Jiang
Liu
a,
Fangchao
Zhao
a,
Huaping
Li
*b and
Qibing
Pei
*a
aDepartment of Materials Science and Engineering, University of California Los Angeles, 420 Westwood Plaza, Los Angeles, CA 90095, USA. E-mail: qpei@seas.ucla.edu
bAtom Nanoelectronics Inc., 440 Hindry Avenue, Unit E, Inglewood, CA 90301, USA. E-mail: huaping.li@atomnanoelectronics.com
First published on 13th November 2017
Adding solid electrolytes into organic semiconductors broadens the scope of material properties and electronic applications. Successful examples include polymer light-emitting electrochemical cells, electrolyte-gated organic transistors, and electrolyte-gated organic light-emitting transistors (EGLETs). EGLETs combine an organic light-emitting device and a transistor with a high capacitance electrolytic dielectric. Here we summarize recent progress in the development of EGLETs in both planar and vertical device architectures. The former offers a lateral geometry and in-plane light-emission for scientific scaffolds in the fundamental study of organic semiconductor opto-physics. The latter features surface emission with a unity aperture ratio, and it can be used in matrix displays without the requirement of external thin-film transistor arrays as the switching circuitry. This strategy paves an easy avenue towards fabricating highly integrated organic optoelectronic devices, and it offers a new test bed for research in iontronics and organic electronics.
Polymer light-emitting electrochemical cells (LECs) were invented in 1994 to mitigate the charge injection problem in organic optoelectronics.6 In the LEC materials system, a solid electrolyte is blended with an organic semiconductor. The presence of ions electrochemically-dopes the semiconductor and facilitates the electron injection without using air-sensitive materials. In fact, electrons could be effectively taken from a stable metal like gold or silver and injected into the lowest, unoccupied molecular orbital of an organic semiconductor,7–10 which is usually 1–2 eV higher than the metal's work function.
LECs are promising candidates for next-generation solid-state lighting and display applications, as they possess several unique properties. First, an efficient charge injection at the electrodes arises from the existence of ions at the interface, and it is insensitive to the work function of the electrodes. Thus, any conductive material can be used as an electrode, varying from stable, high work function metals (Au, Ag and Al) and organic conductors, to graphene and commonly used transparent indium-tin-oxide.11–15 Second, the electrochemical process in an LEC is largely independent of the cathode–anode distance. Therefore, variation in film thickness and roughness is less impactful in large area roll-to-roll processes.16 As a result, LECs are sufficiently robust for stretchable and flexible electronic applications.13,16–19 In addition to the applications in optoelectronic device, solid electrolytes have potential as dielectric materials in transistors due to their high capacitance20 (see Fig. 1a). The transistor drain current equation is presented as:
Organic light-emitting transistors (LETs) represent a novel class of organic optoelectronic devices, wherein the light-emission function of an OLED and the electrical-modulation function of a transistor are inherently combined.30–35 LETs offer higher charge carrier mobilities36,37 over OLEDs and in situ color tuning functionalities.38,39 Conventional LETs are usually gated by inorganic dielectrics, such as aluminum oxide or silicon oxide.31 Solid oxide dielectrics formed by atomic layer deposition or chemical vapor deposition offer high dielectric constants, but their deposition is expensive. Moreover, LETs require ultra-thin dielectrics for low voltage operation, which increases their fabrication complexity. For more information, a detailed review of LET devices has been published elsewhere.31
Combining the high capacitance value of the electrolyte in an EGT configuration and its ability to dope organic semiconductor properties in LECs, an electrolyte-gated light-emitting transistor (EGLET) was proposed and developed. This review will discuss the materials, working mechanisms, device performance, and its applications in organic optoelectronics.
EGLETs include the switching functions of a transistor and the light-emitting abilities of an OLED (or an LEC). Therefore, no additional driving transistors are needed when EGLETs are employed in a display matrix, which provides a method to realize highly-integrated optoelectronic applications. In contrast to the rigid dielectric materials used in conventional OLEDs or LETs, EGLETs use solution-processed electrolytes as the dielectric layer, which permits fully printable and flexible optoelectronics applications.
n-doping: P0 + e− + C+ → P−C+ |
p-doping: P0 + h+ + A− → P+A− |
Fig. 2 Operation principles of an EGLET. (a) The emissive p–n junction is formed in the LEP channel due to charge injection from the planar source (cathode) and drain (anode) electrodes. (b) Positive gate voltage introduces excess positive charge into the channel and effectively shifts the emission zone towards the anode. (c) Similarly, a negative gate voltage shifts the p–n junction towards the cathode. Reprinted with permission from ref. 43. Copyright 2005 Wiley-VCH. |
Since the doped LEP is a good electric conductor, it serves as an extension of the charge-injection electrodes. As the n- and p-doped polymer regions expand towards the center, the intrinsic LEP region in the middle shrinks in size. After the central intrinsic LEP is reduced to a point where the electrons and holes injected from the doped polymer are close enough to recombine, excitons are formed. The excitons will decay to generate light radiatively or heat non-radiatively.
When a positive gate voltage is applied with reference to the cathode (Fig. 2b), an electric potential is generated across the electrolyte dielectric, forcing the anions to drift towards the gate and the cations towards the EML. The cations enhance the n-doped area in the EML, and the emissive p–n junction would move towards the anode. Likewise, a negative gate voltage would drive the anions towards the EML, enlarging the p-doped region and de-doping the n-doped region. As a result, the lighting position move towards the cathode (Fig. 2c).
Since both n-doping and p-doping increase the conductivity of the EML, the channel current flowing between the drain and the source increases under the applied gate voltage. This gate-modulated current change by electrochemical doping is analogous to those in organic electrochemical transistors.48,49 An increase in the channel conductivity also improves electron and hole injection into the channel, generating even more photons.
Structure | Dielectric | Emission layer | Voltage level (V) | I on/off | L on/off | L max (cd m−2) | Ref. |
---|---|---|---|---|---|---|---|
Planar | SiO2 | SY:DCH18C6:LiTf | 40 | 3 | N/A | N/A | 53 |
Planar | SiO2 and PEO:LiTf | Tetracene | 10 | ∼10 | ∼10 | N/A | 54 |
Planar | Polyvinyl alcohol | MDMO-PPV:PEO:LiTf | 50 | ∼800 | N/A | N/A | 55 |
Planar | BCB | MDMO-PPV:PEO:LiTf | 50 | 5 | N/A | N/A | 56 |
Planar | PEO:LiTf | MEH-PPV | 5 | 100 | N/A | N/A | 43, 57 and 58 |
Planar | EMIM-TFSI | F8BT | 3 | 107 | N/A | N/A | 59 |
Vertical | PEO:LiTf:PMMA | SY | 8 | ∼100 | 103 | 8000 | 60,61 |
Vertical | PEO:LiTf:PMMA | A white emitting polymer | 8 | ∼100 | 103 | 6000 | 62 |
Vertical | PEO:LiTf:PMMA | CBP: iridium dopant | 8 | ∼100 | 105 | 104 | 63 |
The desired properties of electrolytes are (1) high ion conductivity, which provides a low voltage operation and a fast device response, and (2) high electrical resistivity, which is necessary to minimize the gate leakage current. Commonly-used electrolytes include polymer electrolytes and ionic liquids; see Fig. 3.
Polymer electrolytes consist of an inorganic salt (e.g. LiCF3SO3) dissolved in an ion coordinating polymer (poly(ethylene oxide) (PEO)), as shown in Fig. 3a. In this electrolyte, the repeating oxygen groups in the PEO chains are strong electron donors, which form a strong coordination bond with the cations. Since PEO possesses a low glass transition temperature of around −65 °C, the PEO chains are fairly flexible at room temperature, and thus allow the ions to move in the PEO matrix.50
Ionic liquids usually possess a higher ionic conductivity than polymer electrolytes, since they are in a molten state at room temperature and do not require an extra ion conductor such as PEO. However, their liquid nature limits their practical applications.51 A small amount of gelation agent is usually added to improve their mechanical integrity.52Fig. 3b displays a typical ionic liquid 1-ethyl-3-methyl imidazolium bis(trifluoromethyl sulfonyl)imide (EMIM:TFSI).
The emissive layer materials of EGLET devices are typically polymers or small molecules; see Fig. 4. To realize the functions of EGLET, (1) EMLs need to be efficient emitters, and (2) they must possess the electrochemical capacity to be reversibly doped by the electrolyte of the system.
Polymer emitters (Fig. 4a) are more attractive than evaporated small molecules in practical applications, because they can be solution-processed (potentially printable), and they are intrinsically flexible. The polymer materials reported in the EGLET publications include: poly[2-methoxy-5-(2-ethylhexyloxy)-1,4-phenylenevinylene] (MEH-PPV) and poly[2-methoxy-5-(3′,7′-dimethyloctyloxy)-1,4-phenylenevinylene] (MDMO-PPV), super-yellow (SY), poly(9,9′-dioctylfluorene-co-benzothiadiazole) (F8BT), and a white emitting polymer with an unlisted structure.
Small molecules (Fig. 4b) are the preferred emitters in current OLEDs partially because of the availability of high purity compounds by repeated sublimation.64 Moreover, the use of high vacuum physical vapor deposition systems allows the formation of multiple ultrathin layers of organic small molecules to precisely manipulate the charge carriers and excitons. However, small molecules are easily washed away when an electrolyte solution is cast onto them. The incompatibility between the small molecules and solution processed electrolyte results in the fact that they are sporadically explored in EGLETs. Examples of a small molecule EML used in EGLETs include tetracene,65 and a mixture of CBP ((4,4-bis(N-carbazolyl)-1,1-biphenyl)) host doped with iridium-containing phosphorescent guests.66
Fig. 5 (a) Structure of the first reported planar EGLET and (b) its electrical and light emission characteristics. The device consists of two layers of dielectric materials: SiO2 and a polymer electrolyte. The light emission intensity increases by 8 times with the gate voltage switching from 0 V to −10 V. The authors performed the experiment at 230 K because the elevated temperature caused the mobile ions to screen the gate potential to a certain extent. Reprinted with permission from ref. 54. Copyright 2005 Wiley-VCH. |
Later on, Liu et al. reported a top-gate EGLET.43 Unlike Hepp's work, this device uses a polymer electrolyte as the sole dielectric material, resulting in an operation voltage within 5 V. As depicted in the inset of Fig. 6a, the device possesses a bottom drain and source made of gold. A light-emitting polymer (MEH-PPV) was spin-cast as the channel. The blend of PEO and KCF3SO3 serving as the dielectric was drop-cast before poly(3,4-ethylenedioxythiophene-2,5-diyl):polystyrene sulfonate (PEDOT:PSS) was laminated on top of the electrolyte layer as the gate. The device exhibits clear transistor behavior as shown in the transfer curve in Fig. 6a. It has a current on/off ratio of around 50 when VG sweeps from 0 to −5 V (p-doping). Most interestingly, the light emission zone, which was initially close to the cathode with an inactive gate, moves towards the anode with a positive VG (n-doping) and backwards with a negative VG (p-doping). The spatial control of the emission zone was also reported in a LET structure with an oxide dielectric, working under 100 V.67,68 Here the low voltage operation (<5 V) is again attributed to the high capacitance of the polymer electrolyte and the electrochemical doping between the electrolyte and the polymer emitter.
Fig. 6 (a) Transfer characteristics of the EGLET reported by Berggren's group; the device structure is shown in the inset. (b) Light emission zone is spatially controlled while the device is alternated between n-doping and p-doping operation. Reprinted with permission from ref. 43. Copyright 2011 American Chemical Society. |
With a structure similar to that of Liu's work, Bhat et al. reported an EGLET59 using F8BT as the light-emitting polymer, and an ionic liquid EMIM:TFSI as the dielectric. A top gate PEDOT:PSS was drop-cast offset to the channel (Fig. 7a) so that the light emission from the channel is not absorbed by the gate. Unlike conventional transistors, electrolyte gated transistors still function with the gate electrode misaligned with the channel69 because the capacitance of the electrolyte is independent of the electrolyte thickness. The authors observed light emission when Vds was over 3 V. Current modulation with different VG is pronounced (Fig. 7b), but no light modulation is reported. The authors employed charge accumulation spectroscopy (CAS) to study the electrochemical doping. After applying a negative gate voltage, the authors observed significant penetration of negative TFSI ions into the light emitting polymer (p-doping), evidenced by the spectrum change in CAS shown in Fig. 7c. However, they observed no evidence of positive ion diffusion with positive gate voltages (n-doping) (Fig. 7d). This hole-current dominated operation leads to light emission occurring in the proximity of the cathode due to the lack of n-doping. As a result, no gate-modulated spatial modulation was reported.
Fig. 7 (a) The EGLET reported by Sirringhaus's group. The gate electrode is misaligned to the channel, so that the light emission is not absorbed by the gate electrode. (b) The transfer characteristics of the device with clear hysteresis. Charge accumulation spectroscopy (CAS) under (c) negative gate voltages and (d) positive voltages. Upon application of a negative gate voltage, significant spectrum changes in CAS are shown in (c), indicating the penetration of negative TFSI ions into the light emitting polymer (p-doping). A positive gate voltage does not induce a similar spectrum change, suggesting that the positive ions do not penetrate the LEP under this voltage. Reprinted with permission from ref. 59. Copyright 2012 American Chemical Society. |
Two major light-emitting polymers employed in the previously mentioned EGLET devices, PPV and F8BT, have been intensively studied under the electrolyte gating configuration by different groups of researchers.22,70–72 Although no light-emitting properties were studied, they thoroughly investigated the electrical property modulation of these LEPs in ionic environments. Frisbie's group used the ionic liquid 1-ethyl-3-methylimidazolium tris(pentafluoroethyl)trifluorophosphate (EMI-FAP) to gate both F8BT and MEH-PPV, with a top Pt gate configuration as shown in Fig. 8a. The conductivities of both the polymers were modulated by electrochemical p-doping as a function of the gate potential in a transistor geometry. In these experiments, very high carrier densities on the order of 1021 cm−3 were obtained in the polymer semiconductors with ionic liquid gating. Interestingly, the polymers displayed a non-monotonic conductivity dependence on the different gate potentials, as shown in Fig. 8b and c. The conductivity initially increases as the potential increases, followed by a peak and an eventual decrease as the doping potential increases. The authors claim that this reversible decrease results from a hole mobility dependent on the carrier density. In a similar experiment, Sayago and coworkers used three ionic liquids containing TFSI anions to gate MEH-PPV.72 Under a drain–source voltage of 1 V, a 103 on–off drain–source current ratio was observed when the gate was swept from 0 V to 1 V because of p-doping electrochemical reactions between MEH-PPV and TFSI. Both Frisbie and Sayago's experiments focus on p-doping of the polymers, but they did not report n-doping of the transistors.
Fig. 8 (a) Experimental setup for hole transport property study of F8BT and MEH-PPV under ionic liquid gating. Measured hole current for (b) F8BT and (c) MEH-PPV as a function of gate voltages. Hysteresis is clearly observed in both devices, which results from the slow movement of the electrolytes. Reprinted with permission from ref. 89. Copyright 2011 American Chemical Society. |
For comparison, it is imperative to mention previous work on transistors with a polymer–electrolyte mixture as the semiconductor and an inorganic oxide as the dielectric. This structure contains ions in the charge transport layer but not in the dielectric layer, and it is equivalent to placing an LEC above an oxide capacitor. This set of devices uses the ion-assisted charge injection in the semiconductor/electrode (drain and source) interface, but using an oxide dielectric with a small capacitance limits the lighting modulation ability and requires a high gate voltage. Back in 2004, Edman et al. reported a bottom gate transistor made on a silicon oxide dielectric.53 The channel material was a mixture of a SY LEP, a crown ether ion conductor dicyclohexano-18-crown-6 (DCH18C6), and LiCF3SO3. The authors observed a hole current modulation with gate voltage sweeping, but no light modulation was reported (Fig. 9a). Similar to Edman's work, Yumusak et al. described a bottom gate field effect transistor with an organic material, poly-vinyl alcohol, as the dielectric55 (Fig. 9b). The channel layer was a mixture of a PPV derivative, an ion conducting PEO, and LiCF3SO3 salt. The drain and source were made using gold, and the gate was made from ITO. Clear hole current modulation was observed, but no light emission change vs. gate voltage was reported. Later, the same group demonstrated a similar device, with the dielectric layer replaced by benzocyclobutene (BCB) with a higher dielectric constant than that of poly(vinyl alcohol).56 They noted a similar current modulation, and finally, an increase in light emission when the gate voltage exceeded 30 V. The light emission was at the proximity of the electrode and no spatial modulation was seen. This suggests that the electrostatic effects (field effect) were not strong enough to control the electrochemical reaction, which is essential for ion-containing light-emitting systems. Both Edman and Yumusak's works required high voltages (>50 V) to fully switch the transistor, because of the oxide dielectric used in the structure.
Fig. 9 (a) An oxide gated LEC reported by Edman et al. (b) An electrochemical light emitting transistor reported by Yumusak et al. Both devices contain electrolytes admixed in the emitter layer, but not in the gate insulator layer. Since solid dielectrics have low dielectric constants, their gate voltage (ranging from 0–50 V) is much larger than gate voltages in electrolyte gated devices (0–5 V). Reprinted with permission from ref. 55 and 56. Copyright 2010 AIP Publishing LLC. |
It is also worth mentioning two examples of planar EGLET devices that employed unconventional gate structures. To further improve the spatial control of the EGLET devices, Liu and his coworkers developed double gate EGLETs,57 with one gate (p-gate) controlling the p-doping and another (n-gate) the n-doping. Using appropriate voltages and timing control, one can precisely define the location of the emissive junction, which was observed in the undoped area between the two gates. Later on, they discovered that omitting the p-gate saw the light emission zone located in proximity to the n-gate.58 This is because p-doping is by nature more active than n-doping. In both devices, they demonstrated a homogeneously centered emission zone.
Vertical EGLETs employing electrolyte as the dielectric layer were not demonstrated until very recently. In 2016, Luan and coworkers developed the first vertical EGLET with a solution processed polymer emitter SY.60 As displayed in Fig. 10a, the device consisted of a OLED stack on the bottom (ITO/PEDOT/SY/Al), above which a polymer electrolyte was employed as the dielectric layer, and an another Al layer as the gate electrode. The polymer electrolyte was a mixture of PEO, LiCF3SO3 and poly(methyl methacrylate) (PMMA) in a weight ratio of 5:1:10. Similar to the planar device, the electrochemical synergy between the SY emitter and the polymer electrolyte at the interface governs the vertical EGLET's functionality. Under a constant Vds of 7 V, this device can be switched on with a positive gate voltage of 8 V, and turned off with a negative gate voltage of −8 V, as shown in Fig. 10b. The maximum luminance can reach as high as 4500 cd m−2 with an external quantum efficiency of 1.7%, demonstrating its potential in signage displays. The Al porous electrode can be replaced with an Al and LiF composite material for better electrochemical intereaction.61 It is also possible to construct a vertical EGLET with a white emitting polymer based on the same structure.62 Besides similar electrical characteristics, they demonstrated a spectrum shift with gate voltage as well as a color coordinate shift attributed to changing light-emitting mechanisms at different gating voltages. The authors claimed that at low Vg, light emission is governed by electromer generation (electromer is a pair of chemically identical molecules charged by electrons and holes, which are statistically independent of each other81) with an imbalanced charge carrier injection. In contrast, electromer emission is suppressed at high Vg and results in a spectral shift. Furthermore, red, green, and blue emission were achieved with different color filters.
Fig. 10 Device structure and electrical characteristics of the vertical EGLETs employing (a) the SY polymer emitter and (b) the small molecule emitter with CBP as the host and Ir(ppy)3 as the guest. In both devices, the electrolyte makes contact with the organic layer through the porous Al source. Reprinted with the permission from ref. 60 (Copyright 2016 Wiley-VCH) and ref. 63 (Copyright 2017 American Chemical Society). |
The same group of scientists also employed a small molecule OLED for the vertical EGLETs63 (Fig. 10b). The OLED stack possessed the following structure: ITO/PEDOT/TAPC/CBP:Ir(ppy)3/CBP/Al, where PEDOT serves as the hole injection layer, TAPC (4,4′-cyclohexylidenebis[N,N-bis(4-methylphenyl)benzenamine]) as the hole transport layer, CBP as the fluorescent host, Ir(mppy)3 as the green emission phosphorescent guest, and an another CBP layer as the electron transport layer. Above the OLED stack sat the same electrolytic capacitor layer as reported in the previous publication. The modulation of light emission is governed by the electrochemical doping of the CBP in the electron injection layer. These phosphorescent OLEDs exhibit tunable luminance between 0.1 and 10000 cd m−2, with the gate voltage shifting from −4 V to 7 V. The phosphorescent materials used in this configuration harvest the triplet excitons, which leads to an increase in efficiency from 6–7 cd A−1 in devices with polymer emitters to 25 cd A−1 in this device.
Fig. 11 Prototype of EGLET devices. (a) A seven-segment display with gate patterned as seven different segments and SY as the emitter. Numbers 0–9 are displayed. (b) A 5 × 5 active matrix display with the pixelized gate electrode and SY emitter. Letters A, T, O and M are displayed. (c) A 3-color module with the patterned emissive layer (small molecules) and patterned gate electrode. The display shows a red “r”, a green “g” and a blue “b”. Reprinted with the permission from ref. 61 (Copyright © 2017 American Chemical Society), ref. 60 (Copyright © 2016 Wiley-VCH) and ref. 59 (Copyright © 2017 American Chemical Society). |
To improve the emission efficiency and maximum luminance, the light emitting molecules used in state-of-the-art OLED devices should be employed in EGLETs. The reported EGLET devices mainly use a fluorescent polymer as the emitter, which fails to collect 75% of the emissive excitons (triplets).64 Polymers blended with triplet harvesters, such as phosphorescent small molecules82 or thermally activated delay fluorescent emitters,83 could address this efficiency problem. Additionally, inorganic emitters such as tungsten diselenide, molybdenum disulfide, zinc oxide,84 and organometal halide perovskites85 also electrochemically inter-react with electrolytes and are worth exploring.
As discussed previously, polymer emitters in direct contact with the electrolyte cause exciton quenching, which decreases device performance and stability. Another charge transport layer solely responsible for electrochemical doping but not light emission would separate the polymer emitter from the electrolyte, which might mitigate this issue.
Also, as noted, vertical EGLETs can serve as a device combining an emitting element and a driving transistor. However, practically speaking, each pixel will need at least one or more switching transistors to ensure full addressing and control. Specifically-designed circuits and driving schemes are required to employ LETs in active matrix displays. Realization of sufficient grey levels with current technology needs further study, especially on gate voltage duration and amplitude. In addition, with a proper choice of electrolyte and semiconductor, electrolyte gated transistors have demonstrated oscillator circuits faster than 5 μs.86 After adopting this technology, the EGLETs could satisfy the technical requirements for application in a 50 Hz display.
Finally, electrolyte gated light emission offers a test-bed for investigation into the fundamental physics of organic semiconductors and optoelectronics, especially under high charge carrier density induced by electrochemical doping. This will advance the development of iontronics87,88 and organic electronics.
This journal is © the Partner Organisations 2018 |