterning in a continuous roll-to-roll setup with the shape and thickness of the obtainedpatterns defined by the engravings in the gravureroller. As the main limitation of gravure printing, thenecessary optimization of the ink’s surface tensionshould be mentioned, since factors such as the inkrheology and the pressure of the gravure roller on thesubstrate affect the quality of the print significantly [1].

Flexographic printing In flexographic printing,the ink is transferred from the bath via an anilox roller,which is a cylinder with ink-collecting micro-cavities,to the relief of a printing roller that then ‘stamps’ itspattern onto the substrate when pressed into contact(see figure 3(i)). Like for gravure printing, this allows

for two-dimensional patterning control in continuousroll-to-roll setups, and the remaining advantages andlimitations are very similar to this too.

4.2. VacuumdepositionPreviously, vacuum steps have been regarded as beingnon-compatible with large-scale fabrication of organicsolar cells, and ‘vacuum-free’ has often been used inliterature as a precondition for scalability [21, 44,51–53]. We would, however, like to challenge thatposition with reference to the numerous commercialphotovoltaic technologies incorporating vacuumdeposition steps such as organic light emitting diodesand silicon solar cells. Furthermore, Heliatek hasdemonstrated with their HeliaFilm® pilot line in 2016and later with a small-scale fabrication line that a fullroll-to-roll setup in inert atmosphere and with severalvacuum steps is indeed realizable and not leastcommercially promising [54]. Very recently, the groupled by M Madsen at the University of SouthernDenmark also demonstrated roll-to-roll vacuum sput-tering using their in-house setup [55]. Vacuum stepsare thus not prohibitive for the upscaling of thefabrication itself, and if the costs related tomaterial useand processing conditions, amongst these high-temp-erature steps, can be kept down as indicated by thecommercial nature of theHeliaFilm® project, vacuum-and inert steps are likely to be part of the future large-scale fabrication of organic solar cells because of theefficiency gains usually seen compared to solutionprocessing in ambient conditions.

Thermal evaporation Also known as vapordeposition, thermal evaporation relies on resistiveheating of an evaporation source, for example silver inthe case of electrode deposition, under vacuum until avapor pressure is reached and the evaporated silver isdeposited on a substrate, forming a thin-film. Thermalevaporation allows for a precise control of the layerthickness and produces highly uniform layers, andpatterning control is achievable through the use ofshadow masks: for one-dimensional control, a sta-tionarymaskwould be sufficient, whereas two-dimen-sional patterning in a continuous setupwould demanda mask moving with the same speed as the substrate.The requirement of shadowmasks for patterning con-trol is, however, linked to a not insignificant materialwaste, but the material deposited on the shadow maskcould potentially be recycled. Even though this is a sig-nificant challenge, especially in terms of economywhen using expensive materials such as silver, we havein the evaluation of this deposition method chosen toput emphasis on the possibility to integrate it into aroll-to-roll setup for continuous deposition and thuslabeled it partly scalable.

Sputtering In sputter deposition, material iseroded off of a target source, e.g. molybdenum, using,in most cases, argon plasma. The sputtered materialwill then deposit on the substrate to form a thin-film.The atmosphere in the sputtering chamber can be

7

Flex. Print. Electron. 5 (2020) 014004 A SGertsen et al

tuned to need—for example in the case ofMoOx hole-transport layers, molybdenum atoms are sputtered ina controlled oxygen atmosphere to obtain a stoichio-metrically desired MoOx layer. Like for thermal eva-poration, the uniformity of the deposited layer is high,and its thickness can be controlled with a very highprecision down to single nanometres. The need forshadowmasks poses problems identical to the ones forthermal evaporation, but it should be noted that theamount of material needed per area is usually sig-nificantly lower for vacuum deposition than for solu-tion processing because of the homogeneity of thevacuumdeposited layers.

5.Overview of scalably fabricated,fullerene-freeOPVs from literature

The availability of equipment and especially the ease ofuse are important explanatory factors for the relativelyfew studies published on fully scalable fabrication oforganic solar cells. However, as discussed above,techniques like blade coating and flat-bed screenprinting are optimal for laboratory-scale optimizationand readily transferable to continuous roll-to-rollsetups, but whereas blade coating has recently beenused routinely for active layer deposition, top electro-des and interfacial layers are still, almost exclusively,thermally evaporated.

Opposed to scalable deposition techniques, the useof non-fullerene acceptor materials is not a pre-requisite for commercial fabrication of organic solarcells, but they have to a large extent simply out-com-peted fullerene acceptors. Fullerene acceptors havehistorically been widely used in the active layers oforganic solar cells owing to their favorable propertiessuch as high electron affinities, high electron mobi-lities, and easy solution processing [56]. However, thelow-energy transitions in fullerenes are dipole for-bidden owing to their high molecular symmetry, inturn leading to weak optical absorbance in the visiblespectrum, which is a significant limitation for furtherimprovements in the efficiencies of fullerene-basedorganic solar cells. The optical properties of non-full-erene acceptors can to a higher degree be tuned by che-mically engineering their molecular structure. Themost widely used design principle for small-molecule,non-fullerene acceptor materials is to utilize a con-jugated internal acceptor–donor–acceptor structurein which two electron withdrawing units (internalacceptors) are separated by a central electron donatingunit (internal donor) and potential bridging units[57, 58]. In this way, the low-energy transitions arered-shifted due to a promotion of charge-transferstates, in turn facilitating an optical absorption profiledominant in the red part of the visible spectrum, com-plimentary to most polymer donor materials, which

absorb in the blue and green parts of the spectrum. Asecond way to achieve this is by employing polymeric,non-fullerene acceptor materials with internally alter-nating donor–acceptor structures [38, 59, 60], similarlyfacilitating low-energy charge-transfer absorptions.Like for the fullerene acceptors, high electron affinitiesof both small-molecule and polymer non-fullereneacceptors are obviously paramount, but the active layerprocessing conditions for which optimal microphaseseparation and domain purity occur to ensure highelectronmobilities can vary greatly for the three types ofacceptors and not least for different deposition techni-ques. These considerationswill be discussedbelow.

Another important consideration for scalableprocessing relates to the device architecture. On flex-ible substrates, and especially in a roll-to-roll context,the inverted device architecture has proven to be themost practical given the available materials and pro-cessing methods. In particular, hole-transport layershave shown to be problematic in normal devicearchitecture solar cells, as they have to be both highlytransparent (all light has to pass through it to reachthe active layer) and mechanically robust (being thefirst layer processed on top of the transparent elec-trode, it is subject to high stress). The commonly usedmaterials like PEDOT:PSS and hole-conductingmetal oxides have so far not proven processable in away where these demands are fulfilled. On the otherhand, the materials for electron-transport layers havenot suffered from the same problems. Highly trans-parent materials like ZnO and TiOx are routinelyused and have proven themselves as good front mate-rials in inverted architecture solar cells while simulta-neously allowing the use of less transparent and lessrobust hole-transport materials such as the abovementioned at the back of the solar cell. The dom-inance of the inverted architecture, as will be obviousfrom the following sections, is thus predominately aconsequence of the availability of suitable materials.If new hole-transport materials with the right prop-erties are found, there is in principle no reason whynormal architecture solar cells could not be used inthe future.

Throughout the coming sections, we have high-lighted groundbreaking works and their resultingdevices in figures 4–8. The molecular structures of alldonor polymers mentioned in these sections areshown in figure 9, of all non-fullerene acceptors infigure 10, and of all molecular interfacial layers infigure 11. The device parameters for all mentioneddevices, including short-circuit currents, open-circuitvoltages, and fill-factors (FF), as well as qualitative esti-mates of the scalability of the materials and depositiontechniques are summarized in table 1. Further detailsregarding this table can be found in section 5.5.

8

Flex. Print. Electron. 5 (2020) 014004 A SGertsen et al

5.1.Devices fabricated using solely roll-to-rollcompatible depositionmethods andno vacuumstepsOnly a few studies on fully roll-to-roll compatible,vacuum-free processing of non-fullerene systems havebeen reported. The first effort in this respect waspublished in 2013 by Liu et al [64], where theyinvestigated the effect of upscaling small area deviceson glass substrates with spin-coated active and inter-facial layers and thermally evaporated electrodes to acomplete roll-to-roll fabrication of large-areamoduleson flexible substrates. For the flexible devices, aninverted structure of indium tin oxide (ITO)/ZnO(NP)/PDI-DTT:PSBTBT/PEDOT:PSS/Ag was used;the polyethylene terephthalate (PET) substrate withITO was purchased from a commercial supplier, thezinc oxide nanoparticles (ZnO(NP)), active layer(PDI-DTT:PSBTBT, seefigures 9 and 10, respectively),and PEDOT:PSS (see figure 11) were slot-die coated,and the silver back electrode was deposited using aroll-to-roll integrated flat-bed screen printer. Theaverage efficiencies of the resulting 4.2 cm2 moduleswere 0.20%, a factor of three lower than the small-area, spin-coated devices on glass substrates, leavingnotable room for improvement. The effect of thesubstrate (PET versus glass) was concluded to be adecisive factor, but probably most problematic forupscaling (disregarding the low performance) was theuse of ITO. ITO has been shown to be both economic-ally and environmentally critical, and in addition, thesignificant fraction ofmore than 80%of the embeddedenergy in similar modules stemming from the ITOcoated PET posed a significant impediment for theprojected energy payback times [6, 41].

This problematic use of ITO had already beenaddressed in fullerene-based OPVs at several occa-sions [65, 66], but the first study of roll-to-roll compa-tible processing of non-fullerene OPVs on flexiblesubstrates without ITO was not published until 2014

by Chen et al [67]. This was additionally the first studylooking to replace the fullerenes in the well-knownmodel system P3HT:PC61BM (see figure 9) withsmall-molecule, non-fullerene acceptors in fully roll-to-roll processed OPVs, but the efficiencies reachedwere lower than 0.1%. However, the deposition tech-niques and the device structure used therein have beenthe dominant in literature since. The processingequipment was introduced in 2012 byDam andKrebs,who reported a laboratory-scale coating/printingmachine enabling the fully scalable processing of alllayers in a stand-alone setup [68], and the device struc-ture was introduced by Carlé et al later that year [69].Using an inverted architecture of PET/Ag/PEDOT:PSS/ZnO(NP)/D:A/PEDOT:PSS/Ag (D: donor, A:acceptor), ITO- and vacuum-free devices could be rea-lized, allowing for lab-scale assessment of new activelayers in the context of large-scale fabrication. Theprocessing is, in principle, straightforward: flexo-graphic printing of a silver paste onto the PET sub-strate, slot-die coating of a ZnO nanoparticle solution,slot-die coating of a PEDOT:PSS ink, slot-die coatingof an active layer ink, slot-die coating of a secondPEDOT:PSS ink, and lastly flexographic printing of asilver paste as the top electrode. This also enables theuse of pre-processed substrate foils with bottom elec-trodes and ZnO electron-transport layers alreadyapplied, making the testing of new systems simple aswell asminimizingmaterial waste.

The above procedure has been used in almost allstudies of fully roll-to-roll compatible, non-full-erene acceptor OPVs published subsequently. In2014, Cheng et al aimed to study the effects of the1,8-diiodooctane (DIO) high boiling-point additiveand to compare spin-coating on glass substrates withslot-die coating on flexible substrates in this type ofsetup for both fullerene-based systems and all-poly-mer systems [70]. Of the four combinations, the scal-ably processed, flexible, all-polymer cells with an

Figure 4. Flexible, ITO-free, vacuum-freeOPVmodule fabricated using continuous roll-to-roll deposition techniques at theTechnical University ofDenmark.

9

Flex. Print. Electron. 5 (2020) 014004 A SGertsen et al

inverted Ag/PEDOT:PSS/ZnO(NP)/PBDTTT-C-T:PDIDTT/PEDOT:PSS/Ag structure (see figure 9for structures; note that PDIDTT is only shown inhere despite of its applicability as an acceptor too)showed the lowest average PCEs of 0.67% for 1 cm2

devices. This was followed by a paper from the sameauthors in 2015 [11], using identical device struc-tures and deposition techniques but with a small-molecule, non-fullerene acceptor (active layer:PBDTTT-C-T:DC-IDT2T, see figures 9 and 10).This led to a champion efficiency of 1.0% for a 1 cm2

device, which was, however, still a factor of twolower than the PC71BM analog. On the other hand,the non-fullerene devices showed a far superior sta-bility under continuous AM 1.5G illumination,maintaining more than 80% of their initial efficiencycompared to the mere 50% of the fullerene-baseddevice. This increased stability has later been shownto be a somewhat general characteristic for small-molecule, non-fullerene acceptors [12, 13, 16], giv-ing them a significant advantage over fullereneacceptors for commercial viability.

The following year in 2016, Liu et al investigateddevices based on the PTB7-Th:IEIC active layer (seefigures 9 and 10 for molecular structures) [71]. Withchampion efficiencies of 6.31% in lab-scale, spin-coateddevices on glass substrates with evaporated electrodes, itwas a good candidate for upscaling to flexible devicesdeposited with fully scalable methods. They employedtwo types of flexible device structures on PET foil: anITO-free one, namely Ag/PEDOT:PSS/ZnO(NP)/PTB7-Th:IEIC/PEDOT:PSS/Ag, and an ITO-con-taining one, namely ITO/ZnO(NP)/PTB7-Th:IEIC/PEDOT:PSS/Ag. As described above, the ITO-freedevices were deposited using flexographic printingfor the electrodes and slot-die coating for the remain-der of the layers, whereas the PET/ITO foil was pur-chased and the remainder of the layers wereprocessed as for the ITO-free devices. For the ITO-free devices, an average PCE of 1.60% with a cham-pion efficiency of 1.79% was obtained for 1 cm2 cells,whereas the ITO containing devices reached an aver-age PCE of 2.05% and a champion efficiency of2.26% for 0.7 cm2 cells—all of these were slightlylower than their fullerene (PC61BM) counterparts,but were at the time the highest reported efficienciesfor flexible, non-fullerene devices. All cells in thisstudy were suffering from low FF of around 35%, butthe fullerene-free cells exhibited slightly higher FFthan the fullerene cells, whereas the fullerene cellshad significantly higher short-circuit currents. Stabi-lity tests were also performed, showing increased sta-bility in the fullerene-free devices, in turn supportingthe statementmade in the previous paragraph.

Later that year, Brandt et al reported the only sec-ond work on combining P3HT and non-fullerene

acceptors using roll-to-roll compatible processing[52]. Here, they investigated variations in absorp-tion, crystallinity, and device performance based onthe geometrical effects of three diketopyrrolopyrroleacceptors with different degrees of ground state pla-narity in a combined study between quantum che-mical calculations, X-ray experiments, and devicecharacterization. Obtaining only low efficiencies of0.54% for the best roll-coated device (using the samedevice structure as described above), the mostimportant conclusion drawn from this study wasthat the less crystalline system performed better inroll-coating, whereas the more crystalline systemperformed better in spin-coating. This underlinesthe need for in situ morphological studies of activelayer deposition to probe the microstructure evol-ution during solvent evaporation [72–75].

For a couple of years after this, no studies on fullyscalably fabricated, fullerene-free devices were pub-lished, but from 2016 onwards, significant effort hasbeen put into synthesis of new and improved non-fullerene acceptors. The impressive efficienciesexceeding 10% reached in lab-scale devices usingIDTBR small-molecule acceptors [13, 76–78] has, in2018, motivated Strohm et al to produce P3HT:O-IDTBR modules using fully scalable depositionmethods (see figure 10 for the structure of IDTBR)[61]. Although the modules were deposited on ITOcoated glass substrates, which prohibits a true industrialfabrication as discussed above, we have chosen toinclude their work in this section because of their effortto upscale both the interfacial layers themselves andtheir deposition as well as the deposition of the top elec-trode. Using a device structure of ITO/ZnO(NP)/P3HT:O-IDTBR/PEDOT:PSS/AgNW (AgNW: silvernanowires), doctor-bladed 0.1 cm2 cells with an averagePCE of 5.25% and an average FF of 66.6% were fabri-cated using a solvent formulation of chlorobenzenewith 5% 4-bromoanisole additive for the active layerprocessing. 59.5 cm2 modules using the same devicestructure and processing conditions exhibited effi-ciencies of an impressive 5.0% (see figure 5), whereasexchanging doctor-blading for slot-die coating yieldedmodules with efficiencies of up to 4.4%. This system isthus indeed interesting for further studies on flexible,ITO-free substrates using true roll-to-roll deposition.

From this limited number of works on scalablyfabricated, non-fullerene OPVs, it is clear that there isroom for significant progress in the field. In line withour recommendations in the introduction, we urge anincreased effort to demonstrate scalability both interms of deposition techniques and materials. As willbe evident from the below sections, promising mat-erial systems and solvent formulations as well asimportant considerations regarding interfacial layers

10

Flex. Print. Electron. 5 (2020) 014004 A SGertsen et al

have been put forth, which, combined, will surely leadto advances for fully scalably fabricatedOPVs.

5.2.Devices fabricated using solely roll-to-rollcompatible depositionmethods but vacuumstepsAs discussed in section 4.2, vacuum depositiontechniques are likely to play a role in future large-scalefabrication of organic solar cells if the processing costscan be kept low; the techniques themselves are notincompatible with roll-to-roll setups. In this section,we will thus highlight studies utilizing vacuum deposi-tion for the top electrodes and hole-transport layersbut using continuous roll-to-roll deposition techni-ques for the active layers.

In 2017, Gu et alwere the first to reach the 5% effi-ciency mark for fullerene-free organic solar cells usingroll-to-roll deposition of the active layer [3]. In thisstudy, different all-polymer active layers were studiedin a PET/ITO/ZnO(NP)/D:A/MoO3/Ag devicestructure, where the ZnO(NP) and active layers wereslot-die coated onto a pre-produced PET/ITO foil in acustom-built roll-to-roll setup and the MoO3 hole-transport layers as well as the silver top electrodes werethermally evaporated. Using sidechain engineering tocontrol crystallinity, the two donor polymers PII2Tand PII2T-PS (see figure 9)were synthesized and cate-gorized as crystalline and low-crystalline, respectively,using grazing-incidence X-ray scattering. They werethen paired with the two acceptor polymers PNDITand PPDIT (see figure 10), similarly categorized ascrystalline and low-crystalline, respectively. Spin-coated, lab-scale devices were then fabricated for each

of the four pairs, showing that suppressing crystallinityled to higher device efficiencies caused by lower phase-separation sizes. The low-crystalline PII2T-PS:PPDITpair was hence identified as the best candidate forupscaling, and cells were fabricated using the devicestructure described above. A small module with acombined area of 10cm2 was characterized, showingan average PCE of 4.1% with a champion PCE of4.24%measured over 12 of the 0.12 cm2 cells that wereconnected to form the 10 cm2 module. With theseimpressive results in mind, they extended the study toencompass the PTB7-Th:PPDIE active layer (seefigure 10 for molecular structures of PPDIE), whichexhibited even lower crystallinity and phase-separa-tion sizes than the PII2T-PS:PPDIT combination. Theroll-to-roll coated devices based on this PTB7-Th:PPDIE active layer showed an average PCE of 5.0%with a champion PCE of 5.1%, at the time a record forflexible organic solar cells with continuously printedactive layers. This work furthermore substantiates thefindings of Brandt et al [52]described in the previoussection, putting additional emphasis on the impor-tance of morphological studies and showcasing thestrength of developing design principles.

Very recently, in 2019, Na et al reported the cur-rent record efficiency of 7.11% for non-fullereneorganic solar cells with roll-to-roll deposited activelayers [33]. Extending their previous work on full-erene-based OPVs [49], their novel modification of aslot-die coater was used to investigate the effects ofdeposition temperature on device parameters of full-erene-free OPVs. By implementing a heating elementin the slot-die head of a modified 3D printer,

Figure 5.RigidOPVmodule with P3HT:O-IDTBR active layers fabricated using roll-to-roll compatible deposition techniques byS Strohm, FMachui, and co-workers. Reproduced from [61]with permission fromTheRoyal Society of Chemistry.

11

Flex. Print. Electron. 5 (2020) 014004 A SGertsen et al

independent temperature control of the solution and thesubstrate was achieved (see figure 6(a)). Optimizationshowed large deviations of several percentage points inPCE with varying substrate temperature, whereas fixingthe substrate at its optimum temperature of 120 °C andvarying the slot-die head temperature showed a coupleof percentage points difference with an optimumaround 90 °C. They used an ITO/ZnO(NP)/PEIE/PBDB-T:ITIC/MoO3/Ag device structure (seefigures 9–11 for the structures)with pre-produced ITO-coated substrates, slot-die coating of the ZnO nano-particles, the PEIE electron-transport layers, and theactive layers, and thermal evaporation of MoO3 and thesilver top electrodes. PCEs of 10.0% on glass substrateswith slot-die batch processing, 8.77% on PET substrateswith slot-die batch processing, and 7.11% on PET sub-strates with full roll-to-roll, continuous slot-die proces-sing were achieved for 0.07 cm2 areas (see figure 6(b)).The latter is close to the current efficiency record forflexible OPV devices with roll-to-roll processed activelayers of 7.32%, which was reached using anITO/AZO:PEIE/PTB7:PC71BM/MoO3/Ag structure(AZO: aluminum-doped zinc oxide) in 2017 [34]. Thediscrepancy between the batch process and the roll-to-roll process for flexible substrates is explained by thephysical contact between the coated film and the back-side of the substrate on the rewinder roll, but this contactwill be avoided in commercial roll-to-roll setups beforethe addition of the remainder of the layers or even beforeencapsulation. These results thus strongly indicate thatdual temperature control can be key in overcoming thelab-to-fab challenge and realizing large-scale fabricationofflexibleOPVswithhigh efficiencies.

5.3.Devices fabricated using partly scalable activelayer depositionmethodsThe vast majority of fully scalable deposition techniquesare not straightforward to adopt in small-scale labora-tory testing of costly material systems of which onlysmall amounts are available. In particular, doctor-

blading has been a popular choice as the first steppingstone towards scalable fabrication of organic solar cellsbecause of its easy transferability to continuous proces-sing. This makes it a strong technique for optimizationof active layer solutions in terms of solvents, additives,material composition, and processing conditions inbatchprocesses prior toupscaling.Hence, in this section,we will keep a principal focus on the active layers andreview the most notable works using doctor-blading orother partly scalable techniques for the active layerdeposition. Unless otherwise mentioned, ZnO(NP)electron-transport layers were spin-coated and MoO3

hole-transport layers and Al or Ag top electrodesthermally evaporated for all revieweddevices below.

In the doctor-blading paragraph of section 4, webriefly touched upon the FLUENCE technique. In 2015,Diao et al used this variation of the doctor-blading tech-nique to alter the morphology of all-polymer active lay-ers [47]. It was found that the flow design with amicrostructured blade increased the crystallinity of neatdonor PII-tT-PS5 thin-films while concurrently redu-cing domain sizes in blend PII-tT-PS5:PPDIT thin-films(PPDIT is also denoted P(TP); see figures 9 and 10,respectively, for structures). Additionally, the surfaceroughness was also reduced significantly compared toregular, unstructured doctor-blading, in turn improvingthe reproducibility of device efficiencies. The combinedeffect of these properties led to a champion PCE of 3.2%in an inverted glass/ITO/ZnO(NP)/PII-tT-PS5:PPDIT/MoO3/Al structure, the record efficiency forblade-coated, all-polymer organic solar cells at the time.

The following year in 2016, Ye et al reached a newrecord efficiency for blade-coated, all-polymer OPVs[79]. By doctor-blading a PBDT-TS1:PPDIODT activelayer (see figures 9 and 10) in a green solvent, namely2-methylanisole, in an inverted glass/ITO/ZnO(NP)/MoO3/Al architecture, a champion PCE of 5.21% wasachieved. Thiswas one of the earlier efforts to replace thehalogenated solvents regularly used for a green solvent,allowing for a more environmentally friendly

Figure 6. (a) Illustration of dual temperature control obtained in hot slot-die coating and (b) J–V characteristics of the differentdevices fabricatedwith slot-die coated active layers by S-INa,DVak, and co-workers. Reproduced from [33]with permission fromWILEY-VCHVerlagGmbH&CoKGaA,Weinhim.

12

Flex. Print. Electron. 5 (2020) 014004 A SGertsen et al

processing, which is particularly desirable for large-scalefabrication. More recently in 2019, Lin et al also focusedon the use of non-halogenated solvents to control themorphology of all-polymer OPVs based on PTzBI:N2200 active layers (seefigures 9 and 10, respectively, forstructures) [80]. Using 2-methyltetrahydrofuran as theprocessing solvent, a champion PCE of 8.36% wasachieved, whereas devices processed from chlor-obenzene only reached 2.92% (both in glass/ITO/ZnO(NP)/D:A/MoO3/Al inverted structures).This significant difference underlines the importance ofexploring alternative—and preferably green—solventformulations.

Moving to small-molecule acceptors, the ITIC non-fullerene acceptor and derivatives hereof have domi-nated the scene of blade-coated devices since late 2017.In December, Zhao et al investigated a methylated ITICderivative, IT-M, in conjunction with the polymericdonor PBTA-TF (see figures 9 and 10) processed ingreen solvent formulations [81]. By comparing spin-coating and blade-coating of this active layer using botha low-boiling point solvent blend, namely tetra-hydrofuran/isopropanol, and a high-boiling point sol-vent blend, namely o-xylene/1-phenylnaphthalene, itwas found that the spin-coated devices performedslightly better when processed from the high-boilingpoint blend, whereas the blade-coated devices per-formed significantly better when processed from thelow-boiling point blend. For these tetrahydrofuran/iso-propanol processed, blade-coated devices, a recordPCE of 11.7% was obtained in a conventional glass/ITO/PEDOT:PSS/D:A/PFN-Br/Al architecture (seefigure 11 forPFN-Br structure) and11.3% in an invertedglass/ITO/ZnO(NP)/D:A/MoO3/Al architecture, bothfor 0.04 cm2 devices. Impressively, large-area conven-tional devices of 1.0 cm2 maintained a high efficiency of10.6%, showing great promise for both thismaterial sys-tem and the tetrahydrofuran/isopropanol solvent for-mulation for blade-coating.

In January 2018, Ye et al also investigated IT-M in anactive layer with the FTAZ donor (see figure 9) similarlyusing a conventional glass/ITO/PEDOT:PSS/D:A/PFN-Br/Al architecture [31]. Once more, chlor-obenzene processing, even with additives, was shown tobe inferior to processing in additive-free, non-haloge-nated solvents, exemplified by pure toluene yielding achampion PCE of 11.0% for a 0.07 cm2 cell, close to theabove mentioned record at the time for blade-coatedOPVs. For an area of 0.56 cm2, an impressive PCE of9.80% was reached, and with a dark stability of 85% ofthe initial PCE after 1000 hours in nitrogen atmosphereas well as only minimal FF reductions with longerannealing times at 150 °C, themorphological stability ofthis material system shows promise for adoption tocommercial fabrication.

Simultaneously in January 2018, Lin et al publishedtheir efforts on doctor-blading flexible, large-area devi-ces based on ITIC [62]. In a comparative study, theyinvestigated the difference of spin-coated and blade-coated active layers as well as that of rigid substrates andflexible substrates (see figure 7). In addition, theyworked with ITO-free PET for the flexible substrates,yielding the overall most scalable cells reviewed in thissection. Starting from the glass substrates, doctor-blad-ing the active layer in an ITO/ZnO(NP)/PTB7-Th:ITIC/MoO3/Ag inverted structure, a champion PCE of9.54% was achieved, slightly higher than the 9.31% ofthe spin-coated analog. For the flexible substrates, thisdifference was more pronounced with a champion PCEof 7.60% for doctor-blading and 5.86% for spin-coatingin an inverted Ag/TiOx/PTB7-Th:ITIC/PEDOT:PSSdevices with large areas of 2.03 cm2

—the former 7.60%a record for large-area, flexible, ITO-free, non-fullereneOPVswith blade-coated active layers.

Later in 2018, Zhang et al, investigated the depend-ence of device characteristics on DIO additive contentin chlorobenzene active layer processing solution forspin- and blade-coated glass/ITO/ZnO(NP)/PBDB-T:ITIC/MoO3/Al cells [32]. Historically, DIO has been

Figure 7. (a)Fabricationof rigid andflexible devices and their corresponding architectureswith blade-coatedactive layers and(b) comparisonof J–V characteristics of small- and large-area devices fabricatedwith spin-coated andblade-coated active layers byYLin, FLiu, FZhang, LHou, and co-workers. Reproduced from [62]withpermission fromWILEY-VCHVerlagGmbH&CoKGaA,Weinhim.

13

Flex. Print. Electron. 5 (2020) 014004 A SGertsen et al

key to achieving high efficiencies in devices depositedfrom halogenated solutions, but the difficulty of remov-ing residual DIO content due to its high boiling-pointhas been shown to lead to acceleratedmorphology evol-ution [82] and degradation of device performancecaused by iodine radicals formed under irradiation [83].Using processing solvent formulations not dependenton DIO additives, as in the above described work of Yeet al [31], is thus the most obvious solution, althoughdecreasing the DIO content would also be a step in theright direction. In the work mentioned in thisparagraph by Zhang et al [32], it was found that blade-coated devices exhibited an optimum PCE of 10.0% fora DIO content of 0.25%, whereas the optimum PCE of9.41% for spin-coated devices was achieved for a DIOcontent of 1.00%. This was reflected in the stability stu-dies of unencapsulated devices, where the blade-coatedcells with 0.25% DIO outperformed the spin-coatedwith 1.00% DIO on all parameters, both under illumi-nation in ambient and in the dark under nitrogenatmosphere, indicating that both of the discussed degra-dation pathways were indeed relevant. These findingsmotivate morphological studies on residual additivecontent in dry OPV thin-films and in particular studieson alternative, additive-free processing solutions.

Jumping to 2019, Ji et al set out to investigate andopt-imize the surface morphology of ZnO electron-transportlayers, leading to some of the highest efficiencies both forblade-coated OPVs in general of 12.88% for 0.12 cm2

cells and for blade-coated, fullerene-free, large-area(>1 cm2)OPVsof 9.22% for 1.04 cm2 cells [84]. Twodif-ferent solutionsof zincoxidenanoparticleswerepreparedand used in the inverted glass/ITO/ZnO(NP)/PBDB-T:IT-4F/MoO3/Al devices (see figure 10 for IT-4F struc-ture): one in acetone (A-ZnO) and one in methanol (M-ZnO). Scanning electron microscopy images revealedinhomogeneities and voids in pristine ZnO films spin-coated from acetone, whereas the methanol processedfilms exhibited compactness and an increased homo-geneity. Additionally, atomic force microscopy showed ahigher surface roughness for A-ZnO than for M-ZnO,which was ascribed to the faster drying process of the for-mer ink. For devices of the above structure, the M-ZnOelectron-transport layers yielded slightly better deviceperformances than the A-ZnO layers when spin-coatingthe PBDB-T:IT-4F active layers (champion PCEs of12.81% and 12.40% for 0.12 cm2 devices, respectively),whereas this improvement was more pronounced whenblade-coating the active layers (champion PCEs of12.88% and 11.74% for 0.12 cm2 devices, respectively).The performance disparity was hypothesized to originatefrom a change in interface charge transport propertiesbetween the active layer and the ZnO layers induced bythe different surface morphologies of these, motivatingfurther studies of this.

Very recently, Pascual-San-José et al published anelaborate study on blade-coated P3HT:NFA deviceswith a range of non-fullerene acceptors in an invertedglass/ITO/ZnO(NP)/D:A/MoO3/Ag architecture [85].

Only the active layer deposition and the deposition ofthe ZnO(NP) electron-transport layer were, however,sought upscaled, and with an optimized efficiency of5.6% for a P3HT:O-IDTBR active layer, the upscalingeffort by Strohm et al [61] described in the abovesection 5.1 remains a stronger contribution on thisspecific system in terms of scalable fabrication. Thestrength of the study by E Pascual-San-José M Cam-poy-Quiles, and co-workers is, however, that theydevise a high-troughput screening method incorpor-ating variable speed blade-coating, enabling activelayer thickness gradients, as well as a Kofler bench,enabling annealing temperature gradients. This con-tinuous change of processing parameters allowed forthe fabrication ofmore than a thousand devices of area8mm2. In addition, an exemplary stability study wasperformed on P3HT:O-IDTBR devices, yielding thedependence of efficiency on encapsulation and activelayer thicknesses. A degradation to 80% of initial per-formance was linearly extrapolated to >5 years forencapsulated devices with thin (80 nm) active layers, ascompared to 8300 h for encapsulated devices withthick (250 nm) active layers, substantially longer thanthe 700 h and 120 h, respectively, for the corresp-onding unencapsulated devices (all based onmeasure-ments up to 3000 h). Interestingly, light-beaminduced-current measurements suggested that P3HTsuffered from a lower degradation rate thanO-IDTBR,meaning that further optimization of IDTBR micro-phases could lead to better stabilities.

As a last work in this section, the 2019 study by DCorzo, D. Baran, et al on digital inkjet printed active lay-ers deservesmentioning [63]. Using the P3HT:O-IDTBRmaterial system also mentioned in section 5.1, inverteddevices with a glass/ITO/ZnO(NP)/P3HT:O-IDTBR/MoO3/Ag architecture were fabricated. By optimizingthe rheologies of the ink formulationswith respect to sol-vents and concentrations, a champion PCE of 6.47% fora 0.1 cm2 cell was obtained for a homogeneous, inkjetprinted active layer processed from a 1,2-dichlor-obenzene solution. Increasing the active area to 2 cm2

yielded only a small drop in device performance, sustain-ing a PCE of 6.00% (see figure 8(a)). In order to illustratethe possibility of full two-dimensional patterning usinginkjet printing, a 2.2 cm2 device in the shape of a marineturtle was fabricated with an efficiency of 4.76% (depic-ted in the inset of figure 8(b)), creatively demonstratinghowversatileOPVcustomdesigns canbe.

The works highlighted above contribute withimportant observations on the path towards upscalingorganic solar cell fabrication. First of all, the use ofgreen, non-halogenated solvents are not only bene-ficial for the performance of fullerene-free OPVsdeposited with scalable methods, but also for thedevice stability due to the elimination of the need forprocessing additives—and not least for the environ-mental friendliness of the fabrication itself. Secondly,and strongly linked to the first point, the volatilities ofthe solvents used during deposition of the active layers

14

Flex. Print. Electron. 5 (2020) 014004 A SGertsen et al

are crucial. Depending on the miscibility and crystal-linity of the donor: acceptor couples, tuning the boiling-point of the solvent formulations is probably necessaryto close the efficiency gap seen between spin-coated andscalably deposited active layers, as the slower evapora-tion times for techniques like blade-coating and slot-diecoating could lead to unfavorable aggregation. This callsformore systematic solvent studies and preferably in situmorphological studies refining the, predominantly phe-nomenological, hypotheses based on indirect observa-tions. Finally, we should strive to improve thesmoothness of interfacial layers to promote chargetransport between these and the active layers throughaltering the processing solvent formulations that areoptimized for spin-coated fullerene devices. Surfacemorphology studies of slot-die coated interfacial layersprocessed from different solvents could hence provideessential insight into what seems to be a significant path-way for performance losses when upscaling depositionmethods, especially when combined with studies of theinterface between these and the active layer components.

5.4. Promisingmaterial systems for upscalingThe absence in literature of ITO-, vacuum-, and full-erene-free devices on flexible substrates fabricated usingsolely fully scalable deposition techniques with efficien-cies of more than 2% underlines the need for aconcentrated effort towards this goal. Most materialsystems presented in sections 5.2 and 5.3 show greatpromise for adoption to fully scalable fabrication, and inaddition to these, a number of recent record-breakinglab-scale systems fabricated solely using spin-coatingand vacuum deposition deserve mentioning. In thissection, we will thus review selected works with donor:acceptor pairs that have, as of now, not been used indeviceswith scalably deposited active layers.

Already in the first months, 2019 proved to be anextraordinary year for organic solar cells, particularlyfueled by the synthesis of a novel non-fullerene acceptorby the name of BTPTT-4F (also denoted Y6; see figure 10for structure) [8]. First, Yuan et al reported this synthesis

and demonstrated, at the time, record device efficienciesfor single-junction OPVs of up to 15.7% (certified to14.9%) for lab-scale cells with spin-coated PBDB-TF:BTPTT-4F active layers (PBDB-TF is also denoted PM6;see figure 9 for structure) [8]. Using both a conventionaldevice architecture of glass/ITO/PEDOT:PSS/PBDB-TF:BTPTT-4F/PDINO/Al (see figure 11 for PDINOstructure) and an inverted device architecture ofglass/ITO/ZnO(NP)/PBDB-TF:BTPTT-4F/MoO3/Ag,average PCEs of 15.6% and 15.5%, respectively, bothwith champion PCEs of 15.7%, were reached for0.07 cm2 cells. This equally high performance in invertedarchitectures is crucial, because the inverted structure sig-nificantly improves the long-term stability in ambientconditions relative to conventional architectures [86],which is a prerequisite for sustainable scalability oforganic solar cells. Although their following studies onconventional architectureswere not carried out for inver-ted architectures, they showed interesting tendencies.First of all, it was found that increasing the active layerthickness did not hamper the device efficiencies sig-nificantly: going from 150 to 300 nm yielded a twopercentage point drop in champion PCE from 15.7% to13.6%. Although the open-circuit voltages and the short-circuit currents remained largely unaffected by theincreased thickness, the FF went down from 74.8% to62.3% and was thus the main reason for the efficiencydrop. Most notable, however, was the impressive perfor-mance of additive-free, as-cast devices: using no anneal-ing steps after deposition, an average PCE of 15.2% wasobtained, showcasing the stability of this material systemwithdifferent processing conditions.

Shortly after, Fan et al reported the current effi-ciency record for single-junction OPVs of 16.0%, alsousing the BTPTT-4F non-fullerene acceptor [87]. Aconventional architecture of glass/ITO/PEDOT:PSS/P2F-EHp:BTPTT-4F/PFNDI-Br/Ag was employed(structures for P2F-EHp and PFNDI-Br can be foundin figures 9 and 11, respectively), leading to PCEs of11.1% for additive-free, as-cast devices and, as men-tioned, the record 16.0% using devices processed with

Figure 8. (a)The average efficiency of P3HT:O-IDTBRdevices as a function of device area and (b) J−V characteristics of an inkjetprinted device in the shape of amarine turtle. Reproduced from [63]with permission fromWILEY-VCHVerlag GmbH&Co.KGaA,Weinhim.

15

Flex. Print. Electron. 5 (2020) 014004 A SGertsen et al

a relative amount of 1% dibenzylether solvent addi-tive. Inverted devices were also fabricated, leading to achampion PCE of 13.1% for glass/ITO/ZnO(NP)/P2F-EHp:BTPTT-4F/MoO3/Ag cells, also processedwith 1% dibenzylether. BTPTT-4F is thus indeed apromising non-fullerene acceptor that holds a greatpotential for application in upscaled systems.

In the previous section 5.1, the IDTBR non-full-erene acceptor was introduced, and cells and modulesutilizing this in conjunction with P3HT donor poly-mers were reviewed. However, coupling the IDTBRacceptor with PffBT4T derivatives (see figure 9) hasconsistently yielded efficiencies of around 10%–11%in lab-scale devices [13, 76, 78], making this an attrac-tive material system for testing in scalable fabricationwhen taking the favorable properties into accountsuch as negligible burn-in efficiency losses [13], highopen-circuit voltages above 1 V [76], and high repro-ducibility and lifetime when processed in green sol-vents (non-halogenated hydrocarbons) [78].

Finally, the ITIC family of small-molecule acceptorsshould be mentioned. As reviewed in section 5.2, thePBDB-T:ITIC system has already proven to be very wellperforming in slot-die coated layers, and coupledwith theimpressive results obtained for blade-coated layers as

presented in section 5.3, it is clearly indicated that varia-tions of this material system hold great potential forscalably processed OPVs. Furthermore, in 2018, the effi-ciency record for single-junction organic solar cells washeld by a cell incorporating an active layer based on theflourinated IT-4F acceptor, namely PDTB-EF-T:IT-4F(see figure 9 for donor structure). In a glass/ITO/ZnO(NP)/PDTB-EF-T:IT-4F/MoO3/Ag inverted struc-ture, average PCEs of 14.0%were obtained with a cham-pionPCEof 14.2% (certified to 13.9%) and an impressiveFF of 76% [29], further profiling ITICderivatives as someof the best acceptor candidates for future fullerene-freeOPVs.

With these high-efficiency material systems inmind, alongside the considerations regarding proces-sing when going from spin-coating to scalable deposi-tion described in the previous section, it seems that theactive layers will not be the limiting factors for large-scale fabrication of organic solar cells with efficienciesof 10% or more. Knowing that the intrinsic chargetransport and -transfer properties of the polymerdonors and non-fullerene acceptors indeed facilitatehigh-efficiency active layers, focus can be put on opti-mizing the morphology of slot-die coated inksthrough systematic studies of processing conditions,

Figure 9.Molecular structures of all donor polymersmentioned in this work.

16

Flex. Print. Electron. 5 (2020) 014004 A SGertsen et al

including dual temperature-control and solvent for-mulations. Furthermore, and probably equally impor-tantly, the interfaces between the active layers and theelectron- and hole-transport layers should be opti-mized with regards to smoothness using scalable, con-tinuous deposition techniques.

5.5. Summary of the reviewed devicesIn table 1, the deposition methods and characteristicsof all devices reviewed in the above sections 5.1–5.4 arelisted. This includes the donor and non-fullereneacceptor (D:A) materials, the deposition method ofthe active layer, the processing solvent formulation,

the substrate material(s), the depositionmethod of thetop electrode, the device areas, and their corresp-onding champion PCEs, FF, open-circuit voltages(Voc), and short-circuit currents (Jsc).

The deposition methods are classified using thesmiley-model presented in section 3 by their colorsgreen, yellow, or red to provide a quick overview. Asimilar classification is used for the processing sol-vents: halogenated solvents are marked with yellowand non-halogenated solvents with green as a repre-sentation of their environmental friendliness. Corre-spondingly, the substrates are marked with red forrigid glass substrates, yellow for PET substrates with

Figure 10.Molecular structures of all non-fullerene acceptorsmentioned in this work.

Figure 11. Structures of allmolecular interfacial layermaterialsmentioned in this work.

17

Flex. Print. Electron. 5 (2020) 014004 A SGertsen et al

Table 1.Overview of thematerials and depositionmethods used for the reviewed devices alongside their performance characteristics.

18

Flex.Print.E

lectron.5(2020)014004

ASGertsen

etal

Table 1. (Continued.)

aAverage value of>10 cells; acharacteristicsmeasured over 0.12 cm2 cells; CB: chlorobenzene, CN: 1-chloronaphthalene, CF: chloroform, o-DCB: 1,2-dichlorobenzene, BrA: 4-bromoanisole, o-MA: 2-methylanisole, (Me)THF: (2-methyl)tetrahydrofuran, IP: isopropanol, DBE: dibenzylether.

19

Flex.Print.E

lectron.5(2020)014004

ASGertsen

etal

ITO, and green for ITO-free PET to indicate theirscalability.

6. Conclusions and outlook

The field of organic solar cells has been moving fast inrecent years, and record efficiencies are publishedregularly using new non-fullerene acceptor materials. Inthis perspective, we have sought to identify focus pointsforovercoming the challengeofupscaling the fabricationof organic solar cells based on these non-fullereneacceptors. By categorizing a wide range of depositiontechniques in terms of their compatibility with contin-uous roll-to-roll setups, their material waste, and theirthroughput as either fully scalable, partly scalable, ornon-scalable, the literature on fullerene-free OPVs wasreviewed using these classifications. Although numerousstudies have been published on laboratory-scale devicesfabricated using non-scalable deposition techniques,only a small number have been published on devicesfabricated using fully- or partly scalable depositiontechniques. However, combining the knowledge gainedfrom these few studies allows us to suggest three mainpriorities formeeting the lab-to-fab challenge.

(i) First of all, implementing dual temperature con-trol, meaning that both the ink- and substratetemperatures can be controlled simultaneouslyand independently, for example through the useof heated slot-die coating heads, has shown to bean impressively effective way of optimizing theactive layer morphologies, leading to some of thehighest efficiencies published forflexibleOPVs.

(ii) Secondly, the use of non-halogenated, i.e. ‘green’,solvents for active layer deposition has in severalcases shown to be superior to using halogenatedsolvents. Some of these studies also point to the factthat processing additives, which are common inhalogenated solvent formulations and which mightcause device performance to deteriorate with timeand illumination, canbemade redundantwith greensolvents. Furthermore, tuning the boiling point ofthe active layer solvent formulation is crucial tofacilitate preferential morphology evolution duringevaporation when depositing active layers withscalable techniques. Systematic in situ studies can aidthe interpretationof such studies.

(iii) Finally, the interfacial layers should be optimized forcontinuous deposition techniques. The well-per-forming systems with roll-to-roll deposited activelayers and evaporated top electrodes reviewed insection 5.2 all utilize thermally evaporated MoO3

hole-transport layers. As vacuum deposition, asdiscussed, could very well be connected to highprocessing costs, solution processable formulationsof molybdenum oxide hole-transport layers have

great potential as replacements of thermally evapo-rated MoO3 layers [88], whereas also solutionprocessable molybdenum sulfide hole-transportlayers show promise with performance comparableto PEDOT:PSS-based devices [89]. Very recently,solution processed tungsten sulfide layers have alsoshown great promise as hole-transport layers [90].We thus recommend that improving solutionprocessed interfacial layers is prioritized goingforward, as significant efficiency gains for scalablyfabricated, flexible organic solar cells are expected ifthe qualities of solution processed charge transportlayers can get close to the ones of the evaporated. Inaddition, the interface between the electron-trans-port layer, usually ZnO nanoparticles, and the activelayer has shown to be important too as reviewed insection 5.3. Simply by changing the processingsolvent, a higher smoothness and fewer voids andinhomogeneities can be achieved in spin-coatedZnO layers, in turn leading to relative efficiencyincreases of almost 10% for devices with blade-coated active layers [84]. Studying the surfacemorphology of these interfacial layers with varyingprocessing conditions when deposited using fullyscalablemethods is thus important going forward, asthe optimal conditions might differ significantlyfromthe spin-coatedones.

If these three points are addressed, we are confidentthat the 10-10 goals of 10% efficiency and 10 yearsstability for scalably fabricated organic solar cells canbe reached [1, 7, 24], making sustainable, large-scalefabrication viable. We urge that large-area devices(>1 cm2) fabricated using scalable deposition techni-ques are reported alongside the laboratory-scalechampion devices, preferably accompanied by stabilityanalyses, in order to move towards these goals andidentify promisingmaterial systems for upscaling.

Acknowledgments

We acknowledge financial support from the H2020European Research Council through the SEEWHIConsolidator grant, ERC-2015-CoG-681881.

Conflicts of interest

There are no conflicts of interest to report.

ORCID iDs

Anders SGertsen https://orcid.org/0000-0002-4712-0339Marcial FernándezCastro https://orcid.org/0000-0003-3294-2994

20

Flex. Print. Electron. 5 (2020) 014004 A SGertsen et al

Roar R Søndergaard https://orcid.org/0000-0003-3567-3400JensWAndreasen https://orcid.org/0000-0002-3145-0229

References

[1] Søndergaard R,HöselM,AngmoD, Larsen-Olsen TT andKrebs FC 2012Roll-to-roll fabrication of polymer solar cellsMater. Today 15 36–49

[2] Lucera L, Kubis P, Fecher FW, Bronnbauer C, TurbiezM,Forberich K, Ameri T, Egelhaaf H and Brabec C J 2015Guidelines for closing the efficiency gap between hero solarcells and roll-to-roll printedmodules Energy Technol. 3373–84

[3] GuX et al 2017Roll-to-roll printed large-area all-polymersolar cells with 5% efficiency based on a low crystallinityconjugated polymer blendAdv. EnergyMater. 7 1602742

[4] EspinosaN,HoselM, AngmoD andKrebs FC 2012 Solar cellswith one-day energy payback for the factories of the futureEnergy Environ. Sci. 5 5117–32

[5] Peng J, Lu L andYangH2013Review on life cycle assessmentof energy payback and greenhouse gas emission of solarphotovoltaic systemsRenew. Sustain. Energy Rev. 19 255–74

[6] Emmott C J, Urbina A andNelson J 2012 Environmental andeconomic assessment of ITO-free electrodes for organic solarcells Sol. EnergyMater. Sol. Cells 97 14–21

[7] Gambhir A, Sandwell P andNelson J 2016The future costs ofOPV—Abottom-upmodel ofmaterial andmanufacturingcosts with uncertainty analysis Sol. EnergyMater. Sol. Cells 15649–58

[8] Yuan J et al 2019 Single-junction organic solar cell with over15% efficiency using fused-ring acceptor with electron-deficient core Joule 3 1140-51

[9] LinY andZhanX 2014Non-fullerene acceptors for organicphotovoltaics: an emerging horizonMater. Horiz. 1 470–88

[10] LinY,Wang J, Zhang Z-G, BaiH, Li Y, ZhuD andZhanX 2015An electron acceptor challenging fullerenes for efficientpolymer solar cellsAdv.Mater. 27 1170–4

[11] Cheng P, BaiH, ZawackaNK,Andersen TR, LiuW,Bundgaard E, JørgensenM,ChenH,Krebs FC andZhanX2015Roll-coated fabrication of fullerene-free organic solarcells with improved stabilityAdv. Sci. 2 1500096

[12] GaspariniN et al 2017 Burn-in free nonfullerene-basedorganic solar cellsAdv. EnergyMater. 7 1700770

[13] ChaH et al 2017An efficient, ‘burn in’ free organic solar cellemploying a nonfullerene electron acceptorAdv.Mater. 291701156

[14] ZhangG, Zhao J, ChowPCY, JiangK, Zhang J, Zhu Z,Zhang J,Huang F andYanH2018Nonfullerene acceptormolecules for bulk heterojunction organic solar cellsChem.Rev. 118 3447–507

[15] Cheng P, Li G, ZhanX andYang Y 2018Next-generationorganic photovoltaics based on non-fullerene acceptorsNat.Photon. 12 131–42

[16] DuX,Heumueller T, GruberW,ClassenA,UnruhT, LiN andBrabecC J 2019 Efficient polymer solar cells based on non-fullerene acceptors with potential device lifetime approaching10 years Joule 3 215–26

[17] FanQ et al2018High-performance as-castnonfullerenepolymersolar cellswith thicker active layer and large area exceeding 11%power conversionefficiencyAdv.Mater.30 1704546

[18] WangN, YangW, Li S, ShiM, Lau T-K, LuX, Shikler R,Li C-Z andChenH2019Anon-fullerene acceptor enablesefficient P3HT-based organic solar cells with small voltage lossand thickness insensitivityChin. Chem. Lett. 30 1277–81

[19] Krebs FC,GevorgyanSAandAlstrup J 2009A roll-to-roll processtoflexible polymer solar cells:model studies,manufacture andoperational stability studies J.Mater.Chem. 195442–51

[20] LiN et al 2013Towards 15% energy conversion efficiency: asystematic study of the solution-processed organic tandem

solar cells based on commercially availablematerials EnergyEnviron. Sci. 6 3407–13

[21] Burgués-Ceballos I, StellaM, Lacharmoise P andMartínez-Ferrero E 2014Towards industrialization of polymersolar cells: material processing for upscaling J.Mater. Chem.A2 17711–22

[22] PoR, Bernardi A, Calabrese A, CarboneraC, CorsoG andPellegrinoA 2014 From lab to fab: howmust the polymer solarcellmaterials design change? An industrial perspective EnergyEnviron. Sci. 7 925–43

[23] Min J, Luponosov YN,Cui C, KanB, ChenH,WanX, ChenY,Ponomarenko SA, Li Y andBrabec C J 2017 Evaluation ofelectron donormaterials for solution-processed organic solarcells via a novelfigure ofmeritAdv. EnergyMater. 7 1700465

[24] Azzopardi B, Emmott C JM,UrbinaA, Krebs FC,Mutale J andNelson J 2011Economic assessment of solar electricityproduction fromorganic-based photovoltaicmodules in adomestic environment Energy Environ. Sci. 4 3741–53

[25] BinH et al 2016 11.4% efficiency non-fullerene polymer solarcells with trialkylsilyl substituted 2D-conjugated polymer asdonorNat. Commun. 7 13651

[26] Li S, Ye L, ZhaoW,Zhang S,Mukherjee S, AdeH andHou J2016 Energy-levelmodulation of small-molecule electronacceptors to achieve over 12% efficiency in polymer solar cellsAdv.Mater. 28 9423–9

[27] ZhaoW, Li S, YaoH, Zhang S, Zhang Y, Yang B andHou J 2017Molecular optimization enables over 13% efficiency in organicsolar cells J. Am.Chem. Soc. 139 7148–51

[28] Zhang S,Qin Y, Zhu J andHou J 2018Over 14%efficiency inpolymer solar cells enabled by a chlorinated polymer donorAdv.Mater. 30 1800868

[29] Li S, Ye L, ZhaoW,YanH, Yang B, LiuD, LiW,AdeH andHou J 2018Awide band gap polymerwith a deep highestoccupiedmolecular orbital level enables 14.2% efficiency inpolymer solar cells J. Am. Chem. Soc. 140 7159–67

[30] ZhengZ et al 2018Ahighly efficient non-fullerene organicsolar cell with afill factor over 0.80 enabled by afine-tunedhole-transporting layerAdv.Mater. 30 1801801

[31] Ye L, XiongY, ZhangQ, Li S,WangC, Jiang Z,Hou J,YouWandAdeH2018 Surpassing 10% efficiency benchmarkfor nonfullerene organic solar cells by scalable coating in airfrom single nonhalogenated solventAdv.Mater. 30 1705485

[32] Zhang L, Lin B,HuB, XuX andMaW2018Blade-castnonfullerene organic solar cells in air with excellentmorphology, efficiency, and stabilityAdv.Mater. 30 1800343

[33] NaS-I et al 2019High performance roll-to-roll producedfullerene-free organic photovoltaic devices via temperature-controlled slot die coatingAdv. Funct.Mater. 29 1805825

[34] HuangY-C, ChaH-C,ChenC-Y andTsaoC-S 2017Auniversal roll-to-roll slot-die coating approach towards high-efficiency organic photovoltaics Prog. Photovolt. 25 928–35

[35] Swartwout R,HoerantnerMT andBulovic V 2019 Scalabledepositionmethods for large-area production of perovskitethinfilmsEnergy Environ.Mater. 2 119–43

[36] LiN,McCulloch I andBrabec C J 2018Analyzing theefficiency, stability and cost potential for fullerene-free organicphotovoltaics in onefigure ofmeritEnergy Environ. Sci. 111355–61

[37] Bundgaard E et al 2015Matrix organization andmerit factorevaluation as amethod to address the challenge offinding apolymermaterial for roll coated polymer solar cellsAdv.EnergyMater. 5 1402186

[38] BentenH,Mori D,OhkitaH and Ito S 2016Recent researchprogress of polymer donor/polymer acceptor blend solar cellsJ.Mater. Chem.A 4 5340–65

[39] YanC, Barlow S,Wang Z, YanH, JenAK,Marder S R andZhanX 2018Non-fullerene acceptors for organic solar cellsNat. Rev.Mater. 3 18003

[40] SunH,Chen F andChen Z-K 2019Recent progress on non-fullerene acceptors for organic photovoltaicsMater. Today 2494–118

[41] EspinosaN,Garc-Valverde R,Urbina A andKrebs FC2011Alife cycle analysis of polymer solar cellmodules prepared using

21

Flex. Print. Electron. 5 (2020) 014004 A SGertsen et al

roll-to-rollmethods under ambient conditions Sol. EnergyMater. Sol. Cells 95 1293–302

[42] EspinosaN,Garc-Valverde R,Urbina A, Lenzmann F,ManceauM,AngmoDandKrebs FC2012 Life cycleassessment of ITO-free flexible polymer solar cells prepared byroll-to-roll coating and printing Sol. EnergyMater. Sol. Cells 973–13

[43] TsangMP, SonnemannGWandBassaniDM2016 Life-cycleassessment of cradle-to-grave opportunities andenvironmental impacts of organic photovoltaic solar panelscompared to conventional technologies Sol. EnergyMater. Sol.Cells 156 37–48

[44] Krebs FC 2009 Fabrication and processing of polymer solarcells: a review of printing and coating techniques Sol. EnergyMater. Sol. Cells 93 394–412

[45] Roth B, Søndergaard R andKrebs F 2015Roll-to-roll printingand coating techniques formanufacturing large-area flexibleorganic electronicsHandbook of Flexible Organic Electronics edS Logothetidis (Oxford:Woodhead Publishing) ch 7 pp 171–97

[46] Diao Y et al 2013 Solution coating of large-area organicsemiconductor thin filmswith aligned single-crystallinedomainsNat.Mater. 12 665

[47] Diao Y et al 2015 Flow-enhanced solution printing of all-polymer solar cellsNat. Commun. 6 7955

[48] GuX, ShawL, GuK, ToneyMF andBaoZ 2018Themeniscus-guided deposition of semiconducting polymersNat. Commun.9 534

[49] Song S, Lee KT, KohCW, ShinH,GaoM,WooHY,VakD andKim J Y 2018Hot slot die coating for additive-freefabrication of high performance roll-to-roll processed polymersolar cells Energy Environ. Sci. 11 3248–55

[50] Aziz F and Ismail A 2015 Spray coatingmethods for polymersolar cells fabrication: a reviewMater. Sci. Semicond. Process. 39416–25

[51] AndersenTR et al 2014 Scalable, ambient atmosphere roll-to-rollmanufacture of encapsulated large area,flexible organictandem solar cellmodulesEnergy Environ. Sci. 7 2925–33

[52] Brandt RG, Zhang F, Andersen TR, AngmoD, ShiM,Gurevich L, Krebs FC, Andreasen JW andYuD2016Rollcoated large area ITO- and vacuum-free all organic solar cellsfromdiketopyrrolopyrrole based non-fullerene acceptors withmolecular geometry effectsRSCAdv. 6 41542–50

[53] Carlé J E et al 2017Overcoming the scaling lag for polymersolar cells Joule 1 274–89

[54] Heliatek, ‘AboutHeliatek.’https://heliatek.com/en/heliatek/about-us (Accessed: 30April, 2019)

[55] Destouesse E, TopM, Lamminaho J, RubahnH,Fahlteich J andMadsenM2019 Slot-die processing andencapsulation of non-fullerene based ito-free organic solarcells andmodules Flex. Print. Electron. 4 045004

[56] HeY and Li Y 2011 Fullerene derivative acceptors for highperformance polymer solar cells Phys. Chem. Chem. Phys. 131970–83

[57] LinY et al 2016High-performance electron acceptor withthienyl side chains for organic photovoltaics J. Am.Chem. Soc.138 4955–61

[58] Dai S et al 2017 Fused nonacyclic electron acceptors forefficient polymer solar cells J. Am.Chem. Soc. 139 1336–43

[59] ZhanX, TanZ,Domercq B, AnZ, ZhangX, Barlow S, Li Y,ZhuD,Kippelen B andMarder S R 2007Ahigh-mobilityelectron-transport polymerwith broad absorption and its useinfield-effect transistors and all-polymer solar cells J. Am.Chem. Soc. 129 7246–7

[60] Genene Z,MammoW,Wang E andAnderssonMR2019Recent advances in n-type polymers for all-polymer solar cellsAdv.Mater. 31 1807275

[61] StrohmS,Machui F, Langner S, Kubis P, GaspariniN,SalvadorM,McCulloch I, EgelhaafH-J and BrabecC J 2018P3HT: non-fullerene acceptor based large area, semi-transparent PVmodules with power conversion efficiencies of5%, processed by industrially scalablemethodsEnergy Environ.Sci. 11 2225–34

[62] LinY et al 2018 Printed nonfullerene organic solar cells withthe highest efficiency of 9.5%Adv. EnergyMater. 8 1701942

[63] CorzoD, Almasabi K, Bihar E,Macphee S, Rosas-VillalvaD,GaspariniN, Inal S andBaranD 2019Digital inkjet printing ofhigh-efficiency large-area nonfullerene organic solar cellsAdv.Mater. Technol. 4 1900040

[64] Liu Y, Larsen-Olsen TT, ZhaoX, Andreasen B,Søndergaard RR,HelgesenM,NorrmanK, JørgensenM,Krebs FC andZhanX 2013All polymer photovoltaics: fromsmall inverted devices to large roll-to-roll coated and printedsolar cells Sol. EnergyMater. Sol. Cells 112 157–62

[65] AngmoD andKrebs FC 2013 Flexible ITO-free polymer solarcells J. Appl. Polym. Sci. 129 1–14

[66] Roth B, BenattoGAdR,CorazzaM,Carlé J E,HelgesenM,Gevorgyan SA, JørgensenM, Søndergaard RR andKrebs FC2016 Improving the operational stability of PBDTTTz-4polymer solar cellsmodules by electrodemodificationAdv.Eng.Mater. 18 511–7

[67] ChenM-R, FanC-C, AndersenTR,DamHF, FuW-F,Lin Y-Z, Bundgaard E, Krebs FC, ZhanX-WandChenH-Z2014 Solvent-resistant smallmolecule solar cells by roll-to-rollfabrication via introduction of azide cross-linkable groupSynth.Met. 195 299–305

[68] DamHF andKrebs FC 2012 Simple roll coater with variablecoating and temperature control for printed polymer solarcells Sol. EnergyMater. Sol. Cells 97 191–6

[69] Carlé J E, Andersen TR,HelgesenM, Bundgaard E,JørgensenMandKrebs FC 2013A laboratory scale approachto polymer solar cells using one coating/printingmachine,flexible substrates, no ITO, no vacuum and no spincoating Sol.EnergyMater. Sol. Cells 108 126–8

[70] Cheng P, Lin Y, ZawackaNK,Andersen TR, LiuW,Bundgaard E, JørgensenM,ChenH,Krebs FC andZhanX2014Comparison of additive amount used in spin-coated androll-coated organic solar cells J.Mater. Chem.A 2 19542–9

[71] LiuK, Larsen-Olsen TT, Lin Y, BeliatisM, Bundgaard E,JørgensenM,Krebs FC andZhanX 2016Roll-coatingfabrication offlexible organic solar cells: comparison offullerene and fullerene-free systems J.Mater. Chem.A 41044–51

[72] Rossander LH, ZawackaNK,DamHF,Krebs FC andAndreasen JW2014 In situmonitoring of structure formationin the active layer of polymer solar cells during roll-to-rollcoatingAIPAdv. 4 087105

[73] ZawackaNK,Andersen TR, Andreasen JW, Rossander LH,DamHF, JorgensenMandKrebs FC2014The influence ofadditives on themorphology and stability of roll-to-rollprocessed polymer solar cells studied through ex situ andin situX-ray scattering J.Mater. Chem.A 2 18644–54

[74] Richter L J, DeLongchampDMandAmassianA 2017Morphology development in solution-processed functionalorganic blendfilms: an in situ viewpointChem. Rev. 1176332–66

[75] Rossander LH,DamHF,Carle J E,HelgesenM, Rajkovic I,CorazzaM,Krebs FC andAndreasen JW2017 In-line, roll-to-rollmorphology analysis of organic solar cell active layersEnergy Environ. Sci. 10 2411–9

[76] BaranD et al 2016Reduced voltage losses yield 10%efficientfullerene free organic solar cells with>1Vopen circuitvoltagesEnergy Environ. Sci. 9 3783–93

[77] BaranD et al 2017Reducing the efficiency-stability-cost gap oforganic photovoltaics with highly efficient and stable smallmolecule acceptor ternary solar cellsNat.Mater. 16 363–70

[78] WadsworthA et al 2017Highly efficient and reproduciblenonfullerene solar cells fromhydrocarbon solventsACS EnergyLett. 2 1494–500

[79] Ye L et al 2016High performance organic solar cells processedby blade coating in air from a benign food additive solutionChem.Mater. 28 7451–8

[80] Lin B et al 2019Molecular packing control enables excellentperformance andmechanical property of blade-cast all-polymer solar cellsNano Energy 59 277–84

22

Flex. Print. Electron. 5 (2020) 014004 A SGertsen et al

[81] ZhaoW, Zhang S, Zhang Y, Li S, LiuX,HeC, Zheng Z andHou J 2018 Environmentally friendly solvent-processedorganic solar cells that are highly efficient and adaptable for theblade-coatingmethodAdv.Mater. 30 1

[82] KimW,Kim JK,KimE, AhnTK,WangDHandPark JH2015Conflicted effects of a solvent additive on ptb7:pc71bmbulk heterojunction solar cells J. Phys. Chem.C 119 5954–61

[83] Jacobs I E,Wang F, Bedolla Valdez Z I, AyalaOviedoAN,BilskyD J and Moulé A J 2018 Photoinduced degradationfrom trace 1, 8-diiodooctane in organic photovoltaics J.Mater.Chem.C 6 219–25

[84] Ji G, ZhaoW,Wei J, Yan L,HanY, LuoQ, Yang S,Hou J andMaC-Q2019 12.88% efficiency in doctor-blade coatedorganic solar cells through optimizing the surfacemorphologyof a ZnO cathode buffer layer J.Mater. Chem.A 7 212–20

[85] Pascual-San-José E, Rodrez-Martínez X, Adel-AbdelaleimR,StellaM,Martínez-Ferrero E andCampoy-QuilesM2019Blade coated P3HT:non-fullerene acceptor solar cells: a

high-throughput parameter studywith a focus on up-scalability J.Mater. Chem.A 7 20369–82

[86] TanM J, Zhong S, Li J, Chen Z andChenW2013Air-stableefficient inverted polymer solar cells using solution-processednanocrystalline ZnO interfacial layerACSAppl.Mater.Interfaces 5 4696–701

[87] FanB, ZhangD, LiM, ZhongW, Zeng Z, Ying L,Huang F andCaoY 2019Achieving over 16% efficiency for single-junctionorganic solar cells Sci. ChinaChem. 62 746–52

[88] YiQ et al 2015Aqueous solution-depositedmolybdenumoxidefilms as an anode interfacial layer for organic solar cellsACSAppl.Mater. Interfaces 7 18218–24

[89] Wei J, Yin Z, Chen S-C, CaiD andZhengQ 2016 Solution-processedMoSx thin-films as hole-transport layers for efficientpolymer solar cellsRSCAdv. 6 39137–43

[90] LinY et al 2019 17% efficient organic solar cells based on liquidexfoliatedWS2 as a replacement for PEDOT:PSSAdv.Mater.31 1902965

23

Flex. Print. Electron. 5 (2020) 014004 A SGertsen et al