Review

Printed Diodes www.advancedscience.com

Printed Diodes: Materials Processing, Fabrication,

and Applications
Yihang Chu, Chunqi Qian, Premjeet Chahal, and Changyong Cao*

by both commercial demand as well as

Printing techniques for the fabrication of diodes have received increasing potential for improvement.[1–4] The widely
attention over the last decade due to their great potential as alternatives for utilized ability of diodes to control the uni-
high-throughput and cost-effective manufacturing approaches compatible directional flow of current originates from
a structural junction permitting the flow
with both flexible and rigid substrates. Here, the progress achieved and
of current in one direction, but not the
the challenges faced in the fabrication of printed diodes are discussed and other.[5,6] Their usage in voltage regulation
highlighted, with a focus on the materials of significance (silicon, metal and voltage surge protection pushed devel-
oxides, nanomaterials, and organics), the techniques utilized for ink deposition opment toward alteration of the reverse
(gravure printing, screen printing, inkjet printing, aerosol jet printing, etc.), breakdown voltage, while the discovery of
and the process through which the printed layers of diode are sintered after light-emitting diodes (LEDs) led to the
opening of an entire dedicated subfield
printing. Special attention is also given to the device applications within
of research in response to the tremen-
which the printed diodes have been successfully incorporated, particularly dous commercial potential. Amid these
in the fields of rectification, light emission, energy harvesting, and displays. influences, the meteoric rise of rectifiers
Considering the unmatched production scalability of printed diodes and relying on diodes and their use in near-
their intrinsic suitability for flexible and wearable applications, significant field communication and wireless power
transmission technology gave rise to a
improvement in performance and intensive research in development and
comprehensive push toward diodes of
applications of the printed diodes will continuously progress in the future. higher switching speeds capable of radio
frequency (RF) operation. In the 1970s, an
associated surge of interest in the develop-
1. Introduction ment of RF identification (RFID) tags intensified the demand
for flexible electronic components. This trend, when combined
As one of the most ubiquitous electronic components, diodes with the ever-ongoing search for less costly methods of mass
have been intensively studied over the past century, driven production, inevitably propelled research on diodes to the field
of printed electronics.
Among diode fabrication methods, printing is an umbrella
Y. Chu, Prof. C. Cao
Laboratory for Soft Machines & Electronics term incorporating all techniques in which solution-processed
School of Packaging materials are deposited in the form of ink on a substrate. Printed
Michigan State University diode designs generally attempt to minimize the number of
East Lansing, MI 48824, USA layers due to the difficulty of multilayer printing originating
E-mail: ccao@msu.edu
from limitations of solvent processing, and typically are based
Y. Chu, Prof. C. Qian, Prof. P. Chahal, Prof. C. Cao
Department of Electrical and Computer Engineering
on vertical structures to reduce demands on lateral printing
Michigan State University resolution.[7] The processing methods of this category generally
East Lansing, MI 48824, USA excel in the scalability, cost-effectiveness, efficiency, and pro-
Prof. C. Qian duction speed that are vital to the widespread use of dependent
Department of Radiology technologies such as RFID tags or flexible LEDs.[1,8] Another
Michigan State University
vital advantage stems from the compatibility of printing with
East Lansing, MI 48824, USA
thin, flexible, and lightweight substrate materials, except in
Prof. C. Cao
Department of Mechanical Engineering cases such as metal oxide semiconductors in which the solu-
Michigan State University tion-processed materials require high temperature annealing
East Lansing, MI 48824, USA which may damage substrates with lower melting points.[9] For
The ORCID identification number(s) for the author(s) of this article ultrahigh frequency (UHF) applications, printing manufac-
can be found under https://doi.org/10.1002/advs.201801653. turing can also potentially bypass component assembly steps by
© 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, directly printing preconnected components.[10,11] Despite such
Weinheim This is an open access article under the terms of the Creative overwhelming advantages, certain drawbacks of printing tech-
Commons Attribution License, which permits use, distribution and re-
production in any medium, provided the original work is properly cited.
niques have limited their use. For instance, the requirement
of solution-processing of needed materials can compromise
DOI: 10.1002/advs.201801653 their electrical properties, especially certain sensitive materials,

Adv. Sci. 2018, 1801653 1801653 (1 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

resulting in lowered device performance. Also, the formation

of excellent metal–semiconductor contacts at room temperature Premjeet Chahal joined the
remains a major challenge. Additionally, due to the difficulty Electrical and Computer
of multilayer printing, a large portion of printed diodes are Engineering Department at
single layer unipolar devices based on a single semiconductor Michigan State University
of either electron or hole conduction properties. in 2009 as an assistant
In this review, we summarize the progress of recent professor, and has been
advances regarding the material processing, fabrication, and an associate professor
applications of printed diodes, with an emphasis placed on since 2015. He received his
the aspects of solution-processing and material performance B.S. and M.S. degrees in
(Figure 1). A brief overview for printed diodes is first pre- electrical engineering from
sented in Section 2 describing the general architectures used Iowa State University and
in printed diodes, as well as the criteria through which they his Ph.D. degree in electrical
are evaluated. Section 3 summaries the major printing pro- engineering from Georgia Institute of Technology in 1999.
cesses for fabricating printed diodes. Afterward, the subse- He was a senior researcher at Raytheon from 1999 to 2006,
quent Section 4 focuses on the primary classes of materials and at Abbott Laboratories from 2006 to 2008. His current
used in printing diodes, with each subsection highlighting the research interests include millimeter-wave and terahertz
notable advances, detailing challenges, and summarizing their electronics, wireless sensors, RF-optical devices, systems
applications. In detail, Section 4.1 is devoted to silicon, the packaging, and additive manufacturing of RF components
dominant material within modern rigid-substrate semicon- and systems.
ductor technology; Section 4.2 focuses on metal oxide semi-
conductors (MOS), a class of materials which has progressed Changyong Cao is currently
greatly over the last decade; Section 4.3 describes a variety of an assistant professor in
nanomaterials, which have demonstrated the greatest elec- the School of Packaging,
trical performance and potential, but also must overcome a Department of Electrical and
few grand challenges for commercially viable applications; Computer Engineering and
Section 4.4 discusses the organic materials widely used Department of Mechanical
in printed diodes, which possess the greatest compatibility Engineering (by courtesy) at
with printing fabrication methods, but suffer from a lack of Michigan State University
high performance and device longevity due to their limited (MSU), where he is directing
material properties. Finally, a brief conclusion and discussion the Laboratory for Soft
for the future of printed diodes is presented in Section 5. Machines and Electronics.
He received his Ph.D. degree
in Mecahnical Engineering and Materials Science from the
Australian National University in 2014, and then worked
as a postdoctoral research associate at Duke University.
His research interests include soft materials and soft
machines, printed electronics, flexible and stretchable
electronics, smart packaging systems as well as the
3D/4D printing of advanced functional materials.

2. Types and Frequency of Diodes

2.1. Schottky Diode

In its simple form, the Schottky diode consists of one layer

of a p-type or n-type semiconductor positioned between two
metal electrode contacts, one of which permitting charge injec-
tion in one bias direction while blocking in the other, and the
other providing low resistance current conduction as an ohmic
contact. Where the rectifying metal electrode and the semi-
conductor are brought together, the originally different Fermi
level of the semiconductor and work function of the metal are
forced to align, theoretically resulting in band bending and the
Figure 1. Outline illustration of the review for printed diodes via different
kinds of materials. The focus of the review is placed on materials creation of the Schottky barrier responsible for rectification
processing, fabrication approaches, and broad applications of the diodes as described by the Schottky–Mott rule.[8] In practical applica-
and their integrated devices. tions, however, the Schottky–Mott rule is not strictly followed

Adv. Sci. 2018, 1801653 1801653 (2 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

by metal–semiconductor interfaces due to the phenomenon comparison to thin-film transistor (TFT)-based alternatives.[19] The
of Fermi level pinning, which originates from the formation use of the negative differential resistance effect for tunnel diodes
of electron states within the bandgap at the termination of the in printed electronics currently remains at the early stage, although
semiconductor material against the rectifying metal contact.[12] multiple hybrid organic–inorganic tunnel diode architectures with
In the unipolar Schottky diode, there is no charge carrier potential for printing fabrication have been reported.[8,20]
depletion region at the metal–semiconductor junction. Thus,
under forward bias, holes can shift without much hinder-
ance from the semiconductor directly into the metal, while 2.3. LEDs
the Schottky barrier blocks hole injection from the metal in
the reverse direction. The lack of a charge carrier depletion region Organic LEDs (OLEDs) are dominating the current printed LEDs
allows for the elimination of the reverse recovery time required field due to their superior contrast ratio, response time, and
in other types of diodes for shifting operation state from for- power efficiency, as well as the unmatched compatibility with
ward bias to reverse bias, resulting in the outstanding switching solution-processing approaches of cheap and lightweight organic
frequency of Schottky diodes. Additionally, as the metal–semi- materials. The most basic single-layered OLED consists of an
conductor barrier is lower in height compared to the junction emissive layer (EML) of either fluorescent or phosphorescent
barriers in p–n junction diodes, a lower activation voltage in organic material sandwiched between two electrodes, the anode
forward bias is required, while yielding a higher forward cur- being transparent and the cathode being reflective or vice versa.[21]
rent density and lower forward bias voltage drop.[1] In return, Once voltage is applied, holes and electrons are injected into the
however, the low barrier height is a disadvantage in reverse bias, organic material layer to form excitons, which become de-excited
being responsible for the higher levels of reverse leakage cur- after the emission of a photon through the transparent electrode.
rent and reduced reverse breakdown voltage which have limited A more efficient architecture requires the addition of a hole-trans-
the Schottky diode to low-voltage applications. These two disad- port layer and an electron-transport layer sandwiching the EML,
vantages can be compensated for architecturally by surrounding and further tuning of emitted light properties such as color or
the diode with semiconductor guard ring structures, which can intensity may require the addition of supplementary emission
guard against leakage current and reverse breakdown by con- layers.[22,23] To combat the short lifespan of organic materials
trolling the breakdown field geometry.[13] The Schottky diode while preserving the light intensity level, the operation current
has emerged as the most popular printed diode design, being is designed to be reduced by stacking two or more OLEDs with
well ahead of most other competitors. This diode architecture a charge generation layer in between to double the number of
has demonstrated both excellent printing compatibility and out- layers,[7] which has inevitably rendered the advancement of mul-
standing switching frequency performance at lower voltages. tilayer printing fabrication, a fiercely pursued research topic.[21,24]
Traditional LEDs formed from inorganic materials usually
have longer lifespans and adequate energy efficiency.[25] How-
2.2. Tunnel Diodes ever, due to the fact that inorganic materials are not compat-
ible with solution-processing, research on printing inorganic
Tunnel diodes have shown the potential for impact in low- LEDs has received less fervor than that of OLEDs. Recently,
energy applications emphasizing durability, in which negative quantum dots (QDs) have emerged as another highly suitable
differential states can be observed when quantum mechanical material category for printed LEDs.[26] QD-LEDs possess many
tunneling occurs through a thin barrier.[14] The initial design of the advantages of OLEDs such as high efficiency, luminosity,
for tunnel diodes took advantage of this with a modified p–n flexibility, and a greater spectrum of visible light, while over-
junction architecture in which the valence band hole states of coming lifespan challenges.[27]
the p-type semiconductor were nearly aligned with the con- The fabrication of LEDs represents a unique challenge to
duction band electron states of the n-type semiconductor, the printed electronics industry. Different types of materials,
either through heavy doping or through employing the natural including organic polymers, organic small molecules, colloi-
electrical properties of the chosen materials. Moreover, the dally synthesized quantum dots, and some inorganic materials,
tunneling mechanism has also been observed and applied to have been explored for boosting the performance of printed
notable effect in metal–oxide–metal and metal–insulator–metal LEDs. However, regardless of material types, the device archi-
(MIM) diode architectures, which rely on the work function tectures of LEDs rely heavily on multiple layers. With the
difference between the narrowly separated metal layers to gen- limitations of solvents, multilayer printing faces difficulties
erate relatively asymmetric I–V characteristics.[15] in aligning different printed layers to the required degree of
As the speed of quantum tunneling far exceeds the speed of accuracy. This shortcoming typically results in devices of lower
conventional drift currents or diffusion currents which rely on power efficiency and greater fragility, a phenomenon which
conduction, tunnel diodes have the potential for unsurpassed has greatly hindered the large-scale commercialization of LEDs
high frequency operation under extremely low bias voltages.[8,16,17] fabricated through printing.[3,7,28]
Tunnel diodes have also demonstrated excellent durability and
longevity as well as superb resistance to ionizing radiation, which
allow them to be used in space and other high-radiation environ- 2.4. Operation Frequency
ments.[18] In addition, MIM tunnel diodes have also been consid-
ered for incorporation in image display technology such as liquid Operation frequency is one of the most important parameters
crystal display screens as switching devices with reduced cost in for evaluating the performance of diodes and for determining

Adv. Sci. 2018, 1801653 1801653 (3 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

charge carrier depletion region, and thus do not require any

reverse recovery time. This advantage is instrumental in estab-
lishing the superiority of Schottky diodes in high frequency
operation. Another parameter is forward voltage drop, which
represents the minimum forward bias voltage at which the
diode begins to conduct.[34] Lower forward voltage drop brings
a positive influence on operation frequency, while also leading
to higher reverse leakage current and lower reverse breakdown
voltages under reverse bias, which may become an issue to the
relevant diode architectures, especially for Schottky diodes.
A key measure of diode frequency of operation is the cut-off
frequency, above which the nonlinear characteristics of the diode
are significantly suppressed due to the associated parasitics
(resistance, capacitance, and inductance).[17,35] To maximize the
cut-off frequency, the effective series resistance which may arise
from poor semiconductor–electrode contact as well as resistivity
Figure 2. Variation of device frequency characteristics and charge carrier of the metal and the semiconductor should be kept as low as
mobility for different kinds of semiconducting materials. Reproduced possible (ideally <1 Ω). In this regard, the fabrication of elec-
with permission.[1] Copyright 2017, IOP Publishing.
trodes becomes of particular importance, because the quality of
the semiconductor–electrode junction alongside the associated
their applications in different fields such as rectification, signal resistance may greatly affect the device performance. Ideally,
modulation, and flexible electronics.[2,29] High frequency (HF) diodes with low turn-on voltage, high cut-off frequency, and low
operation (3–30 MHz) and UHF operation (>30 MHz) have leakage are desired for RFID applications where power dissipa-
been explored intensively in the near-field communication and tion is a critical design challenge. Also requiring minimization
energy transmission/harvesting industries.[1] Being a vital vari- is the diode capacitance, which largely manifests due to the
able of RFID systems, the diode switching frequency, i.e., the depletion region and the spacing between the metal electrodes.
rate at which diodes can shift between forward and reverse In the efforts to reduce the capacitance and effective resist-
bias, typically becomes the limiting factor barring rectifiers ance, two diode architectures have risen to prominence. The
from higher operation frequencies.[1,30] vertical sandwich-like diode architecture, consisting of neatly
Multiple parameters determine the switching frequency of stacked layers, emphasizes the minimization of layer thickness to
diodes, mainly stemming from the characteristics of materials reduce the effective resistance, but it shows a diode capacitance
used as well as the diode structure chosen. First, the two primary that scales with diode area, thus limiting the area-dependent
material-dependent parameters are the charge carrier mobility, current driving capabilities in return for excellent frequency
determining how quickly charge carriers can move through the performance.[36] For some applications in which the effective
material under the influence of an electric field, and charge carrier impedance of the diodes must be tailored for ease of impedance
concentration, referring to the concentration of charge carriers matching to an external circuit or antennas, the manipulation of
within the material per unit of volume. As seen in Figure 2, the dif- diode area for the vertical diode architecture represents an easier
ferent materials possess an array of charge carrier mobility ranges and more realistic route. On the other hand, the coplanar diode
which determine and limit their maximum potential switching architecture is formed by depositing semiconductor material in
frequencies when used in diodes, with carbon nanotubes (CNTs) the gap between two electrodes, using the interelectrode distance
and graphene demonstrating the highest value range, and organic to define the semiconductor channel length.[37] Once fabricated,
polymers displaying the lowest.[8,31] In the research involving the the coplanar diode architecture exhibits ultralow effective resist-
solution-processing of semiconductor materials, much effort is ance and capacitance values which generally result in superior
placed on the preservation of high charge carrier mobility through operating frequencies and lower turn-on voltages. However, the
improving the fabrication process, in particular for organic mate- implementation of this structure, especially for Schottky diodes is
rials.[32] On the other hand, charge carrier concentration depends extremely challenging due to the need for fabricating two asym-
less on the natural performance of the material, and can be arti- metric electrodes within 10–20 nm of distance. This requirement
ficially enhanced during the fabrication process through doping has barred such architecture from any form of large-scale use
with many inorganic semiconductor materials.[8,28] even employing the adhesion lithography technique.[36–39]
Alongside material considerations, the diode structures
chosen are also critical in determining their frequency perfor-
mance. The most vital parameter dependent on diode structure 3. Printing Techniques for Diodes
is the reverse recovery time, which represents the time required
for the diode to shift from the conducting to the noncon- Among viable fabrication techniques for electronics, printing
ducting state following a voltage shift. For p–n junction diodes, approaches have received significant attention from both
a charge carrier depletion region forms at the junction under academia and industry due to their ability to bypass the main-
bias, and must require a reverse recovery time ranging from stream rigid and expensive silicon-based electronics to directly
several microseconds to less than 100 ns to disappear before deposit preconnected device structures on flexible substrates in
switching.[33] By contrast, Schottky diodes do not possess a an efficient, scalable, and cost-effective manner.[1,8,9,27] There

Adv. Sci. 2018, 1801653 1801653 (4 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

fabrication emphasizing scalability, the roll-to-roll (R2R) tech-

niques of gravure, flexographic, and rotary screen printing are
ideal. However, when superior printing resolution is required,
inkjet printing, screen printing, aerosol jet printing, and gravure
printing stand out. A comparison between these printing tech-
niques is summarized in Figure 3, including the printing resolu-
tion, printing speed, minimum layer thickness, and required ink
properties. We briefly discuss the popular printing approaches
used in fabricating printed diodes as below in this section.

3.1. Inkjet Printing

As illustrated in Figure 4a, inkjet printing relies on the ejection

of droplets of ink from a nozzle onto the substrate, which can be
either rigid or flexible. Two different methods of drop-on-demand
ink ejection exist: 1) the piezoelectric method, in which a piezo-
electric ceramic tile is used to generate pressure to force drops of
Figure 3. Phase-diagram of the printing resolutions and speeds demon- ink from a reservoir near the nozzle, and 2) the thermal method,
strated by the major printing techniques. For each method, the minimum in which thermal excitation from a heating element is used to
layer thickness and required ink viscosity are presented. Reproduced with cause rapid vaporization of the ink to form a bubble, resulting in
permission.[27] Copyright 2017, Royal Society of Chemistry. a rapid pressure increase capable of propelling a droplet of ink out
through the nozzle.[27,43,44] Typically, the inks for this technique
are two primary challenges within the process of printing, i.e., are formulated with viscosities on the order of 1–20 centipoise,
the solution-processing of electronics materials into inks for similar to that of water, to maintain good jettability of the inks.
selected printing fabrication methods[1,2,27] and the fabrication As a noncontact printing method, inkjet printing can perform
of multilayer structures with high resolution, good alignment, moderately detailed patterning without the use of any mask.
and favorable compatibility between layers.[7,21,40,41] Compared to most other printing techniques, inkjet printing pos-
Printing fabrication methods can be generally categorized into sesses one of the lowest percentages of wasted materials during
sheet-based and roll-by-roll-based approaches.[42] For high volume fabrication as well as low initial startup costs, thus rendering it

Figure 4. Schematic illustration of the most popular printing techniques. a) Inkjet printing. b) Screen printing. Reproduced with permission.[27]
Copyright 2017, Royal Society of Chemistry. c) Gravure printing. Reproduced with permission.[42] Copyright 2016, IOP Publishing. d) Aerosol jet printing.
Reproduced with permission.[191] Copyright 2017, IOP Publishing.

Adv. Sci. 2018, 1801653 1801653 (5 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

the easiest printing tool to employ when convenience is the first the aerosol mist for jetting onto the target substrate (Figure 4d).
priority.[45] However, due to the limitation of the printing tech- This approach can push the printing precision down to the level
nique, inkjet printers cannot handle highly viscous inks and high of ≈10 µm, and can manufacture layers ranging from tens of
aspect ratio particle inks, such as organic dielectrics and CNTs, nanometers to a few micrometers in thickness.[51] Moreover,
which may lead to severe nozzle clogging.[43,46,47] Though excel- this technique is compatible with a wide range of materials,
ling in neither printing resolution nor scalability, inkjet printing including high-viscosity inorganic inks, organics, and high-
presents a balance between these parameters, a technique most aspect-ratio CNT solutions.[27,52] Another advantage of aerosol jet
suitable for industrial production, definitively surpassed only by printing is its unique ability to print patterns on nonflat (3D) sur-
gravure printing.[43,44] Inkjet printing has been widely used for faces because of the large standoff distance between the printing
printing conductive electrodes with nanoparticle inks.[48] nozzle and the substrate.[27] In addition, the platen stage of the
aerosol jet printer can be heated up to 200 °C for facilitating the
evaporation of inks to achieve better printing quality, and avoid
3.2. Screen Printing the coffee ring problem frequently suffered in inkjet printing.
Recently, it has gained significant interest in the printing of
Screen printing is a technique in which a fine patterned mesh of transistors[52,53] as well as high frequency circuits and systems.[54]
either porous nylon textile or metal stretching over a rigid frame
is used to deliver ink on a substrate, as seen in Figure 4b.[27]
The screen is placed above a substrate and ink is applied to 3.5. Extrusion-Based 3D Printing
it. With the assistance of a squeegee, the mesh is pressed to
the substrate along a line of contact, allowing ink to be trans- The extrusion-based 3D printing technique has been demon-
ferred to the substrate in accordance to the patterns of the mesh strated in the multilayer fabrication of diodes.[41,55] In fused
along the mesh openings.[49] Despite seeing less attention than deposition modeling, the model layers are produced through
inkjet printing, screen printing is noted for the capability to the extrusion of small beads of material from a moving print-
produce thick, patterned layers from materials of high viscosity head, typically heated, which harden immediately to form the
while maintaining throughput and resolution of approximately structure.[41] Unlike other printing techniques which excel in
the same order of magnitude as inkjet printing.[50] The greatest scalability such as inkjet or gravure printing, 3D printing has
strength of screen printing is its versatility, applicable for low throughput and high operating costs.[23,28] In particular,
solution-processed inorganic and organic materials of all viscos- the necessity of interweaving multiple materials into the fab-
ities regardless of layer function or substrate flexibility. rication process is still a major challenge, for which careful
selection of the printing technique and material categories is
required. Despite such issues, 3D printing represents one pos-
3.3. Gravure Printing sible method of overcoming the nearly ubiquitous difficulty
met by printing techniques in multilayer design fabrication.
Gravure printing in particular has received intensive attention The research on the 3D printing of diodes has only recently
due to its excellent scalability and competitive resolution com- begun to appear, and thus the drawbacks and advantages of this
pared with inkjet and screen printing.[27] Employing a printing field, as well as the ideal fabrication methods and most suitable
process like offset printing and flexography, gravure printing material combinations, remain relatively unexplored.
requires the desired image be engraved or loaded on a cylin-
drical roller, which is used to directly transfer ink through
contact pressure to rolls of substrate, as shown in Figure 4c. In 3.6. Postprocessing of Printed Materials
return for unmatched large-scale production and cost-effective-
ness in the long run, alongside a lack of requirements on the After the deposition of solution-processed materials, especially
flexibility of the substrate or properties of the ink, the cylindrical in the case of inorganic materials, thermal annealing becomes
rollers and machinery costs render gravure printing costly to an extremely important step to bringing out the maximum
be employed in the initial phases, and far less accommodating performance of the resulting device.[1,53] The function of the
to design changes in comparison to other printing fabrication annealing is mainly twofold: one is to evaporate the solvent used
methods.[42] Combined with a R2R printing configuration, in solution-processing alongside other impurities to ensure ideal
gravure printing is extremely suitable for high-volume produc- layer quality, and the other is to partially or fully melt the sepa-
tion, and considered ideal for industrial purposes.[49] rated nanoparticles together to get better conductivity or chemi-
cally convert the deposited precursor to a stable semiconductor
layer via thermal energy. Currently there are two major thermal
3.4. Aerosol Jet Printing annealing methods: i) conventional thermal annealing via
furnace, oven, or hotplate;[52,53,56] ii) photothermal annealing via
Aerosol jet printing is a relatively new printing technique laser or flash lamp.[57–59] The first annealing approach has been
developed in recent years for noncontact printing of flexible widely used for most printing applications like transistors or
electronics. In this process, liquid inks are first atomized into sensors with printed electrodes from Ag/Au nanoparticle inks
small droplets of ≈1–5 µm in diameter, then transported to the due to the low-cost and simplicity in implementation of such
printing head through a carrier gas. At the printing head, an a method.[52] However, the drawback of this method is that in
annualer sheath flow is further used to aerodynamically focue multiple layered devices, the curing of newly printed layers will

Adv. Sci. 2018, 1801653 1801653 (6 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

lead to repeated curing of the former printed layers multiple 4. Printing Materials and Their Applications
times and even with different temperatures, which may gene­ for Diodes
rate negative effects on the affected layers as well as the device’s
performance, particularly when certain organic materials are 4.1. Silicon
involved.[60] More significantly, some flexible substrates may
not be able to survive at high curing temperatures, limiting the Single crystalline silicon is the material currently dominating
possible application of this general thermal annealing method. the cutting edge of rigid RF electronics, possessing undoped
Another annealing approach, termed photothermal electron and hole mobilities of 1400 and 450 cm2 V−1 s−1,
annealing, relies on light to directly deliver thermal energy to respectively.[62] However, it is difficult to utilize in flexible
the necessary layer.[57–59] The most common method is to use a electronics due to the challenges of the preservation of its
focused laser beam for the sintering. The highly localized energy performance without severe degradation of scalability. In the
from the laser can be controlled more accurately through adjust- fabrication of diodes, the transference of single crystalline Si
ment of the environmental gases and the laser parameters, nanomembranes (SiNMs) to flexible substrates has yielded
minimizing the negative impact to the target substrate and sur- superior results performance-wise in comparison to solution-
rounding areas.[57] Nevertheless, this method usually has a lower processing methods.[1,27] For example, Qin and co-workers have
throughput due to the time-consuming sequential process.[58] successfully employed transferrable SiNMs to fabricate flex-
Recently, UV light is also employed for the sintering of printed ible microwave P-type-Intrinsic-N-type (PIN) diodes capable
electronics. The short-term UV light pulses emitted from xenon of high-frequency response at above 10 GHz,[63,64] as shown
lamps can be tuned through the pulse duration, flash number, in Figure 5a,b. More recently, Seo et al. further simplified the
and power output parameters to maximize the thermal energy fabrication process by using nanoimprint lithography (NIL)
output to the target layer (>800 °C) while minimizing impact to to make flexible RF electronics with finely defined patterns,
the substrate.[61] The xenon lamps can be combined with roll- thereby potentially broadening RF applications (Figure 5c).[65]
to-roll based processes, resulting in large-area processability and The RF TFTs fabricated with NIL patterned deep sub-micro-
high efficiency.[59] Due to the cost-effective and environmentally scale channel lengths and flexible Si NM on plastic substrates
friendly merits of photothermal annealing method, it has been demonstrated a record-breaking 38 GHz maximum oscilla-
attracting more attention and applications in the past few years. tion frequency.[65] They further showed a 5-stage ring oscillator

Figure 5. Printed diodes from silicon materials. a) Optical image of the finished PIN diodes arrays on a bent PET substrate. b) Microscopic image of the
80 mm2 flexible microwave single-crystalline SiNM PIN diode on plastic substrate (top) and Microscopic image of the 240 mm2-PIN diode on plastic
substrate (bottom). Reproduced with permission.[64] Copyright 2011, Elsevier. c) Schematic illustration of fabrication process for nanotrench Si NM
flexible RF TFTs by NIL. d) i) A microscope image of a bent array of TFTs and ring oscillators on a PET substrate. ii) A microscopic image of a single
5-stage ring oscillator under a flat condition. e) Measured voltage–time characteristic of the 5-stage ring oscillator showing a frequency of 165 MHz
and a delay time of 0.59 ns. Reproduced with permission.[65] Copyright 2016, Springer Nature Group.

Adv. Sci. 2018, 1801653 1801653 (7 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

built on poly(ethylene terephthalate) (PET) with TFTs of The semiconducting film is highly reconfigurable, being easily
200 nm trench width and 20 µm channel width (Figure 5d), delaminated from the PET/Al substrate for transference to
which exhibited an oscillation frequency of 169 MHz and other substrates, and the entire fabrication procedure, as shown
corresponding stage delay of 0.59 ns, respectively (Figure 5e). in Figure 6g, possesses excellent potential for mass production
However, the transference of SiNMs typically require highly due to a lack of temperature constraints while only requiring
complex steps such as ion implantation, photolithography, a calender machine for pressing and laminating the different
dry etching, and HF etching, as well as high temperature or film layers together.
high vacuum conditions. The lack of cost-effectiveness and
throughput has rendered SiNM-based devices incapable of
filling the void for low-cost flexible RF electronics. Thus, great 4.1.2. Solution-Processed Silicon
desire is invested in the development of new fabrication tech-
niques offering higher scalability while compromising on per- Silicon is extremely difficult to utilize in solution processing
formance, overlapping with the field of printed electronics. A due to the challenge of maintaining the superb electrical
general summary of the important performance parameters for properties of single crystalline silicon in a solution form.
the fabricated devices can be seen in Table 1. The prevalent use of silane-based liquid as a precursor for
conversion to high purity amorphous silicon (a-Si) or micro-
crystalline/nanocrystalline silicon (µc-Si/nc-Si) has limited
4.1.1. Silicon Particles compatible precursors to hydrogenated silicon compounds
of the straight-chain (SinH2n+2) or cyclic forms (SinH2n).[71]
Silicon microparticles, defined as particles of Si between Moreover, the pyrolysis procedure in the process requires
0.1 and 100 µm in diameter, have been used for developing high temperature, resulting in incompatibility with most flex-
printed solar cells in the last decade with the dispersion of Si ible substrates, and the production of harmful monosilane
suspension inks.[66,67] However, the exploration of Si micropar- gases. Even then, solution-processed silicon has limited car-
ticles for printed diodes has recently attracted a great deal of rier mobilities and faces great difficulty in achieving opera-
interest.[1,68,69] In particular, Sani et al. demonstrated a printed tion at the UHF band.[71] However, unlike organic materials,
rectifier incorporating Schottky diodes capable of operation at solution-processed silicon can experience enhancement in
1.6 GHz (Figure 6a–c).[68] In their fabrication, a 4 inch n-doped semiconductor properties before and after layer formation
Si wafer with a doping level of 1018 cm−3 was processed through by doping, while silicon possesses longer lifetime and negli-
ball milling into microparticles 0.7 µm in diameter on average. gible degradation, which may render it an available option for
As shown in Figure 6a, these particles were bonded in an printed RF diodes.[67,72]
organic polymer binder (SU8) to ensure charge carriers would The procedures for converting the hydrosilane precursor
only flow through a few individual particles, thus eliminating compound to a printed silicon layer involve four steps:
electron mobility losses due to electron charge scattering. The polymerization, printing, pyrolysis, and optional annealing. Sil-
resulting 4–5 µm layer was stacked with an additional con- icon ink is composed of a chain-like polydihydrosilane and an
ducting buffer layer of similar-size NbSi2 particles in the same organic solvent such as toluene.[73] To convert the hydrosilane
polymer binder, an oxide-free conduction layer of carbon to precursor to polydihydrosilane, UV light irradiation or sonifica-
introduce current rectification, as well as two electrode layers tion can be applied to the precursor to trigger the ring-opening
on the top and bottom on a PET substrate to form the structure polymerization.[74] The resulting hydrogenated polysilanes are
PET/Al (50 nm)/Si&SU8 (cross-linked, <5 µm)/NbSi2&SU8 diluted once more with the monomer precursor and an organic
(<5 µm)/carbon (6 µm)/Ag (3 µm), as shown in Figure 6a. solvent, then filtered to remove the insoluble polysilanes that
An additional Al-foil antenna structure as well as an electro- precipitated in the dilution process to obtain “liquid Si.”[71]
chromic display based on the poly(3,4-ethylenedioxythiophene): Following ring-opening polymerization and the formation of
polystyrene sulfonate (PEDOT:PSS) polymer were included on ink, a printing process such as inkjet or screen printing can be
the PET substrate (Figure 6b), and the resulting e-label dem- used to deposit a layer of the ink, which would stabilize into a
onstrated remote signal and power transfer in concert with a polydihydrosilane film coated in a thin SiO2 layer due to oxida-
smart phone successfully (Figure 6c). This success directly tion.[72] Finally, the polydihydrosilane layer can be transformed
elevated Si microparticles to the most promising material in into pure amorphous silicon through the process of pyrolysis,
the fabrication of Si flexible electronics, possessing a balance in which heat is applied to induce thermal decomposition. If
of adequate performance and suitability for commercial mass conversion from a-Si to µc-Si is desired, additional thermal
production. and laser annealing steps may be employed, while ambient
More recently, the researchers from the same group pre- hydrogen plasma treatment can further enhance the Si layer’s
sented a simplified fabrication procedure capable of producing electronic properties.[71,75]
Schottky diodes of up to 1.8 GHz (Figure 6d–g).[70] In place Among the potential hydrosilane compounds, cyclopen-
of SU8, nano fibrillated cellulose (NFC) and glycerol were tasilane and cyclohexasilane have been identified as suitable
mixed with Si microparticles and water to form a nanocellu- precursors. Both possess superior stability in comparison to
lose–silicon composite film (Figure 6e), which was then lami- other hydrosilane compounds, maintaining relatively lower
nated between a prepatterned bottom Al electrode layer and vapor pressure while remaining stable under ambient light,
a conductive carbon tape top contact coated in Ni to form the and cyclopentasilane in particular is noted to possess a high
structure PET/Al/NFC&Si (10–15 µm)/Ni:carbon (Figure 6d). photoreactivity on UV light irradiation.[71,75] Initial work with

Adv. Sci. 2018, 1801653 1801653 (8 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

Table 1. Comparison of the performances of devices made of a variety of semiconducting materials reviewed in this paper. Although solution-
processed and printed diodes are given priority, other devices with procedures potentially applicable to printed diode fabrication or useful for
comparison are alos included in the table. Note that in nonrectifying diode devices, the Vin parameter is used to indicate the turn-on voltage.

Device Materials Fabrication method Frequency Vin [V] Vin [V] Mobility [cm2 V−1 s−1] Ref.
TFT Si nanomembrane NIL 5000 – – 155–460 [65]
Schottky diode Si microparticles Screen, inkjet printing 1600 1 – – [68]
Schottky diode Si microparticles Lamination 1800 0.5 – – [70]
TFT, Poly-Si Si precursor cyclopentasilane Spin-coating, inkjet printing – – – 6.5–108 [71]
TFT, a-Si Si precursor cyclohexasilane Spin-coating – – – <1.7 [73]
TFT, µc-Si Si precursor cyclohexasilane Spin-coating – – – – [75]
Schottky diode IGZO Spin-coating – <1.5 – <10 [88]
TFT IGZO Gravure printing – – – 0.81 [86]
TFT IGZO Spin-coating – – – 1.73 [197]
Seven-stage ring oscillators–TFT IGZO Spin-coating 340 – – 7.0 [81]
TFT IGZO Spin-coating – – – 11.2 [198]
TFT IGZO “Sol–gel on chip” – – – 7−12 [60]
TFT IGZO Spin-coating – – – 0.96 [82]
TFT IGZO Spin-coating – – – 2.0 [83]
Rectifier–Schottky nanodiode ZnO a-Lith 20 ±4 1.2 – [38]
Schottky diode IGZO Sputtering 2450 – – – [199]
Rectifier–Schottky diode ZnO Gravure printing 13.56 ±5 4–4.5 – [94]
TFT ZnO Spin-coating – – – 25–45 [90]
TFT ZnO Spin-coating – – – 4–11 [89]
TFT ZnO Spin-coating – – – 6 [91]
TFT SWCNTs Inkjet printing 18.21 – – 6 [133]
Schottky diode SWCNTs DEP 18 0.4 – – [134]
TFT SWCNTs Aerosol jet printing 5 – – – [200]
TFT SWCNTs Aerosol jet printing 5 – – – [201]
Schottky diode SWCNTs Direct growth 540 – – – [107]
Schottky diode Graphene DEP 26 – – – [147]
Schottky Diode–Rectifier Pentacene Thermal evaporation 1240 ±10 3.8 0.11 [168]
TFT transdiode TIPS-pentacene Zone-casting 1–9.8 – – 0.03 [169]
TFT transdiode C10-DNBDT Edge-casting 1–22 ±15 8 3.0 [173]
Schottky Diode–Rectifier C16IDT − BT Spin-coating >>13.56 (<68) ±20 5.1 1.5–2.5 [177]
MIS diode PTAA Inkjet printing – ±10 – – [178]
Schottky Diode–Rectifier PTAA Gravure printing >13.56 ±10 3.5 – [11]
Schottky Diode–Rectifier PQT-12 Spin-coating >13.56 ±10 4 0.1–0.7 [175]
Schottky Diode–Rectifier PTAA Gravure printing 10 ±10 2.7 – [174]

cyclopentasilane was first presented by Shimoda et al.,[71] current density at low voltages in comparison to a-Si PN junc-
in which the solution-processed polycrystalline silicon was tion diodes prepared by chemical vapor deposition (CVD), as
combined with printing for the fabrication of flexible elec- shown in Figure 7c. They found that a-Si could be generated
tronics (Figure 7a,b). Despite having yet to reach the perfor- through collimated aerosol beam deposition in lines as thin as
mance of single crystalline silicon on rigid substrates, the 10 µm.[73] Similarly, Iyer et al. further refined the use of thermal
mobilities exceeded most of those found in solution-processed annealing, laser annealing, and especially plasma hydrogena-
organic TFTs, as well as those of amorphous silicon TFTs.[71] tion treatment following pyrolysis on solution-processed Si for
Meanwhile, Han et al. used a similar fabrication process conversion from a-Si to µc-Si (Figure 7d–f).[75] Their results
with cyclohexasilane to form a Si ink which was converted demonstrated a potential conductivity in spin-coated µc-Si films
to a-Si after layer formation through spin coating.[73] The as- of above 10−1 S cm−1 depending on the duration of H-plasma
fabricated a-Si PN junction diodes demonstrated excellent treatment in the plasma hydrogenation step.[75]

Adv. Sci. 2018, 1801653 1801653 (9 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

Figure 6. Flexible diodes from silicon particles. Schematic illustration of a) the cross section and b) the lateral architecture of the Schottky diode
made of two layers of microparticles (NbSi2 and Si) bonded together by SU-8 binder. c) Demonstration of e-label application. The antenna–diode–
display circuit is deposited onto a PET substrate. When the mobile phone is held close to the antenna, the display starts to switch on. Reproduced
with permission.[68] Copyright 2014, National Academia of Science. d) Illustration for the structure of the modified Schottky diode. e) SEM image of
top NFC:Si film surface. f) Altered response of output DC voltage to signal frequency. g) The fabrication process of the diode: Peeling off the Si film;
Attaching it to the substrate; Calendering; Peeling off the Ni/C double side adhesive tape; Attaching the Ni/C tape to the Si film; Calendering once
more. Reproduced with permission.[70] Copyright 2016, Springer Nature Group.

4.2. Metal Oxide Semiconductors 4.2.1. Indium Gallium Zinc Oxide (IGZO)

MOS are typically n-type, and provide relatively high per- Indium nitrate hydrate (In(NO3)3∙χH2O), gallium nitrate
formance as well as a superior oxygen stability.[1,27,76] The hydrate (Ga(NO3)3∙χH2O), and zinc acetate dehydrate
optoelectronic properties of MOS have been widely used in (C4H6O4Zn∙2H2O) are the most prominent precursors
LEDs, while their transparency has made them extremely selected for synthesizing IGZO with specific ratios due to their
attractive for display-type applications.[28] The chemical higher charge carrier concentrations in associated films.[80–83]
properties of MOS materials have rendered them far more It is discovered that IGZO layers can generally retain amor-
suitable for solution-processing than silicon, although the phous phase structure (without grain boundaries), high
issues regarding the high annealing temperatures do still degrees of transparency, and maximum semiconductor per-
exist. However, the presence of grain boundaries in MOS formance in annealing temperatures of ≈300–400 °C.[78,80]
materials and their low uniformity in printing have limited Therefore, only substrate materials capable of withstanding
their applications in large-scale fabrication.[1] The selected high temperatures like silicon or glass may be used, excluding
fabrication process has a tremendous impact on the final most flexible substrates such as PET and other polymers.[84]
performance and quality of metal oxide layers. Currently, the For example, Kim et al. demonstrated inkjet printing and gra-
most common manufacturing techniques include physical vure printing of IGZO to fabricate TFTs possessing carrier
vapor deposition, pulsed laser deposition, and atomic layer mobilities of 0.03 and 0.81 cm2 V−1 s−1, respectively.[85,86] In the
deposition.[12,77,78] The representative solution-processed, annealing process, the IGZO film and substrate were treated
nonphotonic MOS devices fabricated with approaches that under high temperature to induce the chemical decomposi-
can be potentially applicable for printing diodes are summa- tion and dispersion of unnecessary materials, getting IGZO of
ried in Table 1.[79] high purity. With the same materials, Lim et al. optimized the

Adv. Sci. 2018, 1801653 1801653 (10 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

Figure 7. Printed electronics from solution-processed silicon. a) A TEM image of a solution-processed poly-Si film. The film was formed by spin-coating
and baking of the liquid silicon materials followed by laser crystallization. The inset TEM image highlights the atomic image of the silicon crystal.
The grain size in the film is about 300 nm, which is comparable to that of conventional CVD-formed poly-Si film. b) SEM image of a TFT made from
ink-jet printed silicon film. Reproduced with permission.[71] Copyright 2006, Springer Nature Group. c) Comparison of current density–voltage (JV)
characteristic of printed and plasma-enhanced chemical vapor deposition (PECVD) based a-Si diode. Dashed and star curves denote freshly printed
and air aged printed diodes, respectively. Reproduced with permission.[73] Copyright 2008, Elsevier. d) Schematic of the solution-based synthesis of
microcrystalline silicon (µc-Si) thin films. e) SEM images of a cyclohexasilane precursor-based Si film after plasma hydrogenation treatment for 20 min.
Each scale bar is 100 nm. f) Changes in grain size and roughness of the surface in (e) following plasma hydrogenation treatment of 0, 20, and 30 min
(left to right). Reproduced with permission.[75] Copyright 2015, American Chemical Society.

precursor ratio, and revealed that in an In:Ga:Zn ratio of 1:X:1, the sample at 300 °C, superior to thermally annealed
the ideal gallium proportion was 0.5, through which mobility devices with the same precursor solution (In:Ga:Zn =
as high as 2 cm2 V−1 s−1 could be obtained at an annealing 0.63:0.05:0.32).[87] Banger et al. also developed a “sol–gel
temperature of 400 °C.[83] on chip” hydrolysis approach from soluble metal alkoxide
Remarkably, Kim et al. proposed a room-temperature pro- precursors to form amorphous metal oxide semiconducting
cess to achieve condensation and densification of a-IGZO layers thin-films.[60] They utilized aqueous hydrolysis annealing
through photochemical activation.[81] The samples were heated on a hot plate at maximum process temperatures as low as
up to 150 °C by the transmitted energy of the deep-ultraviolet 230 °C to fabricate highly stable IGO and IGZO TFTs with
rays, which could assist in the removal of organic residues from field-effect mobilities of 7–12 cm2 V−1 s−1, which were com-
the precursor. Using this method, they successfully fabricated parable with those similar TFTs made through sputtering.[60]
a-IGZO TFTs with carrier mobility up to 7 cm2 V−1 s−1 on a flex- The process is applicable to a broad range of amorphous
ible polyarylate (PAR) substrate (Figure 8a).[81] To demonstrate MOSs, and should provide new opportunities for integrating
device scalability, the authors also fabricated seven-stage ring amorphous oxide materials into functional electronic and
oscillator circuits using room-temperature-fabricated IGZO optoelectronic devices.
TFTs on the PAR substrates (Figure 8b). The results showed At present, most flexible MOS-based diodes operating in
that with a supply voltage of VDD = 15 V, an oscillation fre- the HF and UHF range are based on IGZO materials, and
quency larger than ≈340 kHz, and corresponding propagation typically fabricated through sputtering deposition. Rare solu-
delay less than ≈210 ns per stage (Figure 8c,d) could be obtained. tion-processed diodes made from IGZO were reported. One
Recently, Su et al. presented a similar photoannealing recent example was the fabrication of p–n junction diodes via
method to obtain a field-effect mobility of 1.73 cm2 V−1 s−1 solution-processing of IGZO as the n-type semiconductor and
for TFTs by means of ultraviolet-ozone treatment of CuO as the p-type semiconductor.[88] A mixing ratio of 1:1:1

Adv. Sci. 2018, 1801653 1801653 (11 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

Figure 8. Flexible electronics from IGZO. a) Optical image and a schematic cross-section of photoannealed IGZO TFTs and circuits on PAR. b) Optical
image of a seven-stage ring oscillator. Gate to source/drain overlap distance is 5 µm. c) Oscillation frequency (red) and per-stage propagation delay
(blue) of the ring oscillator as a function of supply voltage, VDD. d) Output waveforms of the ring oscillator operating with supply voltages of 5 V (left
panel) and 15 V (right panel), and oscillation frequencies of 45 and 341 kHz, respectively. Reproduced with permission.[81] Copyright 2012, Springer
Nature Group. e) AFM image and f) cross section SEM image of IGZO (In:Ga:Zn = 1:1:1) composition film, annealed at 400 °C for 2 h. g) I–V curves
of IGZO based heterojunction didoes. Reproduced with permission.[88] Copyright 2017, Elsevier.

was selected for In:Ga:Zn, and after spin-coating and curing of on temperature-sensitive polyethylene naphthalate (PEN)
IGZO (Figure 8e,f), an additional CuO layer was added on top substrates with field-effect mobilities of 4–5 cm2 V−1 s−1
of the IGZO layer. For compatibility with the flexible polyimide after annealing in temperatures up to 160 °C.[89] This was
substrate, UV exposure was utilized at room temperature in a achieved by using a carbon-free molecular ZnO hydrate
nitrogen environment after preannealing the IGZO films at a precursor to form high-quality polycrystalline ZnO layers
lower temperature of 325 °C. In such a process, only a slight ≈4–5 nm in thickness (Figure 9c). Further work was per-
increase of roughness was observed without much impact formed to exploit the better electron transport properties of
upon other properties (Figure 8e,f), suggesting UV treatment low-dimensional polycrystalline heterojunctions and quasi-
of the films may be an effective means to lower the processing superlattices (QSLs) through depositing alternating layers of
temperature. Figure 8g depicts the characteristics of a repre- In2O3, Ga2O3, and ZnO of less than 10 nm thick on glass
sentative diode, indicating the turn on voltage (VON) of 1.11 V substrates (Figure 9d) by sequential spin casting of dif-
and the rectification ratio (Ion/off) of 20.[88] ferent precursors in air at low temperatures (180–200 °C).[90]
Optimized prototype QSL transistors exhibited band-like
transport with electron mobilities approximately ten times
4.2.2. Other Metal Oxides greater (25–45 cm2 V−1 s−1) than single oxide devices (typi-
cally 2–5 cm2 V−1 s−1), indicating a promising perspective
The solution-processing of other metal oxides such as ZnO, of this method for application in next-generation large-area
IGO, and In2O3 is similar to that of IGZO. As shown in optoelectronics and large-area microelectronics such as
Figure 9a,b, Lin et al. successfully fabricated ZnO transistors ultrahigh definition optical displays.

Adv. Sci. 2018, 1801653 1801653 (12 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

Figure 9. Flexible transistors and diodes based on solution-processed metal oxides. a) Photograph of an actual transistor array containing 35 flexible
ZnO TFTs based on a UV-grown bilayer AlOx/ZrOx gate dielectric and Al source–drain electrodes. b) Schematic of the device architecture used and
c) TEM image of the SiO2/ZnO/Al cross-section showing the ultrathin nature of the ZnO film. Reproduced with permission.[89] Copyright 2013,
Wiley-VCH. d) Representative sets of transfer characteristics measured from transistors based on metal oxide quasi-superlattices (QSL) channels.
Reproduced with permission.[90] Copyright 2015, Wiley-VCH. e) Optical image of a flexible combustion-processed In2O3 device on AryLite (30 nm Al
gate electrode/41 nm a-alumina dielectric, with 30 nm Al source and drain electrodes) and optical image of an inkjet printed In2O3 line on n++Si/41 nm
a-alumina (inset). f) Comparison of energetics of combustion synthesis-based processes versus conventional approaches. g) schematic of top-contact
bottom-gate TFT device structure in (e). Reproduced with permission.[91] Copyright 2011, Springer Nature Group.

Significantly, Kim et al. introduced a completely divergent epoxy-based ink for the insulating layer on a PET foil sub-
tactic, using a redox-based combustion chemical technique strate (Figure 10a,b).[93] The all R2R gravure printing process
generated through the precursor solution itself to perform the was performed under a roll pressure of 0.8 MPa and a speed of
annealing step (Figure 9e–g).[91] By using acetylacetone or urea as 8 m min−1 at annealing temperatures no more than 150 °C.[93]
fuel and metal nitrates as oxidizers, a localized, self-energy-gener- The high rectifying efficiency of one printed diode and capacitor
ating exothermic reaction can trigger conversion of the precursor can reach about 90% at 13.56 MHz, as shown in Figure 10c,
solution to the MOS layer without excessive spread of heat to indicating almost total conversion of AC signal into DC voltage.
the substrate (Figure 9f). This strategy was successfully demon- Further work by this group successfully incorporated this rect-
strated with In2O3, a-Zn-Sn-O, a-In-Zn-O, and indium tin oxide enna circuit into a wireless sensor-signage tag compatible with
(ITO) to yield a maximum electron mobility of ≈6 cm2 V−1 s−1 R2R gravure printing (Figure 10d) alongside a humidity sensor
with processing temperature of around 200–250 °C.[91] and an electrochromic signage unit on a PET film (Figure 10e),
The printed MOS diodes that can operate at the 13.56 MHz opening wide the gateway to the use of printed diodes in inex-
band have been successfully used in RFID-based applica- pensive HF wireless power transmission.[94]
tions.[92] For example, Park et al. used a R2R gravure printing Most recently, Semple et al. successfully fabricated ZnO-
process to fabricate a 13.56 MHz rectenna with layer-by-layer based coplanar Schottky nanodiodes on glass and PET sub-
printing of ZnO semiconducting ink as the active layer for the strates.[36,37] Utilizing adhesion lithography (a-Lith), they
diode, Ag conductive ink for the antenna and one electrode, achieved electrode spacing less than 20 nm in an unprece-
Al-based conductive ink for the top electrode of the diode, dented lateral architecture (Figure 10f,g).[38] This architecture
BaTiO3-based ink for the dielectric layer of the capacitor, and brings a variety of benefits aside from its low surface area,

Adv. Sci. 2018, 1801653 1801653 (13 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

Figure 10. Printed diodes with metal oxides. a) Optical image of the R2R gravure printed rectenna on PET foil. The bottom schematics describe the
components of rectenna: printed capacitor, printed diode, printed bottom Ag electrode. b) Cross-sectional SEM images for the R2R gravure printed diode.
c) Input–output electrical characteristics of rectifier at 13.56 MHz A, indicating a rectifying efficiency of about 90%. Reproduced with permission.[93]
Copyright 2012, IOP Publishing. d) Scheme for R2R gravure to completely print sensor-Signage Tags on PET foils. e) R2R samples by combining
R2R gravure printed rectenna and R2R coated electrochromic signage. Reproduced with permission.[94] Copyright 2014, Springer Nature Group.
f) Photograph of a 2 × 2 cm2 substrate patterned with the use of the adhesion lithography (a-Lith) technique. The substrate incorporates 72 discrete
Al/Au nanogap diodes. g) Optical micrograph of a-Lithfabricated interdigitated electrode structures. h) Current–voltage characteristics of Al/ZnO/Au
nanogap diodes (with different widths) fabricated on the substrate shown in (f). Reproduced with permission.[192] Copyright 2016, SPIE.

including low device series resistance, negation of scattering summary of the parameters and performance for nonpho-
at ZnO grain boundaries, and stable device performance. tonic devices is presented in Table 1. While quantum dots have
As shown in Figure 10h, despite a large electrode width, the found great success in photonic LED devices, the carbon-based
fabricated coplanar nanogap diodes could operate at low nanomaterials are characterized by exceedingly high semicon-
voltages (±2.5 V) and were able to sustain a high forward cur- ductor properties which no other material class may match.
rent of 10 mA while the reverse current remained low, on the
order of 10 nA.[38]
4.3.1. Carbon Nanotubes

4.3. Nanomaterials Since their discovery in 1991, CNTs have been intensively
studied and explored for nearly every type of electronics
Nanomaterials, such as CNTs, graphene, 2D materials, and due to their superior mechanical, thermal, and electrical
quantum dots, have been receiving increased intention in semi- properties.[95,96] CNTs incorporated within diodes have been
conductor devices including diodes over the years. The unique associated with relatively high breakdown voltages due to
structures of these carbon-based nanomaterials allow them- minimal self-heating,[97] while very little quantum capacitance
selves to outperform all competitors in operation frequency has been observed due to their low dimensionality.[1,98] These
and current densities with near-certainty, although some severe superlative parameters have inevitably elevated CNTs to one of
challenges in fabrication, integration, and scalability remain the most popular and competitive semiconductor materials in
in developing various kinds of devices and applications. A future high-speed electronic devices.

Adv. Sci. 2018, 1801653 1801653 (14 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

Figure 11. Schematic illustration of the major CNT synthesis techniques compatible with solution processing: a) arc-discharge, b) laser ablation, and
c) high-pressure carbon monoxide growth. Reproduced with permission.[105] Copyright 2004, Elsevier. The CNTs purification approaches for removing
the mixed metallic CNTs: d) functionalization via polymer wrapping, e) chromatography and f) density-gradient ultracentrifugation. Reproduced with
permission.[109] Copyright 2008, Springer Nature. The alignment methods for CNT applications in high performance devices: g) selective positioning
via SEM/Raman spectroscopy. Reproduced with permission.[112] Copyright 2009, American Chemical Society. h,i) Alternating current dielectrophoresis
(DEP). Reproduced with permission.[115] Copyright 2014, IEEE.

One grand challenge to hinder the wide applications of electron microscopy (SEM) and Raman spectroscopy to selec-
CNTs is the lack of an economically viable and high-throughput tively position individual SWCNTs with ultrahigh precision for
production technique for CNTs.[99,100] Current fabrication perfect alignment, but this method completely discarded the
techniques include arc-discharge (Figure 11a),[101] chemical hope of scalability or commercial viability.[112] With alternating
vapor deposition,[100,102] laser ablation (Figure 11b),[103] and gas- current dielectrophoresis (AC-DEP), as shown in Figure 11h,[113]
phase catalytic growth by carbon monoxide (Figure 11c),[104] all in conjunction with photolithographic patterning, Li et al.
of which are not ready for large-scale mass production due to reported HF Schottky diodes possessing excellent alignment
the prohibitive costs and substandard efficiency.[96,105] Another (Figure 11i).[114,115] Similarly, Stokes et al. demonstrated the
big obstacle for using CNTs in diodes and other semiconductor successful alignment of ultrahigh-density SWCNT arrays using
devices is the presence of metallic tubes in every fabricated the DEP method, reaching a maximum density of 30 SWCNT
batch.[106–108] In the past decades, a few promising methods per µm for aligned arrays.[116]
have been developed to purify the semiconducting single-walled The dispersion of CNTs within a solvent to form desirable
CNT (SWCNT) for inks used in printing electronics, including ink is a critical step in any solution-based fabrication process.
functionalization via polymer wrapping (Figure 11d), chroma- The ink parameters such as stability, material quality, disper-
tography (Figure 11e), and density-gradient ultracentrifugation sion degree, surface tension, and viscosity must be carefully
(Figure 11f).[106,109,110] However, significant effort is required to optimized with regard to the selected printing method, sub-
enhance the purification throughput and reduce the cost to the strate material, drying procedure, and platen temperature for
levels reasonable for mass production. achieving maximum device performance.[47,117] However, CNTs
Furthermore, the electrical performance of CNTs is greatly possess high interstructure interaction energy due to internal
affected by their alignment due to their typical tubular struc- van der Waals forces, thus rendering dispersion in liquid diffi-
ture. Thus, controlling the alignment of CNT assembly and cult.[118] In addition, the high aspect ratio of CNTs makes them
patterning becomes another difficulty for developing high- prone to clumping and entanglement.[119] Therefore, different
performance CNT-based electronics. “Blind” deposition without methods have been explored for dispersing CNTs into a stable
alignment control was used in a photolithographic route and and uniform solution ink.
confirmed to result in a low yield of functional devices.[111] By Ultrasonication (Figure 12a) has been intensively used to
contrast, as shown in Figure 11g, Jiao et al. utilized scanning disperse CNT bundles within a solvent and evenly spread

Adv. Sci. 2018, 1801653 1801653 (15 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

Figure 12. Methods for dispersing CNTs into a stable and uniform ink. a) Schematic diagram of the sonication procedure for CNTs. b) Plot of
the decrease in CNT conductivity with increased physical mixing time. c) CNT length distribution after 1 h of sonication treatment (inset: SEM
image of lightly damaged CNTs). Reproduced with permission.[193] Copyright 2012, MDPI. d) Schematic of the functionalization procedure for CNTs.
e) Cylindrical miscelle configuration of surfactants encapsulating a SWCNT. f) Hemimicellar adsorption configuration of surfactants on a SWCNT.
g) Random adsorption configuration of surfactants on a SWCNT. h) Adsorption spectra of SWCNTs in an SDS surfactant suspension. Lines A–C
represent SWCNTs of different diameters, while line D represents aggregated SWCNTs bundles. i) TEM cross-sectional image of a SWCNT bundle.
j) Separation procedure through which surfactants may disperse SWCNT bundles. Reproduced with permission.[125] Copyright 2006, Elsevier.

them throughout.[120] Although ultrasonication is cost-effec- of polymers and surfactants have been tested for adsorption
tive, easy to use, and reliable, it can cause damage to internal and bonding on the nanotube surface. Depending on the
CNT structures ranging from defects or deformation in the nature of the surfactant, its concentration, the identity of the
cylindrical formation to the breakage, fragmentation, and short- solvent, the role of the stabilizing polymers, the treatment
ening of CNTs, as shown in Figure 12b,c, negatively impacting of the surface of the substrate, and other parameters such
the CNTs’ performance.[121] In addition, sonication may also as solution temperature, the properties of the resulting CNT
negatively affect the solvents and dispersants in the solution, ink may vary greatly, requiring highly precise tuning to opti-
resulting in alteration of physical and chemical properties mize ink parameters for each device application and printing
together with potential ink instability depending on the incor- method.[126]
porated materials.[122] The deposition of CNT thin films for flexible electronics has
The chemical functionalization of CNTs or the alteration of been performed by a number of different printing methods,
their surface properties for higher solubility through chem- including inkjet printing,[127] aerosol jet printing,[52,53,56,128]
ical moieties is also employed for the dispersion of CNTs.[123] gravure printing,[129] screen printing,[130] and transfer
Despite being far less damaging to the nanotube structures printing.[131] Though CNTs have rarely been used in printed
than physical dispersion methods, high-temperature function- CNT diodes up to now, the performance of other devices has
alization utilizing aggressive chemicals like acids may cause absolutely shown the great promise of this material in this
defects in CNT structures or negatively impact their semi- field.[132] For example, Grubb et al. printed CNT-based TFTs
conductor properties.[124] Adding surfactants is perhaps the on a Kapton substrate by inkjet printing, with very small chan-
most commonly utilized method in solution-dispersed CNT nels (≈1 µm) formed by employing the chemical forces between
inks to increase solubility and reduce clumping (Figure 12d). inks (Figure 13a).[133] High-purity SWCNTs were added to a pro-
Figure 12e–g shows the main mechanism of surfactants in prietary nonaqueous solution possessing similar properties to
dispersing SWCNTs in different solvents.[125] The dispersion N-cyclohexyl-pyrrolidone (CHP) at a ratio of 20% concentration
process can be monitored by checking transient fluorescent by weight, and then dispersed through sonication. After inkjet
emission as a function of various parameters, like the type of printing, a thermal annealing procedure was applied to remove
surfactant used, sonication time, as well as surfactant concen- the proprietary solution. The printed transistors finally were
tration and functionalization (Figure 12h–j).[125] A great variety capable of reaching operation speeds up to 18.21 GHz.[133]

Adv. Sci. 2018, 1801653 1801653 (16 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

Figure 13. CNT-based Schottky diode. a) Illustration of a fabrication approach using the chemical forces between inks. i) Schematic description for
inkjet printing transistors with small channel gaps via chemical forces. ii,iii) optical images of two parallel lines forming a chemical gap. The lines are
electrically isolated. iv) SEM image of an electronically isolated gap. Reproduced with permission.[133] Copyright 2017, Springer Nature. b) Illustration
for a CNT based Schottky diode detector. c) Photomicrograph of a log-periodic antenna with CNT Schottky diodes, d) AFM scan of the CNT channel
region of the diode, and e) close-up view of CNT rich region. e) I–V characteristics for the Schottky diode of (b) alongside its effective resistance.
f) Output voltage of the diode at 18 GHz as a function of input power. Reproduced with permission.[134] Copyright 2014, IEEE.

While having yet to see significant results in printing, the equivalent model derived from this structure shows that the
many CNT-based diodes fabricated by other methods do exist, diode has high cut-off frequency well into the terahertz (THz)
operating in the UHF, microwave, and terahertz bands. For region,[135] highlighting the potential of CNT diodes for RF and
instance, Yang and Chahal developed CNT Schottky diodes THz circuit applications.
with the unique undercut and self-alignment method.[134] In
their work, the DEP technique was used to align SWCNTs in
between asymmetric, closely spaced (1 µm) electrodes to form 4.3.2. Graphene
the coplanar device structure (Figure 13b). A polyether ether
ketone (PEEK) substrate was used due to its high compatibility Graphene, a monolayer 2D honeycomb lattice formed from
with chemicals used in processing and its low loss-characteris- carbon atoms, possesses optoelectronic, plasmonic, mechan-
tics at high frequency. Alignment of multiple CNTs in parallel ical, and electrical properties highly valued in many applica-
between the electrodes allows for better impedance control for tions.[136] Like CNTs, graphene is a material deemed to possess
RF circuit designs. As an example, multiple CNTs are incor- incredible potential for electronics. Existing synthesis methods
porated within a log-periodic structure (Figure 13c,d). Using a for graphene mainly include mechanical, solution-based or
coplanar structure, a rectified voltage of 9 mV is demonstrated chemical-assisted exfoliation,[137,138] chemical synthesis,[139]
at the strongest nonlinear region of the I–V curve (0.4 V) with epitaxial growth,[140] and pyrolysis.[138,139] From the perspective
input RF power of 13 dBm at 18 GHz (Figure 13e,f).[134] Also, of solution-processing and ink production, the most suitable

Adv. Sci. 2018, 1801653 1801653 (17 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

Figure 14. Current techniques in fabricating graphene nanosheets and Schottky diodes fabricated with graphene. a) Cascade centrifugation. Reproduced
with permission.[194] Copyright 2016, American Chemical Society. b) Density gradient ultracentrifugation Reproduced with permission.[195] Copyright
2009, American Chemical Society. c) Photograph of a polymer PmPV/DCE solution with graphene nanoribbons stably suspended in the solution.
d) Chemically derived graphene nanoribbons down to sub 30 nm width. e) Transfer characteristics of the transistor made of graphene nanoribbon with Pd
contacts. (Inset) The AFM image of this device. Scale bar is 100 nm. Reproduced with permission.[143] Copyright 2008, AAAS. f) SEM image of a Schottky
diode fabricated on a graphene monolayer. Reproduced with permission.[146] Copyright 2012, APS. g) Partially reduced graphene oxide-based Schottky
diodes on a PEEK substrate. h) Magnified SEM image of a single Schottky diode from (d). Reproduced with permission.[147] Copyright 2013, IEEE.

synthesis route with potential for cost-effective mass produc- noncovalent polymer functionalization (Figure 14c,d). After
tion lies in solution-based exfoliation techniques on flakes of the purification process, they found that all the sub 10 nm
2D graphene (Figure 14a,b).[141] graphene nanoribbons produced were semiconductors and
The lack of a bandgap in graphene makes it difficult to uti- afforded graphene FETs on–off ratios of about 107 at room tem-
lize in semiconducting applications, including diodes. Thus perature (Figure 14e).
far, the only possible solutions are to use partially reduced gra- Due to the lack of a bandgap, there are few graphene
phene oxide (rGO)[142] or to open a bandgap through forming Schottky diodes reported up to date, and most operated only at
graphene nanoribbons.[143,144] However, chemically rGO must low frequencies and exhibited poor performances.[1] In a recent
undergo a variety of thermal and chemical treatments to retain work, Dragoman et al. reported a Schottky-like diode able
a diminished degree of electrical conductivity, but is noted to to withstand currents at mA level and operate at millimeter
possess a highly disordered structure and electrical properties wave frequencies, greatly surpassing corresponding currents
which are extremely difficult to tune. On the other hand, the in CNT-based Schottky diodes.[146] To achieve higher perfor-
carrier mobility and bandgap observed in graphene nanorib- mance, they used a graphene monolayer with asymmetric
bons depend on the width of the ribbon, and can be compa- metallic contacts deposited on a high-resistivity Si substrate
rable to those of CNTs.[145] For example, Dai and co-workers (Figure 14f). Remarkably, Chahal and co-workers demonstrated
developed a chemical route to produce graphene nanoribbons solution-processed flexible Schottky diodes having high cut-off
with width below 10 nm, stably suspended in solvents with frequency using reduced graphene oxide (Figure 14g).[147] By

Adv. Sci. 2018, 1801653 1801653 (18 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

Figure 15. LEDs from quantum dots. a) Electroluminescence image of a full-color QD display. b) Flexible LED with RGB QDs patterned by transfer
printing (Inset: Magnified image of emission layer rows). Reproduced with permission.[196] Copyright 2011, Springer Nature. c) 3D printed QD-LED on
a 3D scanned onto a curvilinear substrate. d) Color coordinates of the green (0.323, 0.652) and orange-red (0.612, 0.383) QD-LEDs marked by stars
on the Commission International de l′Eclairage (CIE) 1931 chromaticity diagram, with dashed lines representing the HD television color standard
saturation windows as defined by the National Television System Committee. Reproduced with permission.[41] Copyright 2014, American Chemical
Society. e) Schematic diagram shows the structure and materials of an archetypical QD-LED. f) A high-resolution AFM micrograph of a close-packed
monolayer of QDs deposited on top of the CBP hole transporting layer. g) Electroluminescent red, green, and blue QD-LED pixels with the device
structure shown in (e). h) Chromaticity diagram shows the positions of red, green, and blue QD-LED colors, an HDTV color triangle is shown for
comparison. Reproduced with permission.[154] Copyright 2008, American Chemical Society.

adding hydrazine hydrate to a graphene oxide dispersion to Although QD-LEDs have been integrated into electroemis-
reduce the required annealing temperatures to 100 °C, the use sive QD-LED displays in laboratory settings (Figure 15a,b), the
of a thin polymer substrate (PEEK) was enabled on which the device efficiency and lifetime are not satisfactory and need to be
ink was deposited and aligned between Ti (Schottky) and Pd further improved for future commercialization.[149] In pursuit
(ohmic) contact electrodes via the DEP technique (Figure 14h). of excellent cost-effectiveness, further research is also ongoing
As the graphene was not fully reduced, a bandgap was left to in fully inkjet-printed or gravure-printed QLED displays.[151]
ensure the formation of the Schottky barrier, sacrificing carrier Two primary methods exist for the excitation of QDs for light
mobility in the process. Even so, the final device demonstrated emission that determines the structure of the devices. The first
frequency operation up to 26 GHz and could output a 2 mV DC one is optical excitation, which relies on the use of external
voltage given a 10 dBm input power.[147] light to trigger a response in QDs, and has not seen large-scale
applications in printed diodes.[152] The second one is direct elec-
trical excitation, based on the application of current, resulting
4.3.3. Quantum Dots in QD-LEDs with an emission layer sandwiched by electron
transport and hole transport layers.[153]
QDs, defined as very small semiconductor particles only a few The small size of QDs typically allows for relatively easy dis-
nanometers in diameter, are believed to be a good candidate persion within solvents for ink preparation and printing. For
for the light-emission layers of LEDs. The small size of QDs example, McAlpine and co-workers developed a 3D printing
leads to their optical and electronic properties which differ procedure to fabricate printed QD-LEDs on a curvilinear sub-
from those observed in bulk materials.[148] As individual QDs strate (Figure 15c).[41] By interweaving five different materials
with varied sizes can naturally produce monochromatic light of to make up organic polymer charge transport layers, i.e., a QD
specific gaussian distributions, greater precision and efficiency emissive layer, an elastomeric matrix, solid and liquid metal
can be achieved due to the lack of necessity for color filtration leads, as well as a UV-adhesive transparent substrate layer
with a larger spectrum of colors. A weakness does exist in the (Figure 15c), they demonstrated that the printed QD-LEDs can
production of blue light, as most high-quality blue QDs possess generate highly saturated color emissions, enabling the creation
peak emission wavelengths below 455 nm.[149,150] This would of displays that can subtend the color gamut at levels greater
result in weaker luminescence for shades of blue in compar- than the high definition television standard (Figure 15d).
ison to other display technologies. Unfortunately, the lack of scalability of 3D printing techniques

Adv. Sci. 2018, 1801653 1801653 (19 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

is blocking this method from industrial adoption, though deposition.[1,159] Although organic materials may be lacking
it does form a basis for multilayer device fabrication via 3D in durability and lifespan, they are known for high flexibility,
printing in the future. impressive uniformity, stretchability, and low weight, all of
In another work done by Bulovic and co-workers, a solvent- which are highly coveted features in modern flexible electronics
free contact printing process for deposition of colloidal QD applications such as RFID tags.[1]
thin films as the electroluminescent layers within QD-LEDs
was proposed (Figure 15e,f).[154] As shown in Figure 15g, red
(CdSe/ZnS core–shell15), green (ZnSe/CdSe/ZnS core–double 4.4.1. Small Molecules
shell16,17), and blue (CdS/ZnS core–shell18) QD-LED pixels
were contact printed layer by layer to form identical structures. In many cases, organic materials can be easily solution-pro-
It was found that when the wide bandgap organic semicon- cessed and prepared as inks for use in printing. However,
ductor, 4,4′-N,N′-dicarbazole-biphenyl (CBP), replaced N,N′- some difficulties may appear in the solution-processing of
diphenyl-N,N′-di(m-tolyl)benzidine (TPD) as a hole transport organic materials in the small molecule category.[1,160] These
material, more efficient charge and exciton confinement, as materials are typically deposited through evaporation along-
well as an improvement in color saturation of the QD-LEDs side the use of shadow masks, but efforts toward solution-
could be achieved (Figure 15h). The solvent-free deposition processing for higher scalability are in progress. In this
of the QD monolayers is compatible with a wide variety of regard, like most inorganic materials, a precursor route can
organic semiconductors that are not compatible with solution be used.[161] For instance, the functionalization approach used
processing methods, and thus provides a new flexibility in for CNTs can also be applied to organic materials, in which
choosing organic materials for improved device performance. solubilized side chains are mixed with the small molecules for
chemical attachment to prevent clumping and promote solu-
bility.[162] Kjellander et al. applied this approach with the inkjet
4.4. Organic Materials printing of 6,13-bis(triisopropyl-silylethynyl)pentacene and
polystyrene as the active layer of transistors, displaying a sat-
Organic materials possess the highest scalability of any semi- uration mobility of 0.5 cm2 V−1 s−1.[163] They further success-
conducting material seen thus far, as well as the highest com- fully built inverters, NANDs, and oscillators, then integrated
patibility with solution processing and printing.[8] The raw nearly 300 TFTs in a matrix possessing the surface area of
materials can typically be synthesized from solution on a large 34 mm2 for 8 bit RFID transponder circuits (Figure 16a,b).
scale without high processing temperatures, enabling devel- However, due to the limitations of inkjet printing in con-
opment of high-throughput, cost effective fabrication through trolling the morphology of 6,13-bis(triisopropylsilylethynyl)
established printing techniques. Different from the fragility (TIPS)-PEN crystals, a large variation of performance in
of inorganic materials commonly encountered after printing transistors was observed.[163]
and thermal annealing, most organic materials are capable of Another useful solution-processing approach is the blending
retaining a superb degree of flexibility which renders them of the molecules with a solution of polymers, which acts as a
perfectly suited for flexible device applications.[3] Moreover, binder to form a framework for higher stability.[68] In many
organic polymers allow a quick modification of their chemical cases, this strategy can assist in substrate compatibility and film
structure, electronic bandgap, ink solution properties, as well formation, ensuring the deposition of an even active layer that is
as mechanical properties through chemical synthesis in accord- able to stably bind to the substrate.[164] For instance, Niazi et al.
ance to the necessary device requirements.[3,155] splendidly proved the efficacy of this procedure with a blend of
In comparison to inorganic materials, however, organic sem- the 2,8-difluoro-5,11-bis(triethylsilylethynyl)anthradithiophene
iconductor materials possess some disadvantages in electrical (diF-TES-ADT) organic small molecule with the amorphous insu-
performance due to their different generating mechanism of lating polymer polystyrene/poly(α-methylstyrene) (PS/PαMS)
semiconducting properties.[155–157] Within organic materials, which was deposited via blade coating (Figure 16c,d).[165] As
conjugated π bonds of alternating single and double bonds exist shown in Figure 16e,f, the blended mixture of polymer and
between the carbon atoms, generating the delocalization of elec- small molecule did give a higher hole mobility and on–off cur-
tronic charge in the relevant areas of the molecules.[5,156,158] As rent ratio, but also smaller threshold voltage. The best TFT
a result, electrons in organic materials may demonstrate carrier device fabricated with this approach reached a carrier mobility as
mobility by hopping between localized states of molecules or high as 6.7 cm2 V−1 s−1.[165] Another work by Hwang et al. used
thermally assisted tunneling methods, which typically result in a mixture of TIPS–pentacene, poly (α-methyl styrene), and poly
lower mobility than that in the inorganic semiconductors with (triarylamine) polymer matrices, along with tetralin as a solvent
band transport mechanism. The weaker intermolecular bonds to fabricate solution-processed organic field-effect transistors
in organic materials have also led to inferior mechanical and (OFETs) of the structure in Figure 16g,h with mobility values up
thermodynamic properties and higher deterioration rates com- to 2.82 cm2 V−1 s−1.[166] As the methods have been demonstrated
pared to those of inorganic materials.[155] Based on the basic to be capable of consistently generating well-performing TFTs of
transport mechanism of organic semiconductors, it is believed low variability, the overall strategy can be considered potentially
that the most important parameters affecting charge carrier suitable for high-throughput manufacturing on a larger scale.
mobility in fabricated devices include the degree of crystallinity Pentacene is a polyacene analogue which has distinguished
in organic layers, the degree of precision in control over the itself as the primary organic material candidate for high-perfor-
molecular structure, and the quality of chain alignment in layer mance HF and UHF operation due to its high charge carrier

Adv. Sci. 2018, 1801653 1801653 (20 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

Figure 16. Electronics applications of the organic small molecules. a) Photograph of 8 bit RFID transponder chips on a plastic foil, based on a blend
of TIPS-pentacene and the polymer polystyrene. Each transponder has a footprint of 34 mm2. b) Output oscillations for an 8-bit RFID transponder,
with ink-jet printed TIPS-PEN:PS blend as the active layer. Reproduced with permission.[163] Copyright 2013, Elsevier. c) BCBG device architecture and
d) blade-coating set-up. e) Transfer characteristics of OTFTs prepared using neat diF-TES-ADT, diF-TES-ADT:PαMS (100 kDa) and diF-TES-ADT:PS
(123 kDa) blends at a blade speed of 1.5 mm s−1 and a stage temperature of 70 °C. We employed Vds = −10 V for blends and Vds = −20 V for neat OSC.
f) Hole mobility of OTFTs fabricated by blade-coating diF-TES-ADT:PS blends using different Mw of PS both in different blade speeds. Reproduced
with permission.[165] Copyright 2015, Springer Nature. g) Schematic illustration of OFETs with the dual gate geometry made of TIPS-pentacene/
polymer blend film. h) TIPS-pentacene/polymer blend films morphologies of PTAA blend layers processed from tetralin. i) Transfer characteristics of
TIPS-pentacene/poly-(triarylamine) (PTAA) blend OFETs with top gate geometry processed from tetralin; CYTOP (700 nm) single layer gate dielectric.
Reproduced with permission.[166] Copyright 2012, Royal Society of Chemistry.

mobilities.[167] Nonsolution processed pentacene-based diodes deposited through zone-casting (Figure 17c), a large-area com-
have reached into the HF range, even touching the UHF range, patible deposition method and nanoimprint lithography was
while solution-processed printed pentacene-based diodes can at used for patterning the device, allowing channel lengths down to
most be expected to remain in the 13.56 MHz band in the HF 375 nm.[169] Frequencies in the lower end of the HF band ranging
range. Up to now, the highest-performing organic diode was from 1 to 9.8 MHz depending on the biasing voltage were
built for a 1.24 GHz rectifier, which included a pentacene-based observed, along with the mobility values shown in Figure 17d.
diode with Au and Al contact electrodes (Figure 17a).[168] The Another prominent semiconducting small organic molecule
pentacene active layer was deposited through thermal evapo- is buckminsterfullerene (C60) which has distinguished itself
ration on a shadow mask with a glass substrate, while the Au through higher mobility values.[170] de Zerio Mendaza et al.
electrode was given treatment with self-assembled monolayers demonstrated TFTs formed from a blend of C60 and 9% poly-
to ensure efficient charge injection. While possessing a 3 dB fre- styrene of high molecular weight which possessed mobility up
quency at 1.24 GHz, the peak output voltage of 3.8 V was reached to 1 cm2 V−1 s−1 on a SiO2 substrate (Figure 17e–g).[171] Other
at a 1 GHz AC input frequency (Figure 17b), fully demonstrating work by Smith et al. employed a blend of the diF-TES-ADT
the strong potential of pentacene-based diodes for future UHF small organic molecule with poly(dimethyl-triarylamine) (PTAA)
applications. Prominently, Higgins et al. produced state-of- and poly(dialkyl-fluorene-co-dimethyl-triarylamine) (PF-TAA) as
the-art solution-processed OFETs based on TIPS–pentacene amorphous polymeric semiconducting binders to form polymer

Adv. Sci. 2018, 1801653 1801653 (21 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

Figure 17. Electronics made of pentacene and C60. a) Structural schematic of a pentacene diode with self-assembled monolayers of
2,3,4,5,6-pentafluorobenzenethiol (PFBT) coated on the Au anode. b) The current density response of the structure in (a) to various biasing voltages
(Inset: Corresponding diode rectification ratios). Reproduced with permission.[168] Copyright 2016, Wiley. c) Optical micrograph image of zone-cast
TIPS-pentacene crystals on a self-aligned structure. d) Recorded mobility for zone-cast and spin-coated TFT devices with the structure of (c). Reproduced
with permission.[169] Copyright 2015, Wiley. e) Saturation mobility of C60:C70 films as a function of the weight ratio (Inset: Images of C60:C70 film (above,
full coverage granted) and C60 film (below, incomplete coverage)). f) Ternary phase diagram for C60:C70:O-DCB with a blue-shaded single-phase region
and a two-phase region corresponding to transmission optical images seen to the top left and right, respectively. g) Diagram of C60:C70 solubility as a
function of the weight ratio. Reproduced with permission.[171] Copyright 2015, Wiley. h) Saturation and linear carrier mobilities of diF-TES-ADT:PTAA
and diF-TES-ADT:PF-TAA blend OTFTs. Reproduced with permission.[172] Copyright 2012, Wiley. i) Transfer characteristics of a C10-DNBDT transistor.
Reproduced with permission.[173] Copyright 2015, Wiley.

matrices.[172] A significant enhancement of hole transport was technique, in which a crystalline domain is grown along the
observed, leading to field effect mobilities above 5 cm2 V−1 s−1 in direction of an inclined substrate from a solution droplet sus-
the fabricated top-gate OFETs, which can be considered encour- tained on an edge (Figure 17i).[173] The utilization of these TFTs
aging for the future of solution-processed organic small mole- within half-wave rectifiers yielded impressive operation frequen-
cules (Figure 17h). Uno et al. incorporated decyldinaphthoben- cies up to 22 MHz with a 2 µm channel length, and later integra-
zodithiophene (C10-DNBDT) within TFTs possessing mobilities tion of the rectifiers with a 13.56 MHz transponder successfully
reaching 16 cm2 V−1 s−1 through an “edge-casting” fabrication powered a 5-stage ring oscillator.

Adv. Sci. 2018, 1801653 1801653 (22 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

4.4.2. Polymers incorporation into a rectifier with a function generator input,

an output-smoothing capacitor, an oscilloscope probe, and a
In comparison to small molecules, polymers have seen far more 1 MΩ load resistance, the resulting rectification ratio reached
research in the fabrication of solution-processed semiconductor as high as 2 × 104 with a forward current density of 6 A cm−2
devices capable of RF operation. An early example of gravure while assuming an AC input voltage of 5 V in amplitude
printing was presented by Lilja et al., who used roll-to-roll gra- (Figure 18b,c). The operation frequency reached up to 14 MHz.
vure printing to form rectifying organic Schottky diodes on a Building on previous work, Heljo et al. progressed even
metallized polyester (PET) film.[174] PTAA was used as the active further, again fabricating 13.56 MHz organic Schottky diodes
layer, with a silver layer as the anode and a patterned copper through gravure printing.[11] PTAA was used as the active layer
layer as the cathode. After incorporation within a rectifier, the with Cu and Ag electrode layers, and incorporation into two rec-
operation frequency reached about 10 MHz with an input tifier configurations with full wave and half wave circuits both
AC voltage 10 V in amplitude, alongside a rectification ratio resulted in functional 13.56 MHz operation (Figure 18d,e). The
over 104. Later, Lin et al. employed the lower-throughput spin maximum output voltage achieved reached 3.5 V at 13.56 MHz
coating approach to form vertical Schottky diodes employed with an AC input voltage of 10 V in amplitude, and the fabri-
in rectifiers which did reach the 13.56 MHz operation band, cation process itself demonstrated excellent reliability at a 97%
thus qualifying for standard HF operation.[175] As presented in device yield.[176] Higgins et al. later successfully took device per-
Figure 18a, they utilized poly(3,30 00didodecylquaterthiophene) formance up another notch, again at the cost of throughput,
(PQT-12) as the active layer, along with indium zinc oxide (IZO) employing spin coating with an indacenodithiophene-benzo-
and Al as the anode and cathode electrodes, respectively. After thiadiazole copolymer (C16-IDT-BT) in a Schottky diode using

Figure 18. Schottky diode from organic polymers. a) Schottky diode structure including a PEDOT:PSS hole injection layer. b) Fully functional flexible
wireless power transmission sheet with PQT-12-based Schottky diodes fabricated via spin-coating with the structure of (a). c) Current density as a
function of bias voltage for the diodes with three polymer varieties. Reproduced with permission.[175] Copyright 2011, Elsevier. d) Photograph of gravure
printed full-wave circuit including four PTAA-based Schottky diodes. e) Output voltage as a function of operation frequency for various rectifier setups
with the diodes of (d). Reproduced with permission.[11] Copyright 2013, IEEE. f) C16IDT-BT-based diode structure. g) Output voltage as a function of
frequency for half-wave rectifiers with spin-coated Schottky diodes based on the design of (f). h) Image of the diode on a flexible substrate and the
I–V characteristics for the diode both with and without a PEIE interlayer separating the cathode and semiconductor. Reproduced with permission.[177]
Copyright 2017, Wiley. i) Fully inkjet-printed PTAA-based MIS diode architecture. j) Plot of current density as a function of voltage and the number of
MMAcoMAA layers for the device structure in (i). Reproduced with permission.[178] Copyright 2017, IOP Publishing.

Adv. Sci. 2018, 1801653 1801653 (23 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

Au and molybdenum trioxide/silver (MoO3/Ag) electrodes as need for complex interlacing techniques was reduced with an
the cathode and anode, respectively (Figure 18f–h).[177] A high additional hole-conduction layer for charge transport. More
reverse breakdown voltage nearly reaching 15 V was observed, recently, Kopola et al. reported the fabrication of OLED struc-
as well as a rectification ratio exceeding 106 with input voltages tures with modified PEDOT:PSS layers for lighting applications
between −5 and 5 V. through gravure printing on glass substrates, reaching lumi-
As an alternative to gravure printing, Mitra et al. demon- nosities up to 1000 cd m−2 with a 5.4 V input voltage.[184] They
strated fully inkjet-printed metal-insulator-semiconductor demonstrated large-area fabrication up to 30 cm2 with fully
(MIS) diodes with a PTAA active layer, Ag electrodes, and a poly gravure printing of the polymer layers, further reinforcing the
(methylmethacrylate-methacrylic acid) (MMAcoMAA) insu- feasibility of OLED production via printing.[185]
lating layer,[178] as seen in Figure 18i. Given 10 V amplitude Research regarding the printed solar cells has also steadily
AC input voltages, the maximum forward current density was leaned toward the use of organic materials due to their process-
measured to be 40 mA cm−2 after layer structure optimization ability in low temperatures, compatibility with flexible substrates,
as indicated in Figure 18j, corresponding to rectification ratios and relatively lower cost in photovoltaic applications.[186,187] In
in the range of 103–104. These results suggest that organic poly- performance, the most efficient organic solar cells (OSCs) have
mers are nearly certain to enter the commercial sector of RF demonstrated power conversion efficiencies in the range of
semiconducting devices operating within the HF band in the 10–13% in single junction cells owing to the employment of non-
future. The results achieved by groups utilizing organic semi- fullerene acceptor and conjugated polymer donor materials in
conductors in nonphotonic applications are included in Table 1. combination with a buffer layer for separating the active layer from
the electrodes to increase device efficiency.[188,189] In one notable
example, by modifying the molecular structure of the low bandgap
4.4.3. Optoelectronics Applications n-type organic semiconductor 3,9-bis(2-methylene-(3-(1,1-
dicyanomethylene)-indanone))-5,5,11,11-tetrakis(4-hexylphenyl)-
In recent years, great interest has been focused on the develop- dithieno[2,3-d:2′,3′-d′]-s-indaceno[1,2-b:5,6-b′]dithiophene (ITIC)
ment of organic photodiodes for existing imaging technologies via side-chain isomerization with meta-alkyl-phenyl substitu-
due to the combination of their lightness of weight, flexibility, tion, Yang et al. developed spin-coated polymer solar cells (PSCs)
biocompatibility, tunable photophysical or optoelectronic prop- that could reach power conversion efficiency of 11.77%, one of
erties, and compatibility with both large and small areas.[179] the highest reported efficiency values in nonfullerene PSCs.[190]
Recent research has mainly concentrated on the tuning of solu- Despite great strides in performance, printed organic solar
tion-processable organic semiconducting materials, fabrication cells are still far from competing with the state-of-the-art single-
procedures, and device architectures for the optimization of junction tandem solar cells which have already shown a power
performance parameters. For example, Eckstein et al. recently conversion efficiency over 27.3%.[186,188]
demonstrated fully printed, flexible, and semi-transparent
organic photodiodes with an aluminum-doped zinc oxide (AZO)
nanoparticle hole-blocking layer, a blended polythieno[3,4-b]- 5. Conclusions
thiophene-co-benzodithiophene (PTB7) and [6,6]-phenyl C71-
butyric acid methyl ester (PC70BM) bulk heterojunction active In this review, we have systematically examined the unique
layer, and PEDOT:PSS electrodes for a PEDOT:PSS (200 nm)/ materials, printing technologies, and different kinds of devices
AZO NP (50 nm)/PTB7:PC70BM (310 nm)/PEDOT:PSS and applications of printed diodes. In all the materials used
(200 nm) structure.[180] The all-printed organic photodiode pos- for diode applications, organics are undoubtedly the most
sessed an external quantum efficiency (EQE) of 50%, device promising candidates for printing methodologies due to their
transparency of ≈15%, a specific detectivity (D*) above 1011, a solution-processing compatibility and lower cost. However,
spectral response of 0.26 A W−1, and a ≈300 kHz band width organic diodes will likely remain locked in the HF applica-
at −3 V reverse bias, exceeding nearly all comparable devices tions in the foreseeable future due to their inherent drawbacks.
of its category. The performance of organic photodiodes has The current fabrication for printed organic diodes is close to
caught up to, and even exceeded that of inorganic photodiodes maturity, and further development may remain reliant on the
in multiple areas. Higher EQE values up to 80% have been introduction of new materials such as the high-mobility soluble
observed in fully inkjet-printed organic photodetectors using a small molecules or novel polymer/small molecule blends. By
poly(3-hexyltiophene) and [6,6]-phenyl-C 61 butyric acid methyl combining the high uniformity of organic materials suited
ester bulk heterojunction photoactive layer with an Ag bottom for large-area deposition and the high-throughput traits of
electrode and PEDOT:PSS top electrode built on a PEN sub- printing, printed organic diodes can be expected to boast the
strate.[181] Combined with the inexpensive and scalable nature least expensive prices of all material classes in the future. In the
of printing, the future integration of organic photodiodes with field of photonics, printed OLED-based device applications in
suitable applications in large area flexible digital imagers, bio- the textile, lighting, display sectors are unlikely to outperform
electronics, surveillance systems, and other fields is rapidly peers in lifespan or overall performance, but possess a signifi-
approaching realization.[179,182] cant chance of becoming competitive within the next decade of
In the OLED field, Dongwon et al. developed full-color development by virtue of far more accessible cost.
polymer LED displays deposited through inkjet printing after After organics, silicon may be another good material for
careful optimization of ink formation and drying procedures printed semiconducting devices. While current focus is on
for uniform film profiles and uniform photoemission.[183] The the transfer of prefabricated Si nanomembranes to flexible

Adv. Sci. 2018, 1801653 1801653 (24 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

substrates, this method has only yielded excellent results Keywords

on a small scale and faces severe issues in throughput and
nanomaterials, organic light-emitting diodes (OLEDs), printed diodes,
cost-effectiveness. Depending on development, solution pro-
printing technologies, radio frequency identifications (RFIDs)
cessing could become a scalable fabrication process while
remaining suitable for applications in the UHF and per- Received: September 24, 2018
haps even microwave bands. Although MOS such as IGZO Revised: November 2, 2018
showed excellent performance in HF, UHF, and even higher Published online:
applications in rectifiers and additional circuitry with the
vacuum deposition method, solution processed fabrication
is still facing difficulties in the form of the high annealing
temperatures needed for MOS precursor solutions, especially [1] J. Semple, D. G. Georgiadou, G. Wyatt-Moon, G. Gelinck,
T. D. Anthopoulos, Semicond. Sci. Technol. 2009, 206, 588.
in the case of IGZO. However, the possibility does exist that
[2] W. S. Wong, A. Salleo, Flexible Electronics: Materials and
binary or more complex oxides may yield a new material
Applications, Vol. 11, Springer Science & Business Media, New
able to provide an easier route to solution-processed ink for York, USA 2009.
printing. [3] M. Berggren, D. Nilsson, N. D. Robinson, Nat. Mater. 2007,
The unique electrical and mechanical properties of nano- 6, 3.
materials, such as excellent durability, unmatchable flexibility, [4] a) H. Rost, W. Mildner, Kunstst. Int. 2008, 98, 60; b) D. L. Sengupta,
and high charge carrier mobilities as well as solution-pro- T. K. Sarkar, D. Sen, Proc. IEEE 1998, 86, 235.
cessing compatibility, render them exceptionally suitable for [5] W. Brütting, Physics of Organic Semiconductors, WILEY-VCH Verlag
flexible semiconducting applications and may represent the GmbH & Co. KGaA, Weinheim, Germany 2006.
greatest hope of printing diodes capable of microwave-band [6] a) A. Jonscher, Br. J. Appl. Phys. 1961, 12, 654; b) M. Shur,
Introduction to Electronic Devices, John Wiley & Sons, Hoboken,
operation, competing with silicon. Nevertheless, their unique
NJ 1996; c) E. H. Rhoderick, IEEE Proceedings I—Solid-State
properties have also led to frustrating fabrication challenges,
and Electron Devices 1982, 129, (accessed: February 1982);
such as the impedance mismatch, alignment, and metallic d) B. Sharma, Metal–Semiconductor Schottky Barrier Junctions and
impurity issues seen in CNTs. While printed diodes are nearly Their Applications, Springer Science & Business Media, New York,
certain to enter large-scale production for device applications USA 2013.
in the HF band employing organic materials as well as in [7] S. Höfle, A. Schienle, C. Bernhard, M. Bruns, U. Lemmer,
photonics applications, their level of involvement in UHF and A. Colsmann, Adv. Mater. 2014, 26, 5155.
microwave bands will greatly depend on the degree of success [8] T. M. Kraft, P. R. Berger, D. Lupo, Flexible Printed Electron. 2017, 2,
achieved in research involving nanomaterials and inorganic 033001.
materials. [9] M. A. Leenen, V. Arning, H. Thiem, J. Steiger, R. Anselmann, Phys.
Status Solidi A 2009, 206, 588.
Furthermore, in printing fabrication, to achieve high perfor-
[10] S. Abbasiniazare, O. Manoochehri, K. Forooraghi, Microwave Opt.
mance, it is critically important to carefully tune the ink prop-
Technol. Lett. 2014, 56, 1151.
erties, printing procedure type, and substrate surface during [11] P. S. Heljo, M. Li, K. E. Lilja, H. S. Majumdar, D. Lupo, IEEE Trans.
deposition as well as later annealing or other treatment steps to Electron Devices 2013, 60, 870.
dry or remove impurities from the deposited ink. The materials [12] M. Shen, Schottky Nanodiodes Based on Zinc Oxide Thin Films,
used tend to suffer from fabrication issues of a severity degree Doctoral dissertation, University of Alberta, 2017.
inversely proportional to their degree of potential, with carbon [13] J. Saltich, L. Clark, Solid-State Electron. 1970, 13, 857.
nanomaterials being associated with the most difficulties and [14] L. Esaki, Phys. Rev. 1958, 109, 603.
organic polymers the least. In addition, the need to achieve [15] S. Krishnan, in Electrical Engineering, Master of Science University
good metal–semiconductor interfaces is still a major challenge of South Florida, USA 2004.
[16] a) T. Sollner, W. Goodhue, P. Tannenwald, C. Parker, D. Peck,
using printing technologies, and low-temperature processes
Appl. Phys. Lett. 1983, 43, 588; b) P. R. Berger, A. Ramesh,
are critically needed. Taking into consideration the significant
Negative Differential Resistance Devices and Circuits. Comprehensive
commercial potential of printed diodes and their suitability for Semiconductor Science and Technology 2011, 5. 176.
flexible applications, strong interest will only continue to rise in [17] P. Chahal, F. Morris, G. Frazier, IEEE Electron Device Lett. 2005,
the coming future. 26, 894.
[18] a) N. Holonyak, I. Lesk, Proc. IRE 1960, 48, 1405; b) R. J. Locker,
G. C. Huth, Appl. Phys. Lett. 1966, 9, 227; c) R. Minton,
R. Glicksman, Solid-State Electron. 1964, 7, 491.
Acknowledgements [19] D. R. Baraff, N. M. Serinken, R. W. Streater, C. J. Miner, R. J. Boynton,
B. K. MacLaurin, W. D. Westwood, Google Patents, 1983.
C.C. acknowledges the financial support of the start-up fund from the
[20] a) P. P. Heinonen, Fabrication of organic tunnel diodes with ultra-
MSU School of Packaging and the financial support from the USDA
thin TiO2 interfacial layers, Tampereen teknillinen yliopisto,
National Institute of Food and Agriculture, Hatch Project (Grant No.
1016788). Tampere University of Technology 2015; b) J. J. Guttman,
C. B. Chambers, G. N. C. Santos, P. R. Berger, Org. Electron. 2017,
47, 228; c) L. L. Chang, L. Esaki, R. Tsu, Appl. Phys. Lett. 1974, 24,
593; d) R. Tsu, L. Esaki, Appl. Phys. Lett. 1973, 22, 562.
[21] K. Meerholz, Nature 2005, 437, 327.
Conflict of Interest [22] a) M. A. Baldo, D. O’brien, Y. You, A. Shoustikov, S. Sibley,
The authors declare no conflict of interest. M. Thompson, S. Forrest, Nature 1998, 395, 151; b) J. Burroughes,

Adv. Sci. 2018, 1801653 1801653 (25 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

D. Bradley, A. Brown, R. Marks, K. Mackay, R. Friend, P. Burns, [51] M. S. Onses, E. Sutanto, P. M. Ferreira, A. G. Alleyne, J. A. Rogers,
A. Holmes, Nature 1990, 347, 539. Small 2015, 11, 4237.
[23] B. Geffroy, P. Le Roy, C. Prat, Polym. Int. 2006, 55, 572. [52] C. Cao, J. B. Andrews, A. D. Franklin, Adv. Electron. Mater. 2017, 3,
[24] X. Gong, S. Wang, D. Moses, G. C. Bazan, A. J. Heeger, Adv. Mater. 1700057.
2005, 17, 2053. [53] C. Cao, J. B. Andrews, A. Kumar, A. D. Franklin, ACS Nano 2016,
[25] M. H. Crawford, IEEE J. Sel. Top. Quantum Electron. 2009, 15, 1028. 10, 5221.
[26] I. Ugarte, I. Castelló, E. Palomares, R. Pacios, in Quantum Dots—A [54] a) H. Yuxiao, C. Oakley, P. Chahal, J. Albrecht, J. Papapolymerou,
Variety of New Applications, IntechOpen Limited, London, UK 2012. presented at 2017 International Workshop on Antenna Technology:
[27] W. Wu, Nanoscale 2017, 9, 7342. Small Antennas, Innovative Structures, and Applications (iWAT),
[28] J. Perelaer, P. J. Smith, D. Mager, D. Soltman, S. K. Volkman, Athens, Greece, March 2017; b) C. Oakley, A. Kaur, J. A. Byford,
V. Subramanian, J. G. Korvink, U. S. Schubert, J. Mater. Chem. P. Chahal, presented at 2017 IEEE 67th Electronic Components and
2010, 20, 8446. Technology Conference (ECTC), Orlando, Florida, 30 May–2 June
[29] a) L. Hu, J. Li, J. Liu, G. Grüner, T. Marks, Nanotechnology 2010, 2017.
21, 155202; b) T. Cheng, Y. Zhang, W. Y. Lai, W. Huang, Adv. Mater. [55] J. A. Lewis, B. Y. Ahn, Nature 2015, 518, 42.
2015, 27, 3349; c) J.-S. Park, H. Chae, H. K. Chung, S. I. Lee, [56] J. B. Andrews, C. Cao, M. A. Brooke, A. D. Franklin, IEEE Sens. J.
Semicond. Sci. Technol. 2011, 26, 034001. 2017, 17, 4612.
[30] a) K. Myny, S. Steudel, P. Vicca, M. J. Beenhakkers, N. A. Van Aerle, [57] a) H. Pan, S. H. Ko, N. Misra, C. P. Grigoropoulos, Appl.
G. H. Gelinck, J. Genoe, W. Dehaene, P. Heremans, Solid-State Phys. Lett. 2009, 94, 071117; b) H. Pan, N. Misra, S. H. Ko,
Electron. 2009, 53, 1220; b) J. O. McSpadden, L. Fan, K. Chang, C. P. Grigoropoulos, N. Miller, E. E. Haller, O. Dubon, Appl.
IEEE Trans. Microwave Theory Tech. 1998, 46, 2053. Phys. A 2009, 94, 111; c) N. Misra, L. Xu, M. Rogers, S. H. Ko,
[31] C. Biswas, Y. H. Lee, Adv. Funct. Mater. 2011, 21, 3806. C. Grigoropoulos, Phys. Status Solidi (c) 2008, 5, 3264.
[32] a) B. Lüssem, M. Riede, K. Leo, Phys. Status Solidi A 2013, 210, [58] a) H. Pan, D. Lee, S. H. Ko, C. P. Grigoropoulos, H. K. Park,
9; b) H. Méndez, G. Heimel, A. Opitz, K. Sauer, P. Barkowski, T. Hoult, Appl. Phys. A 2011, 104, 29; b) B. Kang, S. Han, J. Kim,
M. Oehzelt, J. Soeda, T. Okamoto, J. Takeya, J. B. Arlin, Angew. S. Ko, M. Yang, J. Phys. Chem. C 2011, 115, 23664.
Chem., Int. Ed. 2013, 52, 7751. [59] a) S. A. Patil, H.-J. Hwang, M.-H. Yu, N. K. Shrestha, H.-S. Kim,
[33] a) Y.-C. Liang, V. J. Gosbell, IEEE Trans. Power Electron. 1990, 5, 346; RSC Adv. 2017, 7, 6565; b) J. H. Park, S. Han, D. Kim, B. K. You,
b) P. O. Lauritzen, C. L. Ma, IEEE Trans. Power Electron. 1991, 6, 188. D. J. Joe, S. Hong, J. Seo, J. Kwon, C. K. Jeong, H. J. Park, Adv.
[34] S. Cheung, N. Cheung, Appl. Phys. Lett. 1986, 49, 85. Funct. Mater. 2017, 27, 1701138.
[35] a) T. J. Krutsick, Google Patents, 2000; b) M. Lepselter, S. Sze, Bell [60] K. Banger, Y. Yamashita, K. Mori, R. Peterson, T. Leedham,
Syst. Tech. J. 1968, 47, 195. J. Rickard, H. Sirringhaus, Nat. Mater. 2011, 10, 45.
[36] J. Semple, S. Rossbauer, C. H. Burgess, K. Zhao, L. K. Jagadamma, [61] M. Benwadih, R. Coppard, K. Bonrad, A. Klyszcz, D. Vuillaume,
A. Amassian, M. A. McLachlan, T. D. Anthopoulos, Small 2016, 12, ACS Appl. Mater. Interfaces 2016, 8, 34513.
1993. [62] N. D. Arora, J. R. Hauser, D. J. Roulston, IEEE Trans. Electron
[37] J. Semple, S. Rossbauer, T. D. Anthopoulos, ACS Appl. Mater. Devices 1982, 29, 292.
Interfaces 2016, 8, 23167. [63] a) H.-C. Yuan, G. K. Celler, Z. Ma, J. Appl. Phys. 2007, 102,
[38] D. Georgiadou, J. Semple, T. Anthopoulos, SPIE Newsroom 2017, 16. 034501; b) L. Sun, G. Qin, J. H. Seo, G. K. Celler, W. Zhou, Z. Ma,
[39] D. J. Beesley, J. Semple, L. K. Jagadamma, A. Amassian, Small 2010, 6, 2553; c) Z. Ma, K. Zhang, J.-H. Seo, H. Zhou,
M. A. McLachlan, T. D. Anthopoulos, Nat. Commun. 2014, 5, L. Sun, H.-C. Yuan, G. Qin, H. Pang, W. Zhou, ECS Trans.
ncomms4933. 2011, 34, 137; d) H.-C. Yuan, Z. Ma, Appl. Phys. Lett. 2006, 89,
[40] a) L. Qian, Y. Zheng, J. Xue, P. H. Holloway, Nat. Photonics 2011, 5, 212105.
543; b) S. Sax, N. Rugen-Penkalla, A. Neuhold, S. Schuh, E. Zojer, [64] G. Qin, H.-C. Yuan, G. K. Celler, W. Zhou, J. Ma, Z. Ma,
E. J. List, K. Müllen, Adv. Mater. 2010, 22, 2087; c) A. A. Zakhidov, Microelectron. J. 2011, 42, 509.
J. K. Lee, H. H. Fong, J. A. DeFranco, M. Chatzichristidi, [65] J.-H. Seo, T. Ling, S. Gong, W. Zhou, A. L. Ma, L. J. Guo, Z. Ma,
P. G. Taylor, C. K. Ober, G. G. Malliaras, Adv. Mater. 2008, 20, Sci. Rep. 2016, 6, 24771.
3481; d) R. D. Farahani, M. Dubé, D. Therriault, Adv. Mater. 2016, [66] a) H. Hieslmair, S. Chiruvolu, H. Du, Google Patents, 2013;
28, 5794. b) K. Kunze, S. Haubrich, F. Zurcher, B. Ridley, J. Rockenberger,
[41] Y. L. Kong, I. A. Tamargo, H. Kim, B. N. Johnson, M. K. Gupta, Google Patents, 2008.
T.-W. Koh, H.-A. Chin, D. A. Steingart, B. P. Rand, M. C. McAlpine, [67] K. Alberi, G. Scardera, H. Moutinho, R. Reedy, M. Romero,
Nano Lett. 2014, 14, 7017. E. Rogojina, M. Kelman, D. Poplavskyy, D. Young, F. Lemmi,
[42] G. Grau, J. Cen, H. Kang, R. Kitsomboonloha, W. J. Scheideler, presented at 35th IEEE Photovoltaic Specialists Conference (PVSC),
V. Subramanian, Flexible Printed Electron. 2016, 1, 023002. Honolulu, HI, USA, December 2010.
[43] P. Calvert, Chem. Mater. 2001, 13, 3299. [68] N. Sani, M. Robertsson, P. Cooper, X. Wang, M. Svensson,
[44] M. Singh, H. M. Haverinen, P. Dhagat, G. E. Jabbour, Adv. Mater. P. A. Ersman, P. Norberg, M. Nilsson, D. Nilsson, X. Liu, Proc.
2010, 22, 673. Natl. Acad. Sci. USA 2014, 111, 11943.
[45] E. B. Secor, P. L. Prabhumirashi, K. Puntambekar, M. L. Geier, [69] A. Facchetti, Proc. Natl. Acad. Sci. USA 2014, 111, 11917.
M. C. Hersam, J. Phys. Chem. Lett. 2013, 4, 1347. [70] N. Sani, X. Wang, H. Granberg, P. A. Ersman, X. Crispin,
[46] a) B. J. De Gans, P. C. Duineveld, U. S. Schubert, Adv. Mater. 2004, P. Dyreklev, I. Engquist, G. Gustafsson, M. Berggren, Sci. Rep.
16, 203; b) R. P. Tortorich, J.-W. Choi, Nanomaterials 2013, 3, 453; 2016, 6, 28921.
c) J.-W. Song, C.-S. Han, Int. J. Precis. Eng. Manuf. 2008, 9, 79. [71] T. Shimoda, Y. Matsuki, M. Furusawa, T. Aoki, I. Yudasaka,
[47] J.-W. Song, J. Kim, Y.-H. Yoon, B.-S. Choi, J.-H. Kim, C.-S. Han, H. Tanaka, H. Iwasawa, D. Wang, M. Miyasaka, Y. Takeuchi,
Nanotechnology 2008, 19, 095702. Nature 2006, 440, 783.
[48] A. Kamyshny, S. Magdassi, Small 2014, 10, 3515. [72] S. Tatsuya, M. Takashi, Jpn. J. Appl. Phys. 2014, 53, 02BA01.
[49] S. Khan, L. Lorenzelli, R. S. Dahiya, IEEE Sens. J. 2015, 15, 3164. [73] S. Han, X. Dai, P. Loy, J. Lovaasen, J. Huether, J. M. Hoey,
[50] D. A. Pardo, G. E. Jabbour, N. Peyghambarian, Adv. Mater. 2000, A. Wagner, J. Sandstrom, D. Bunzow, O. F. Swenson, J. Non-Cryst.
12, 1249. Solids 2008, 354, 2623.

Adv. Sci. 2018, 1801653 1801653 (26 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

[74] a) A. C. Bedini, S. Muthmann, J. Allgaier, K. Bittkau, F. Finger, C. Dekker, Phys. Rev. Lett. 2000, 84, 2941; h) L. Nougaret,
R. Carius, Ultrason. Sonochem. 2017, 34, 289; b) A. P. Cádiz Bedini, H. Happy, G. Dambrine, V. Derycke, J.-P. Bourgoin, A. Green,
S. Muthmann, J. Flohre, B. Thiele, S. Willbold, R. Carius, M. Hersam, Appl. Phys. Lett. 2009, 94, 243505; i) M. J. Treacy,
Macromol. Chem. Phys. 2016, 217, 1655. T. Ebbesen, J. Gibson, Nature 1996, 381, 678.
[75] G. R. Iyer, E. K. Hobbie, S. Guruvenket, J. M. Hoey, K. J. Anderson, [96] E. T. Thostenson, Z. Ren, T.-W. Chou, Compos. Sci. Technol. 2001,
J. Lovaasen, C. Gette, D. L. Schulz, O. F. Swenson, A. Elangovan, 61, 1899.
ACS Appl. Mater. Interfaces 2012, 4, 2680. [97] a) S. Hasan, M. A. Alam, M. S. Lundstrom, IEEE Trans. Electron
[76] a) M. Prins, K. O. Grosse-Holz, G. Müller, J. Cillessen, Devices 2007, 54, 2352; b) A. Vignes, O. Dufaud, L. Perrin,
J. Giesbers, R. Weening, R. Wolf, Appl. Phys. Lett. 1996, 68, 3650; D. Thomas, J. Bouillard, A. Janes, C. Vallieres, Chem. Eng. Sci.
b) E. H. Nicollian, J. R. Brews, E. H. Nicollian, MOS (Metal Oxide 2009, 64, 4210.
Semiconductor) Physics and Technology, Vol. 1987, Wiley, New York [98] a) R. Misra, D. Depan, J. Shah, Phys. Chem. Chem. Phys. 2013, 15,
1982. 12988; b) I. Szleifer, R. Yerushalmi-Rozen, Polymer 2005, 46, 7803.
[77] a) J. T. Cheung, in Handbook of Vacuum Science and Technology, [99] A. Aqel, K. M. A. El-Nour, R. A. Ammar, A. Al-Warthan, Arabian
Elsevier, Amsterdam, Netherlands 1998, p. 694; b) A. Suresh, J. Chem. 2012, 5, 1.
P. Gollakota, P. Wellenius, A. Dhawan, J. F. Muth, Thin Solid Films [100] S. Park, M. Vosguerichian, Z. Bao, Nanoscale 2013, 5, 1727.
2008, 516, 1326; c) D. M. Mattox, Handbook of Physical Vapor [101] a) E. G. Gamaly, T. W. Ebbesen, Phys. Rev. B 1995, 52,
Deposition (PVD) Processing, Elsevier, Amsterdam, Netherlands, 2083; b) H. Lange, M. Sioda, A. Huczko, Y. Zhu, H. Kroto,
2010; d) X. Yu, T. J. Marks, A. Facchetti, Nat. Mater. 2016, 15, 383; D. Walton, Carbon 2003, 41, 1617; c) M. Keidar, A. M. Waas,
e) S. M. George, Chem. Rev. 2010, 110, 111; f) R. W. Johnson, Nanotechnology 2004, 15, 1571; d) N. Arora, N. Sharma, Diamond
A. Hultqvist, S. F. Bent, Mater. Today 2014, 17, 236; g) J. Sheng, Relat. Mater. 2014, 50, 135.
H.-J. Lee, S. Oh, J.-S. Park, ACS Appl. Mater. Interfaces 2016, 8, [102] a) M. Kumar, Y. Ando, J. Nanosci. Nanotechnol. 2010, 10, 3739;
33821; h) Y.-Y. Lin, C.-C. Hsu, M.-H. Tseng, J.-J. Shyue, F.-Y. Tsai, b) G. Che, B. Lakshmi, C. Martin, E. Fisher, R. S. Ruoff, Chem.
ACS Appl. Mater. Interfaces 2015, 7, 22610. Mater. 1998, 10, 260; c) Y.-L. Li, I. A. Kinloch, A. H. Windle, Science
[78] K. Nomura, H. Ohta, A. Takagi, T. Kamiya, M. Hirano, H. Hosono, 2004, 304, 276.
Nature 2004, 432, 488. [103] a) R. Vander Wal, G. Berger, T. Ticich, Appl. Phys. A: Mater.
[79] M.-G. Kim, H. S. Kim, Y.-G. Ha, J. He, M. G. Kanatzidis, Sci. Process. 2003, 77, 885; b) A. De Giacomo, A. De Bonis,
A. Facchetti, T. J. Marks, J. Am. Chem. Soc. 2010, 132, 10352. M. Dell’Aglio, O. De Pascale, R. Gaudiuso, S. Orlando,
[80] a) S. Hwang, J. H. Lee, C. H. Woo, J. Y. Lee, H. K. Cho, Thin Solid A. Santagata, G. Senesi, F. Taccogna, R. Teghil, J. Phys. Chem.
Films 2011, 519, 5146; b) Y. Wang, S. Liu, X. Sun, J. Zhao, G. Goh, C 2011, 115, 5123; c) Y. Sakakibara, S. Tatsuura, H. Kataura,
Q. Vu, H. Yu, J. Sol-Gel Sci. Technol. 2010, 55, 322. M. Tokumoto, Y. Achiba, Jpn. J. Appl. Phys. 2003, 42, L494;
[81] Y.-H. Kim, J.-S. Heo, T.-H. Kim, S. Park, M.-H. Yoon, J. Kim, d) T. Okada, J. Suehiro, Appl. Surf. Sci. 2007, 253, 7840.
M. S. Oh, G.-R. Yi, Y.-Y. Noh, S. K. Park, Nature 2012, 489, 128. [104] a) O. V. Yazyev, A. Pasquarello, Phys. Rev. Lett. 2008, 100,
[82] G. H. Kim, H. S. Shin, B. Du Ahn, K. H. Kim, W. J. Park, H. J. Kim, 156102; b) P. Nikolaev, J. Nanosci. Nanotechnol. 2004, 4, 307;
J. Electrochem. Soc. 2009, 156, H7. c) A. G. Nasibulin, D. P. Brown, P. Queipo, D. Gonzalez, H. Jiang,
[83] J. H. Lim, J. H. Shim, J. H. Choi, J. Joo, K. Park, H. Jeon, E. I. Kauppinen, Chem. Phys. Lett. 2006, 417, 179.
M. R. Moon, D. Jung, H. Kim, H.-J. Lee, Appl. Phys. Lett. 2009, 95, [105] Y. Ando, X. Zhao, T. Sugai, M. Kumar, Mater. Today 2004, 7, 22.
012108. [106] A. E. Islam, J. A. Rogers, M. A. Alam, Adv. Mater. 2015, 27, 7908.
[84] I. Yakimets, D. MacKerron, P. Giesen, K. J. Kilmartin, M. Goorhuis, [107] H. M. Manohara, E. W. Wong, E. Schlecht, B. D. Hunt, P. H. Siegel,
E. Meinders, W. MacDonald, Polymer Substrates for Flexible Nano Lett. 2005, 5, 1469.
Electronics: Achievements and Challenges, Vol. 93, Trans Tech [108] C. Lu, L. An, Q. Fu, J. Liu, H. Zhang, J. Murduck, Appl. Phys. Lett.
Publications, Reinhardstr, Zürich, Switzerland 2010. 2006, 88, 133501.
[85] G. H. Kim, H. S. Kim, H. S. Shin, B. D. Ahn, K. H. Kim, H. J. Kim, [109] M. C. Hersam, Nat. Nanotechnol. 2008, 3, 387.
Thin Solid Films 2009, 517, 4007. [110] a) S. Ghosh, S. M. Bachilo, R. B. Weisman, Nat. Nanotechnol.
[86] S. J. Kim, S. Yoon, H. J. Kim, Jpn. J. Appl. Phys. 2014, 53, 02BA02. 2010, 5, 443; b) A. A. Green, M. C. Hersam, Adv. Mater. 2011,
[87] B.-Y. Su, S.-Y. Chu, Y.-D. Juang, H.-C. Chen, Appl. Phys. Lett. 2013, 23, 2185; c) M. Ouyang, J.-L. Huang, C. M. Lieber, Acc. Chem.
102, 192101. Res. 2002, 35, 1018; d) Y. Cao, G. J. Brady, H. Gui, C. Rutherglen,
[88] I. Choudhary, Mater. Sci. Eng., B 2017, 218, 64. M. S. Arnold, C. Zhou, ACS Nano 2016, 10, 6782.
[89] Y. H. Lin, H. Faber, K. Zhao, Q. Wang, A. Amassian, M. McLachlan, [111] a) E. Cobas, M. S. Fuhrer, Appl. Phys. Lett. 2008, 93, 043120;
T. D. Anthopoulos, Adv. Mater. 2013, 25, 4340. b) E. D. Cobas, S. M. Anlage, M. S. Fuhrer, IEEE Trans. Microwave
[90] Y. H. Lin, H. Faber, J. G. Labram, E. Stratakis, L. Sygellou, Theory Tech. 2011, 59, 2726.
E. Kymakis, N. A. Hastas, R. Li, K. Zhao, A. Amassian, Adv. Sci. [112] L. Jiao, X. Xian, Z. Wu, J. Zhang, Z. Liu, Nano Lett. 2009, 9, 205.
2015, 2, 1500058. [113] R. Krupke, F. Hennrich, H. v. Löhneysen, M. M. Kappes, Science
[91] M.-G. Kim, M. G. Kanatzidis, A. Facchetti, T. J. Marks, Nat. Mater. 2003, 301, 344.
2011, 10, 382. [114] J. Li, Q. Zhang, D. Yang, J. Tian, Carbon 2004, 42, 2263.
[92] R. Want, IEEE Pervasive Comput. 2006, 5, 25. [115] M. Li, J. Zhang, presented at 12th IEEE International Conference on
[93] H. Park, H. Kang, Y. Lee, Y. Park, J. Noh, G. Cho, Nanotechnology Solid-State and Integrated Circuit Technology (ICSICT), Guilin, China
2012, 23, 344006. October 2014.
[94] H. Kang, H. Park, Y. Park, M. Jung, B. C. Kim, G. Wallace, G. Cho, [116] a) P. Stokes, S. I. Khondaker, Nanotechnology 2008, 19, 175202;
Sci. Rep. 2015, 4, 5387. b) S. Shekhar, P. Stokes, S. I. Khondaker, ACS Nano 2011, 5, 1739.
[95] a) T. W. Odom, J. L. HUANG, C. M. Lieber, Ann. N. Y. Acad. Sci. [117] a) N. T. Dinh, E. Sowade, T. Blaudeck, S. Hermann,
2002, 960, 203; b) T. W. Odom, J.-L. Huang, P. Kim, C. M. Lieber, R. D. Rodriguez, D. R. Zahn, S. E. Schulz, R. R. Baumann,
Nature 1998, 391, 62; c) S. Iijima, Nature 1991, 354, 56; O. Kanoun, Carbon 2016, 96, 382; b) T. Wang, M. A. Roberts,
d) E. T. Thostenson, T.-W. Chou, J. Phys. D: Appl. Phys. 2002, I. A. Kinloch, B. Derby, J. Phys. D: Appl. Phys. 2012, 45, 315304.
35, L77; e) S. Iijima, Phys. B 2002, 323, 1; f) T. Dürkop, S. Getty, [118] a) B. Chen, M. Gao, J. Zuo, S. Qu, B. Liu, Y. Huang, Appl. Phys.
E. Cobas, M. Fuhrer, Nano Lett. 2004, 4, 35; g) Z. Yao, C. L. Kane, Lett. 2003, 83, 3570; b) L. Girifalco, M. Hodak, R. S. Lee, Phys. Rev.

Adv. Sci. 2018, 1801653 1801653 (27 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

B 2000, 62, 13104; c) D. Tasis, N. Tagmatarchis, V. Georgakilas, [143] X. Li, X. Wang, L. Zhang, S. Lee, H. Dai, Science 2008, 319, 1229.
M. Prato, Chem. - Eur. J. 2003, 9, 4000. [144] M. Y. Han, B. Özyilmaz, Y. Zhang, P. Kim, Phys. Rev. Lett. 2007, 98,
[119] a) R. Rastogi, R. Kaushal, S. Tripathi, A. L. Sharma, I. Kaur, 206805.
L. M. Bharadwaj, J. Colloid Interface Sci. 2008, 328, 421; b) J. Li, [145] Y. Lu, Y. Feng, J. Phys. Chem. C 2009, 113, 20841.
P. C. Ma, W. S. Chow, C. K. To, B. Z. Tang, J. K. Kim, Adv. Funct. [146] M. Dragoman, G. Deligeorgis, A. Muller, A. Cismaru, D. Neculoiu,
Mater. 2007, 17, 3207. G. Konstantinidis, D. Dragoman, A. Dinescu, F. Comanescu,
[120] L. Jiang, L. Gao, J. Sun, J. Colloid Interface Sci. 2003, 260, 89. J. Appl. Phys. 2012, 112, 084302.
[121] a) G. Pagani, M. J. Green, P. Poulin, M. Pasquali, Proc. Natl. Acad. [147] A. Kaur, X. Yang, K. Y. Park, P. Chahal, presented at IEEE 63rd
Sci. USA 2012, 109, 11599; b) K. Lu, R. Lago, Y. Chen, M. Green, Electronic Components and Technology Conference, Las Vegas,
P. Harris, S. Tsang, Carbon 1996, 34, 814. Nevada, USA, May 2013.
[122] D. S. Hecht, L. Hu, G. Irvin, Adv. Mater. 2011, 23, 1482. [148] A. P. Alivisatos, Science 1996, 271, 933.
[123] a) Z. Spitalsky, D. Tasis, K. Papagelis, C. Galiotis, Prog. Polym. Sci. [149] L. Qian, Y. Yang, W. Cao, C. Xiang, X. Xie, Z. Liu, S. Chen, L. Wu,
2010, 35, 357; b) K. Balasubramanian, M. Burghard, Small 2005, X. Yan, presented at SID Symposium Digest of Technical Papers,
1, 180. Vol. 48, May 2017, pp. 55–57.
[124] J. Hilding, E. A. Grulke, Z. George Zhang, F. Lockwood, [150] J. Chen, D. Zhao, C. Li, F. Xu, W. Lei, L. Sun, A. Nathan, X. W. Sun,
J. Dispersion Sci. Technol. 2003, 24, 1. Sci. Rep. 2015, 4, 4085.
[125] L. Vaisman, H. D. Wagner, G. Marom, Adv. Colloid Interface Sci. [151] a) Y. Liu, F. Li, Z. Xu, C. Zheng, T. Guo, X. Xie, L. Qian, D. Fu,
2006, 128–130, 37. X. Yan, ACS Appl. Mater. Interfaces 2017, 9, 25506; b) X. Dai,
[126] a) S. K. Eshkalak, A. Chinnappan, W. Jayathilaka, M. Khatibzadeh, Y. Deng, X. Peng, Y. Jin, Adv. Mater. 2017, 29.
E. Kowsari, S. Ramakrishna, Appl. Mater. Today 2017, 9, 372; [152] G. M. Vanacore, J. Hu, W. Liang, S. Bietti, S. Sanguinetti,
b) S. Rana, R. Prasanth, L. P. Tan, in Handbook of Polymer F. Carbone, A. H. Zewail, Struct. Dyn. 2017, 4, 044034.
Nanocomposites: Processing, Performance and Application, Springer, [153] P. Wang, Y. Zhang, C. Ruan, L. Su, H. Cui, W. Y. William, IEEE J.
New York, USA 2015, p. 103; c) T. Hertel, A. Hagen, V. Talalaev, Sel. Top. Quantum Electron. 2017, 23, 1.
K. Arnold, F. Hennrich, M. Kappes, S. Rosenthal, J. McBride, [154] L. Kim, P. O. Anikeeva, S. A. Coe-Sullivan, J. S. Steckel,
H. Ulbricht, E. Flahaut, Nano Lett. 2005, 5, 511. M. G. Bawendi, V. Bulovic, Nano Lett. 2008, 8, 4513.
[127] Y. Qin, H.-J. Kwon, A. Subrahmanyam, M. M. R. Howlader, [155] W. Brütting, Semiconductors 2005, 6, 11.
P. R. Selvaganapathy, A. Adronov, M. J. Deen, Mater. Lett. 2016, [156] V. Coropceanu, J. Cornil, D. A. da Silva Filho, Y. Olivier, R. Silbey,
176, 68. J.-L. Brédas, Chem. Rev. 2007, 107, 926.
[128] K. Wang, Y.-H. Chang, C. Zhang, B. Wang, Carbon 2016, 98, 397. [157] a) A. Salleo, Mater. Today 2007, 10, 38; b) R. Coehoorn, W. Pasveer,
[129] P. H. Lau, K. Takei, C. Wang, Y. Ju, J. Kim, Z. Yu, T. Takahashi, P. Bobbert, M. Michels, Phys. Rev. B 2005, 72, 155206.
G. Cho, A. Javey, Nano Lett. 2013, 13, 3864. [158] K.-C. Kao, W. Hwang, Electrical Transport in Solids: With Particular
[130] X. Cao, H. Chen, X. Gu, B. Liu, W. Wang, Y. Cao, F. Wu, C. Zhou, Reference to Organic Semiconductors, Vol. 14, Elsevier, Oxford, UK
ACS Nano 2014, 8, 12769. 1981.
[131] A. Abdelhalim, A. Abdellah, G. Scarpa, P. Lugli, Carbon 2013, 61, [159] C. Reese, Z. Bao, Mater. Today 2007, 10, 20.
72. [160] C. Mitsui, T. Okamoto, M. Yamagishi, J. Tsurumi, K. Yoshimoto,
[132] L. Cai, C. Wang, Flexible Stretchable Med. Devices 2018, 7. K. Nakahara, J. Soeda, Y. Hirose, H. Sato, A. Yamano, Adv. Mater.
[133] P. M. Grubb, H. Subbaraman, S. Park, D. Akinwande, R. T. Chen, 2014, 26, 4546.
Sci. Rep. 2017, 7, 1202. [161] a) E. Cantatore, T. C. Geuns, G. H. Gelinck, E. van Veenendaal,
[134] X. Yang, P. Chahal, presented at IEEE 61st Electronic Components A. F. Gruijthuijsen, L. Schrijnemakers, S. Drews, D. M. de Leeuw,
and Technology Conference (ECTC), Lake Buena Vista, FL, USA, 31 IEEE J. Solid-State Circuits 2007, 42, 84; b) P. T. Herwig, K. Müllen,
May–3 June 2011. Adv. Mater. 1999, 11, 480.
[135] A. Kaur, X. Yang, P. Chahal, IEEE Trans. Compon., Packag., Manuf. [162] a) J. E. Anthony, J. S. Brooks, D. L. Eaton, S. R. Parkin, J. Am. Chem.
Technol. 2016, 6, 1766. Soc. 2001, 123, 9482; b) M. M. Payne, S. R. Parkin, J. E. Anthony,
[136] a) A. C. Neto, F. Guinea, N. M. Peres, K. S. Novoselov, A. K. Geim, C.-C. Kuo, T. N. Jackson, J. Am. Chem. Soc. 2005, 127, 4986.
Rev. Mod. Phys. 2009, 81, 109; b) R. Cheng, J. Bai, L. Liao, H. Zhou, [163] B. C. Kjellander, W. T. Smaal, K. Myny, J. Genoe, W. Dehaene,
Y. Chen, L. Liu, Y.-C. Lin, S. Jiang, Y. Huang, X. Duan, Proc. Natl. P. Heremans, G. H. Gelinck, Org. Electron. 2013, 14, 768.
Acad. Sci. USA 2012, 109, 11588. [164] N. Shin, J. Kang, L. J. Richter, V. M. Prabhu, R. J. Kline,
[137] a) M. Yi, Z. Shen, J. Mater. Chem. A 2015, 3, 11700; b) M. Lotya, D. A. Fischer, D. M. DeLongchamp, M. F. Toney, S. K. Satija,
Y. Hernandez, P. J. King, R. J. Smith, V. Nicolosi, L. S. Karlsson, D. J. Gundlach, Adv. Funct. Mater. 2013, 23, 366.
F. M. Blighe, S. De, Z. Wang, I. McGovern, J. Am. Chem. Soc. [165] M. R. Niazi, R. Li, E. Q. Li, A. R. Kirmani, M. Abdelsamie, Q. Wang,
2009, 131, 3611; c) H. Wang, J. T. Robinson, X. Li, H. Dai, J. Am. W. Pan, M. M. Payne, J. E. Anthony, D.-M. Smilgies, Nat. Commun.
Chem. Soc. 2009, 131, 9910. 2015, 6, 8598.
[138] W. Choi, I. Lahiri, R. Seelaboyina, Y. S. Kang, Crit. Rev. Solid State [166] D. K. Hwang, C. Fuentes-Hernandez, J. D. Berrigan, Y. Fang,
Mater. Sci. 2010, 35, 52. J. Kim, W. J. Potscavage, H. Cheun, K. H. Sandhage, B. Kippelen,
[139] N. O. Weiss, H. Zhou, L. Liao, Y. Liu, S. Jiang, Y. Huang, X. Duan, J. Mater. Chem. 2012, 22, 5531.
Adv. Mater. 2012, 24, 5782. [167] a) O. Ostroverkhova, D. Cooke, S. Shcherbyna, R. Egerton,
[140] K. Emtsev, F. Speck, T. Seyller, L. Ley, J. D. Riley, Phys. Rev. B 2008, F. Hegmann, R. Tykwinski, J. Anthony, Phys. Rev. B 2005,
77, 155303. 71, 035204; b) M. Orrit, J. Bernard, Phys. Rev. Lett. 1990, 65,
[141] Y. Hernandez, V. Nicolosi, M. Lotya, F. M. Blighe, Z. Sun, 2716.
S. De, I. McGovern, B. Holland, M. Byrne, Y. K. Gun’Ko, Nat. [168] C.-m. Kang, J. Wade, S. Yun, J. Lim, H. Cho, J. Roh, H. Lee,
Nanotechnol. 2008, 3, 563. S. Nam, D. D. C. Bradley, J.-S. Kim, C. Lee, Adv. Electron. Mater.
[142] a) C. Gómez-Navarro, R. T. Weitz, A. M. Bittner, M. Scolari, 2016, 2, 1500282.
A. Mews, M. Burghard, K. Kern, Nano Lett. 2007, 7, 3499; [169] S. G. Higgins, B. V. O. Muir, J. Wade, J. Chen, B. Striedinger,
b) G. Eda, G. Fanchini, M. Chhowalla, Nat. Nanotechnol. 2008, 3, H. Gold, B. Stadlober, M. Caironi, J.-S. Kim, J. H. G. Steinke,
270. A. J. Campbell, Adv. Electron. Mater. 2015, 1, 1500024.

Adv. Sci. 2018, 1801653 1801653 (28 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.advancedsciencenews.com www.advancedscience.com

[170] H. Kroto, J. Heath, S. O’Brien, R. Curl, R. Smalley, Nature 1985, pp. 1196–1199; c) M. Montanino, G. Sico, C. Prontera,
318, 162. A. De Girolamo Del Mauro, S. Aprano, M. Maglione, C. Minarini,
[171] A. D. de Zerio Mendaza, A. Melianas, S. Rossbauer, O. Bäcke, eXPRESS Polym. Lett. 2017, 11, 518.
L. Nordstierna, P. Erhart, E. Olsson, T. D. Anthopoulos, [186] S. E. Habas, H. A. Platt, M. F. Van Hest, D. S. Ginley, Chem. Rev.
O. Inganäs, C. Müller, Adv. Mater. 2015, 27, 7325. 2010, 110, 6571.
[172] J. Smith, W. Zhang, R. Sougrat, K. Zhao, R. Li, D. Cha, [187] a) S. Shaheen, D. Ginley, G. Jabbour, MRS Bull. 2005, 30,
A. Amassian, M. Heeney, I. McCulloch, T. D. Anthopoulos, Adv. 10; b) M. Finn III, C. J. Martens, A. V. Zaretski, B. Roth,
Mater. 2012, 24, 2441. R. R. Søndergaard, F. C. Krebs, D. J. Lipomi, Sol. Energy Mater. Sol.
[173] M. Uno, Y. Kanaoka, B.-S. Cha, N. Isahaya, M. Sakai, H. Matsui, Cells 2018, 174, 7.
C. Mitsui, T. Okamoto, J. Takeya, T. Kato, M. Katayama, Y. Usami, [188] H. Youn, H. J. Park, L. J. Guo, Small 2015, 11, 2228.
T. Yamakami, Adv. Electron. Mater. 2015, 1, 1500178. [189] a) J. Y. Lee, T. Lee, H. J. Park, L. J. Guo, Org. Electron. 2014,
[174] K. E. Lilja, T. G. Bäcklund, D. Lupo, T. Hassinen, T. Joutsenoja, 15, 2710; b) F. C. Krebs, Sol. Energy Mater. Sol. Cells 2009, 93,
Org. Electron. 2009, 10, 1011. 394.
[175] C.-Y. Lin, C.-H. Tsai, H.-T. Lin, L.-C. Chang, Y.-H. Yeh, Z. Pei, [190] Y. Yang, Z.-G. Zhang, H. Bin, S. Chen, L. Gao, L. Xue, C. Yang,
Y.-R. Peng, C.-C. Wu, Org. Electron. 2011, 12, 1777. Y. Li, J. Am. Chem. Soc. 2016, 138, 15011.
[176] P. Heljo, K. E. Lilja, H. S. Majumdar, D. Lupo, Org. Electron. 2014, [191] S. Agarwala, G. L. Goh, W. Y. Yeong, presented at IOP Conference
15, 306. Series: Materials Science and Engineering, Beijing, China, October
[177] S. G. Higgins, T. Agostinelli, S. Markham, R. Whiteman, 2017.
H. Sirringhaus, Adv. Mater. 2017, 29, 1703782. [192] D. G. Georgiadou, J. Semple, G. Wyatt-Moon, T. D. Anthopoulos,
[178] K. Y. Mitra, C. Sternkiker, C. Martínez-Domingo, E. Sowade, presented at SPIE Organic Photonics + Electronics, San Diego, CA,
E. Ramon, J. Carrabina, H. L. Gomes, R. R. Baumann, Flexible USA, August 2016.
Printed Electron. 2017, 2, 015003. [193] Y. Y. Huang, E. M. Terentjev, Polymers 2012, 4, 275.
[179] R. D. Jansen-van Vuuren, A. Armin, A. K. Pandey, P. L. Burn, [194] C. Backes, B. M. Szydłowska, A. Harvey, S. Yuan, V. Vega-Mayoral,
P. Meredith, Adv. Mater. 2016, 28, 4766. B. R. Davies, P.-l. Zhao, D. Hanlon, E. J. Santos, M. I. Katsnelson,
[180] R. Eckstein, T. Rödlmeier, T. Glaser, S. Valouch, R. Mauer, ACS Nano 2016, 10, 1589.
U. Lemmer, G. Hernandez-Sosa, Adv. Electron. Mater. 2015, 1, [195] A. A. Green, M. C. Hersam, Nano Lett. 2009, 9, 4031.
1500101. [196] T.-H. Kim, K.-S. Cho, E. K. Lee, S. J. Lee, J. Chae, J. W. Kim,
[181] G. Azzellino, A. Grimoldi, M. Binda, M. Caironi, D. Natali, D. H. Kim, J.-Y. Kwon, G. Amaratunga, S. Y. Lee, B. L. Choi, Y. Kuk,
M. Sampietro, Adv. Mater. 2013, 25, 6829. J. M. Kim, K. Kim, Nat. Photonics 2011, 5, 176.
[182] K. J. Baeg, M. Binda, D. Natali, M. Caironi, Y. Y. Noh, Adv. Mater. [197] R. Theissmann, S. Bubel, M. Sanlialp, C. Busch, G. Schierning,
2013, 25, 4267. R. Schmechel, Thin Solid Films 2011, 519, 5623.
[183] L. Dongwon, C. JinKoo, R. Jungsoo, Dig. Tech. Pap. - Soc. Inf. Disp. [198] P. K. Nayak, M. N. Hedhili, D. Cha, H. N. Alshareef, Appl. Phys.
Int. Symp. 2005, 36, 527. Lett. 2012, 100, 202106.
[184] P. Kopola, M. Tuomikoski, R. Suhonen, A. Maaninen, Thin Solid [199] J. Zhang, Y. Li, B. Zhang, H. Wang, Q. Xin, A. Song, Nat. Commun.
Films 2009, 517, 5757. 2015, 6, 7561.
[185] a) C. Sekine, Y. Tsubata, T. Yamada, M. Kitano, S. Doi, Sci. [200] C. S. Jones, X. Lu, M. Renn, M. Stroder, W.-S. Shih, Microelectron.
Technol. Adv. Mater. 2014, 15, 034203; b) H. Nakajima, S. Morito, Eng. 2010, 87, 434.
H. Nakajima, T. Takeda, M. Kadowaki, K. Kuba, S. Handa, [201] J. Vaillancourt, H. Zhang, P. Vasinajindakaw, H. Xia, X. Lu, X. Han,
D. Aoki, presented at SID Symposium Digest of Technical Papers, D. C. Janzen, W.-S. Shih, C. S. Jones, M. Stroder, Appl. Phys. Lett.
Vol. 36, Blackwell Publishing Ltd. Oxford, UK, May 2005, 2008, 93, 243301.

Adv. Sci. 2018, 1801653 1801653 (29 of 29) © 2018 The Authors. Published by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim