CHAPTER 1: Introduction and Fundamental Principles of Ambipolar Materials

In the modern era, extensive research in the emerging internet of things and artificial intelligence has provided significant impetus for technological advances in information networks. In order to achieve the complicated functions for the internal circuits required, various stable materials and primary circuit elements are essential as basic functional modules. Therefore, exploration of novel multifunctional materials and devices, as well as miniaturization and high density integration of functional materials and devices, has been the goal of scientific researchers. When incorporating inherent materials into electronic devices, the charge transport properties can be evaluated. In terms of their electrical properties and the principal conductive carriers in devices, materials or devices can be classified into two types: unipolar and ambipolar. Unipolar devices include p-type and n-type devices, where the predominant conductive charges are holes and electrons, respectively. This type of device exhibits relatively unitary electrical characteristics and hence their application ranges have certain limitations. Ambipolar materials and devices, however, can realize comparable simultaneous transfer of electrons and holes, and thereby display p-type and n-type characteristics within a single device, which makes them useful in many different fields.^1,2  In the last few years, diverse ambipolar materials, such as organic materials,³  carbon nanotubes,⁴  two-dimensional (2D)⁵  and perovskite materials,⁶  and various device architectures, for instance, bilayer⁷  and blended structures,⁸  have been produced to achieve ambipolar carrier conduction and further applications, comprising solar cells,⁹  logic devices,¹⁰  neuromorphic devices,¹¹  light-emitting transistors (LETs),¹²  gas sensors¹³  and ambipolar flash memory¹⁴  (see Figure 1.1). Ambipolar materials and devices have been investigated and applied in printed electronics owing to the fact that their intrinsic properties, with concurrent transport of holes and electrons, can greatly enhance the electrical performance of devices, reduce the fabrication complexity, as well as generate novel electrical and optoelectronic phenomena.

This book focuses on ambipolar materials and devices. The outline of this book is described below. An introductory description on the fundamental properties and theories of ambipolar materials and devices is presented first. Then, the synthesis and properties of various state-of-the-art ambipolar materials (e.g., organic polymers and 2D materials), the operation principles and electrical performance of their devices, as well as several ambipolar device architectures are reviewed and analyzed in detail. Within these described material and device systems, despite inherent challenges, the fundamentals and physics of the ambipolar materials and devices have been well understood and thereby numerous functional applications have begun to emerge. Therefore, many applications based on ambipolar devices, for instance, gas sensors and LETs, are also discussed. Lastly, we point out the remaining challenges for material design and device fabrication that are faced in achieving ambipolar charge transport, and propose several possible solution strategies.

Organic materials, including small molecules and polymers, can be engineered and modified to acquire specific energy levels and charge transport characteristics.¹⁵  Besides, organic materials are also readily scaled up and can be separated with relatively high purity. Thus, organic materials, such as conjugated polymers containing isoindigos and diketopyrrolopyrroles (DPPs), can be utilized to realize good ambipolar behaviors, with large charge mobility and on/off ratio, when exploited as semiconductors in thin-film transistors.^16,17  In order to obtain ambipolar carrier conduction, both types of charge carriers should be injected effectively from the metal electrodes to the organic semiconducting materials.¹  This means that the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energy levels of the organic semiconductors are supposed to be aligned with the work function of the metal electrodes for efficient hole and electron injection, respectively.²  However, most organic materials still present unipolar phenomena (mostly hole-dominated p-type behavior) because they possess relatively large energy gaps (2–3 eV) and hence at least one type of charge carrier cannot be injected efficiently for given specific metal electrodes.¹⁸  In this regard, organic materials with a considerably smaller energy gap of less than 2 eV are favorable for realizing a decreased injection barrier for holes and electrons and, thereby, ambipolar charge transport. In addition, a narrow band gap with a lowered LUMO level is also capable of improving the air stability of ambipolar materials significantly, as a result of the reduced captured charges by O₂ or H₂O in ambient air.¹⁹  Furthermore, the morphology of organic thin films (with various molecular packing modes and crystallinities) can also impact the ambipolar performance, for instance, the mobilities of charge carriers.²⁰  For the sake of forming a superior morphology for the organic active layer, the appropriate design of molecular structures,²¹  optimized fabrication methods and preparation conditions,²²  as well as substrate modification,²³  are all essential.

Semiconducting single-walled carbon nanotubes (SWNTs) possess remarkable mobilities of charges, and energy gaps of approximately 0.7–1 eV dominated by their intrinsic size.²⁴  Additionally, they are intrinsic semiconductors with a direct band gap property, which makes it possible for them to transport both two types of charge carriers. The SWNT networks can be manufactured via utilizing various approaches, such as chemical vapor deposition and solution deposition.²⁵  For the latter, different deposition approaches give rise to plentiful and various carbon nanotubes. Thus selective separation of the metallic nanotubes and realization of SWNTs with single chirality and regulated length are both necessary to acquire pure and well-dispersed nanotube inks. So far, many effective post-synthetic separation methods have been proposed to sort and purify the nanotubes, for instance, polymer sorting,²⁶  gel chromatography,²⁷  density gradient ultracentrifugation,²⁸  DNA-based separation²⁹  and electrophoresis.³⁰  Among them, selective solubilization by polymer sorting is more efficient and the backbones, as well as the side chains, of organic polymeric materials can select and purify the nanotubes based on the non-covalent interactions between them.^31,32  However, intrinsic ambipolar nanotubes generally exhibit hole accumulation behavior under ambient conditions by reason of the existence of abundant H₂O and O₂, the interface capture of electrons induced by hydroxyl groups, as well as the formation of a relatively large energy barrier for electron injection. Thus, several effective strategies, such as usage of high k-blocking dielectric,³³  n-type doping,³⁴  suitable electrodes with relatively little contact resistance,³⁵  as well as ionic gating,³⁶  have been put forward to achieve ambipolar carrier conduction. In addition, a dual-gate configuration, composed of original and control gates, is also effective and can regulate and manage the longitudinal electrical field and hence the internal electrons and holes, which finally results in adjustable charge polarity.³⁷ 

Two-dimensional layered materials, for instance, transition metal dichalcogenides (TMDs) and graphene, have been extensively studied and exploited as significant functional layers within various electronic and optoelectronic devices as a result of their superior optoelectronic and mechanical properties.⁹  Besides, because the effective mass of electrons can be ignored in a Fermi surface, relatively large mobilities of both types of charge carriers can be achieved in 2D monolayer materials. Moreover, the basic energy gap of 2D materials can be effectively adjusted while the number of layers of the 2D material gradually changes. Furthermore, due to the fact that the atoms of 2D monolayer materials are exposed under ambient environment, the electrical and optical characteristics of 2D materials can also be tuned easily utilizing diverse approaches, such as chemical doping,³⁸  interface modification³⁹  and surface treatment.⁴⁰  All these advantages make 2D materials very suitable for the realization of ambipolar charge transport with large charge mobility and on/off ratio. During the past few years, plenty of strategies have been employed to realize and control ambipolar carrier conduction, for instance, electric double layer,⁴¹  surface charge transfer doping³⁸  and electrostatic doping.⁵  As previously stated, the work function of electrode materials, the energy gap of 2D semiconducting materials (related to the number of layers of 2D materials), as well as the interface characteristics, are all essential for high-performance ambipolar behavior. The Schottky barrier formed between 2D materials and electrode materials and the Fermi level offset caused by doping and gate bias stress are the most frequently utilized models to interpret ambipolar phenomena in 2D materials.^42,43 

Organic–inorganic hybrid perovskite materials integrate the favorable nature of organic materials and inorganic solids into just one composite, with molecular scale. The inorganic constituent induces an expanding framework, which is controlled via intense ionic or covalent mutual effects and hence ensures relatively large charge mobilities. The organic constituent promotes the self-assembly capability of the hybrids and thereby makes them easily prepared. In addition, the organic constituent can modulate the electrical characteristics of the extended framework by means of regulating the decreased dimensionality and adjusting the electronic coupling of different inorganic components. Charge mobilities and regulation of conductance can be maximized by appropriately designing the hybrid perovskite materials at the molecular scale. Incorporation of large charge mobilities and facile preparation in organic–inorganic hybrids makes them promising candidates for use in electronic devices, for good ambipolar carrier conduction with high carrier mobilities.⁴⁴  So far, various hybrid perovskite materials, such as CH₃NH₃PbI₃⁶ and formamidinium lead iodide (FAPbI₃)⁴⁵  have been exploited to manufacture field-effect transistors with well-balanced carrier mobilities of more than 2 cm² V⁻¹ s⁻¹. The constituents and structure of the organic–inorganic hybrid, the grain size of the perovskite, as well as the interface properties of the insulating layer and hybrid, have all been found to have significant influence on the ambipolar charge transport and the mobilities of both types of charge carriers.

Apart from these four types of single-component materials for effective ambipolar charge transport, combinations of different types of materials can also be employed to implement ambipolar behavior. One strategy is to blend two or three unipolar (p-type or n-type) materials and then form a single layer by solution preparation or vacuum evaporation.⁴⁶  The p-type and n-type materials are responsible for hole and electron conduction, respectively, to realize overall ambipolar performance. The blending ratio can be adjusted to modulate the ambipolar carrier conduction and the mixing microscopic morphology should be well controlled to achieve efficient transfer of electrons and holes, and thus high carrier mobilities. Another strategy is to deposit p-type and n-type materials successively to form a bilayer architecture, where these two materials control the motion of holes and electrons separately.⁷  In contrast to the blending structure, this method is more efficient for completely isolated transport of holes and electrons. The deposition sequence, respective layer thickness, as well as the interface quality between the p- and n-type materials, require elaborate design to display high-performance ambipolar behavior.

Until now, organic solar cells and perovskite solar cells have been the most widely investigated solar cells, as organic solar cells can generate good efficiency at low cost and perovskite solar cells can achieve longer charge diffusion length, larger power conversion efficiency, higher throughput, as well as minimized variation between different batches.^47,48  The prominent features of 2D materials, such as outstanding flexibility and transparency, large conductance, excellent scalability, as well as roll-to-roll manufacture, enable them to be utilized and integrated into solar photovoltaics.⁴⁹  Ambipolar graphene can be exploited as an electrode material in solar cells on the basis of its transparent and conductive nature, and can also function as both positive and negative electrodes on account of its concurrent electron and hole transport.⁵⁰  Other monolayer 2D materials apart from graphene, for instance, TMDs, possess direct band gap characteristics and can behave as active layers and interfacial layers to modulate the extraction of holes and electrons.⁵¹  In addition, 2D materials are also capable of being employed as a third constituent within organic solar cells through utilizing materials that possess complementary light absorption as the second acceptor or donor,⁵²  for efficient acquisition of photons with different energy states (enhancing the light absorption of active organic layers) and as stabilizing agents to increase the environmental stability of perovskite materials.

Complementary metal–oxide–semiconductor (CMOS) inverters are elementary constituents within state-of-the-art integrated circuits. CMOS inverters with superior performance are made up of hole-dominated and electron-dominated field-effect transistors (FETs) that possess similar charge mobilities and reaction rates, as well as similar threshold voltages. During the working process, the hole-dominated and electron-dominated transistors are concatenated at their gate and drain the electrodes, which function as input and output terminals. The source electrodes of the driver (usually n-type) and load (usually p-type) FETs are linked with the ground and electrical power, respectively. The states of the hole-dominated and electron-dominated transistors are capable of being controlled via modulating the input bias, and then NOT logic can be realized within complementary inverters. CMOS inverters are generally fabricated by utilizing p- and n-type metal–oxide–semiconductor FETs (MOSFETs) within a silicon platform in which the doping of electrons and holes coexists and is performed in various spatial regions. To manufacture CMOS inverters on the basis of FETs of various p- and n-type materials, preparation procedures are more sophisticated due to the integration of two divided unipolar transistor devices. However, fabrication complexity and expense could be drastically reduced by exploiting ambipolar materials and transistors to build CMOS inverters and modulating a single device to exhibit unipolar electron or hole conduction. In order to build high-performance complementary inverters, the fundamental ambipolar transistors should possess balanced high carrier mobilities (to guarantee normal Z-shaped electronic characteristics, high DC gain and noise immunity)²¹  and comparable turn-on bias under electron-enhancement and hole-enhancement mode (to ensure a comparatively small potentiating voltage and thereby decreased energy dissipation),⁵³  as well as low off-current in any unipolar operation mode (to ensure relatively low power consumption and large gain value).⁵⁴  Therefore, it is indispensable to realize selective transfer of electrons and holes and hence selective high-performance unipolar mode of ambipolar transistors. A sequence of methods, such as modification of insulating layer and metal electrodes, insertion of carrier infusion layers and utilization of self-assembled monolayers, has been shown to obtain targeted conduction polarity.^55,56  In addition, chemical doping can also be employed to tune carrier injection, carrier mobilities and, thus, the transporting polarity of ambipolar transistors.⁵⁷  Furthermore, multiple gates, for instance, tri-gate and split-gate configurations, are also valid for efficient electrostatic regulation of charge carriers and hence of device polarity.⁵⁸  This strategy is reversible when the electric field is removed and the intervals between various gates should be carefully engineered (small intervals are better).

Chemical synapses are specific cellular connections between neurons and key structures for physiological activities. Chemical information can be propagated in the nervous system via the release and reception of neurotransmitters. At the same time, the chemical information also can be saved and disposed synchronously within the identical biological synapse via modulating the synaptic weight, and this process is mutual and self-adapting. This extraordinary feature of chemical synapses (known as synaptic plasticity) furnishes a more efficient and valid method to implement information processing, which is distinct from the conventional von Neumann structure where processor and memory are spatially disconnected.⁵⁹  Thus, biological synaptic emulation is indispensable for building high-performance neuromorphic computing systems. Fortunately, ambipolar devices (mainly ambipolar transistors) can serve as fundamental functional blocks to mimic diverse chemical synaptic behaviors, such as excitatory postsynaptic current, paired-pulse facilitation, spike-rate-dependent plasticity and spike-time-dependent plasticity.^11,60  During operation, successive electrical pulses are applied on the gate electrode (to emulate the stimuli exerted on the presynaptic membrane) to modulate the carrier concentration in the semiconducting channel and hence realize the learning function of the organism, while the source drain current functions as the postsynaptic current to achieve the signal transmission function (to mimic the resultant potentiation or depression of the postsynaptic membrane).⁶¹  In contrast to unipolar transistors, ambipolar transistors can realize more efficient trapping/detrapping of both types of charge carriers and thereby induce wider threshold voltage shifts, which results in increased scope for conductance modulation and hence augmented tunable synaptic plasticity.⁶²  In addition, the injected charges, which possess inverse polarity, are capable of neutralizing the preceding trapped carriers and thereby significantly decreasing power consumption. Furthermore, owing to the attainable p- and n-type operation modes within ambipolar transistors, they can also be utilized to implement dynamic reconfigurable synaptic plasticity and achieve completely reversed reaction under the same electrical pulse, which eventually can simulate different body sensations under various environmental conditions.^63,64 

Ambipolar LETs refer to novel multifunctional configurations that combine the electrical switching property of transistors with the light-emitting ability of light-emitting diodes (LEDs). The light-emission capability is ascribed to the recombination of positive and negative charge carriers within the semiconductor.⁶⁵  In theory, each ambipolar field-effect transistor can behave as a light-emitting device and the electrical characteristics of the semiconductor materials, such as quantum efficiency and current density, dominate the whole luminous intensity within the semiconducting channel. In comparison with traditional LED geometry, ambipolar LET architecture possesses plenty of merits, for instance, greater channel conductance,⁶⁶  gate-regulated well-balanced positive and negative charges, gate-controlled emission-zone location,⁶⁷  prominently decreased loss of excitons, as well as greater external quantum efficiency.⁶⁸  Therefore, ambipolar LET geometry is extremely promising and attractive for utilization to achieve electrically driven organic lasers, which have many advantages compared to traditional semiconductor lasers, such as greater laser efficiency, wider emissive spectral scope and more possibilities for integration. In the meantime, ambipolar LETs can also be exploited to integrate with some other optical configurations (e.g., a distributed Bragg reflector microcavity).⁶⁹  In order to obtain high-performance ambipolar LETs for electrically driven lasers, various strategies have been employed, for instance, current crowding,⁷⁰  spin-polarized injection,⁷¹  modulation of recombination dimension,⁷²  external magnetic effect⁷³  and microcavity polariton lasing,⁷⁴  to improve the transistor geometry and operation and thereby augment the ampere density, as well as the formation of singlet excitons. In addition, the planar geometry of the ambipolar light-emitting devices can be utilized to investigate the charge influx efficiency and the recombination procedure within the semiconducting materials by studying the physical properties (e.g., width, shape, etc.) of the recombination zone.^75,76  Moreover, ambipolar LETs are also capable of serving as surface-susceptible probes. Both the configurational morphology of semiconductors and the existing carrier traps can be analyzed by observing the shift of the light-emitting area,⁷⁷  which provides an effective approach to research the carrier transfer process within the semiconducting channel.⁷⁸ 

Although the investigation of ambipolar materials in organic electronic devices is extensive, their applications in conductometric gas sensors are largely ignored and extremely rare. When chemical substances (e.g., water, oxygen, ethanol and ammonia) within the surrounding environment come into contact with ambipolar materials or devices, the observed device current under p- or n-channel accumulation mode usually changes and thus ambipolar devices can also function as conductometric gas sensors.¹³  For example, a hybrid ambipolar transistor consisting of zinc oxide and pentacene as n- and p-type semiconducting materials was utilized to sense ethanol steam.⁷⁹  The device current gradually reduced or increased under hole-enhancement or electron-enhancement mode, respectively, when the ethanol vapor was introduced. In this situation, ambipolar materials or devices could implement chemosensing properties with two modes of operation. Another case is ambipolar transistors composed of p- and n-type phthalocyanines. It was found that the p/n bilayer structure was more susceptible to ethanol vapor, in comparison with the counterpart n/p bilayer architecture, in spite of the fact that the device current could be augmented in either case.⁸⁰  These devices could also present reduced conductance in ammonia atmosphere, but the sensitivity exhibited was relatively low. These examples illustrate the great potential of ambipolar materials or devices for chemosensing, and various kinds of ambipolar materials and structures can sense different gases with diverse sensitivities.

Non-volatile memory devices based on FET structures have been researched due to the fact that they are compatible with integrated circuits and the present CMOS technology, and they can also achieve non-destructive read-out and single-transistor implementation.^2,81,82  The operation of transistor-type memories is based on the regulation of the shift of threshold voltage. In order to achieve and optimize the adjustable threshold voltage, numerous studies have utilized a floating gate (e.g., metal nanoparticles and fullerene),^11,14,83  a charge-trap polymer electret (e.g., poly(vinyl alcohol) and poly(2-vinyl naphthalene)),^84,85  as well as ferroelectric materials with polarizable characteristics (e.g., poly(vinylidenefluoride-co-trifluoroethylene)).⁸⁶  Generally speaking, the threshold voltage can only be transferred to one side of the initial state, which is often called unipolar transistor memory due to the induced unipolar charge trapping. However, for non-volatile bipolar transistor memories, their threshold voltages can be shifted to the left and right sides of the initial state under opposite longitudinal voltage polarities, and this ambipolar charge-storing behavior can be realized by exploiting materials with ambipolar charge trapping or transporting properties and by adjusting the interface of various functional layers.^14,87  Seemingly, in comparison with unipolar memories, bipolar transistor memories can exhibit a greater memory window by virtue of their two-way offset of the threshold voltage, and hence provide easier implementation of multi-bit data storage. These advantages make ambipolar memory favorable and attractive for use to miniaturize state-of-the-art commercial memories and so augment their storage density and capacity.