2
Interfacial Modification Methods

Over the past two decades, extensive efforts [6, 8, 9, 16] have been devoted to the modification of interfaces in organic field-effect transistors (OFETs), including physical modification methods, such as encapsulation or insertion on the surface, and chemical modification methods, such as covalently bonded self-assembled monolayer (SAM) modification.

2.1 Noncovalent Modification Methods

2.1.1 Charge Insertion Layer at the Electrode Surface

Because the injection barrier originates generally from the gap between the Fermi level of metal electrodes and the highest occupied molecular orbital (HOMO)/lowest unoccupied molecular orbital (LUMO) energy levels of organic semiconductors (OSCs) according to the energetic alignment, a reasonable solution is to insert a suitable additional layer to modify the electrode and function as a springboard for charge carriers, thus possibly overcoming the poor injection [70]. Initially, this method of introducing a thin buffer layer of several nanometers at metal OSC/electrode interfaces, referred to as a charge insertion layer (CIL) [71, 72], had already been applied to increase the efficiency of organic light-emitting diodes by improving the contact.

A preliminary study [73] was carried out by Hajlaoui et al., who inserted a thin film of tetracyanoquinodimethane (TCNQ) (1), working as molecular dopant, between Au and OSC active layers in OFETs to achieve an increased hole mobility. By using an ultrathin LiF interlayer as CIL between Al electrodes and F16CuPc active layers [33], enhanced n-channel performance was achieved. Since then, CIL modification has attracted more and more attention in terms of enhancing the hole/electron injection efficiency. Various materials, such as inorganic compounds (metal oxides and inorganic salts) and organic compounds (small molecules and polymers), have been used as CILs, as shown in Figure 2.1.

Figure 2.1 Chemical structures of the charge insertion layer (CIL). (a) Metal oxides, (b) metal salts, (c) small molecules, (d) polymers, and (e) other materials.

Metal oxides [74] (Figure 2.1a) provide a variety of candidates for CILs because of their wide range of work functions. Transition metal oxides in particular [75, 76], whose work functions range from very low (? ~ 3 eV for ZrO2) to extremely high (? ~ 7 eV for V2O5), are suitable for modifying different metals with matched energy levels. After introducing the CIL, the thickness of the depletion layer between the metal and metal oxide decreases to a few nanometers, producing a reachable distance for tunneling injection, and the contact resistance could be significantly reduced once the tunneling injection is dominant. The CIL layer can also act as a protection layer against the diffusion [77] of metals from electrodes or the permeation [78] of O2 or H2O into the active layer, thus improving the device stability.

Metal salts are another big family of inorganic materials working as CILs (Figure 2.1b) [79, 80]. They experience a similar mechanism as metal oxides. One of the benefits of using metal salts as CILs is their compatibility with the solution-process method. For example, CuSCN [81, 82] can be spin-casted on Au electrodes as a novel CIL with an intrinsic electron-blocking property. However, there are some drawbacks of metal salts when applying them as CILs. First, the thickness of the CIL affects electron transport. A thicker layer of CsF [83] (more than 0.6 nm) decreased the mobility owing to its insulating nature. Second, when depositing a CIL into ambipolar OFETs by using a single component, the outcome performance of p- and n-channels cannot be enhanced simultaneously. This problem could be solved by mixing two different materials to form a hybrid CIL [84]. The injection barrier for different types of carriers was tuned finely according to the molecular ratio of the two compounds.

When it comes to organic small molecules as CILs (Figure 2.1c), i.e. charge-transfer (CT) complex (2) [85], metal complex (3-5) [86, 87], metal tetraphenylporphyrin and phthalocyanine (6-10), [88] and N-heterocyclic compounds (11-15) [89], direct CT occurs at the interface between the organic CIL and the semiconductor, which helps to reduce the contact resistance by the additional interface dipole. However, a problem that still exists is the possibility of matrix-dopant hybridization [90, 91]. The organic polymers with dipolar groups [23, 92, 93] (16 and 17) (Figure 2.1d) can also act as electrode surface modifiers to achieve electrodes with a low work function. Polyelectrolytes (18-22), which comprise the groups capable of being charged with cations or anions, were proven recently to be a promising CIL to achieve ohmic contacts with a wide range of work functions (3.0-5.8 eV) [23].

Other materials (Figure 2.1e), such as DNA [94] and graphene oxides (GOs) [95-97], have also been used as CILs. The presence of phosphate groups in the DNA leads to a dipole layer, which could change the orientation in response to applied gate voltages and enhance the charge injection of both types of carriers. When GOs were applied as CILs in OFETs, the improvement in device performance could be attributed to the good contact nature of carbon-based materials with semiconductors, which will be discussed in Section 3.3.2.

2.1.2 Dielectric Surface Passivation Methods

Polymer encapsulation is a main choice for modifying the inorganic dielectric surface. Most of the polymer modifier, for example poly(imide-siloxane) [98], poly(4-vinylpyridine) (PVPyr) [99], and polystyrene (PS) [100], are used to change the surface energy of the dielectric and thus influence the grain size of OSCs growing whereon. There is still a debate on how the surface energy of polymer modifiers influence performance of OSC thin films; but as a general rule, the closer the surface energy of the dielectric layer is to that of the OSC, the higher the carrier mobility [101]. More fundamentally, it is a competition between the aggregation force within the OSC and the adsorption of OSC on the substrate.

The polymer modifiers, however, suffer from high roughness and trap density, which hamper charge transport on the surface. The same problem exists when directly using insulating polymers as gate dielectrics [102-105]. These applications call for the efficient surface modification methods of polymer films. However, the lack of functional groups makes it very hard to modify polymer dielectrics. Several strategies have been developed recently to solve this problem:

1. (i) By directly depositing a bilayer dielectric consisted of a bulk dielectric layer and an ultrathin interface layer (Figure 2.2a,b);
2. (ii) By using blends of two components to induce a phase-separated bilayer structure (Figure 2.2c,d);
3. (iii) By utilizing a monolayer capable of self-assembly on the polymer surface (Figure 2.2e-g).

Figure 2.2 Modification of polymer dielectrics. (a) Schematic representation showing the bilayer dielectric made from an initiated chemical vapor deposition process. (b) Molecular structures of the interfacial layer and the bulk dielectric layer. (c) A schematic structure of the vertical phase separation of the blended dielectric. (d) Schematic representation showing the effect of the polaronic disorder and carrier concentration in OFETs with three dielectrics. (e) A few-layer triptycene film featuring 2D nested hexagonal packing and one-dimensional (1D) layer stacking on a parylene dielectric. (f) Molecular structure of the triptycene molecule. (g) Atomic force microscopy (AFM) height images of evaporated OSC films on parylene (top) and triptycene-coated parylene (bottom) gate dielectrics.

Source: (a, b) Adapted from Pak et al. [34]. (c, d) Adapted from Khim et al. [106]. (e-g) Reproduced from Yokota et al. [107], © 2018 Springer Nature.

For the first strategy, a bilayer polymer dielectric (Figure 2.2a) was employed [34], which consisted of a bulk layer of electropositive copolymer (26) for inducing electrostatic polarization along the channel and an ultrathin interface layer of 23 on top to form a good interface with OSCs. The polymer dielectric was prepared via a one-step, low-temperature, solvent-free chemical vapor deposition process with different ratios of two monomers 24 and 25 (Figure 2.2b) [35].

The second strategy can be realized by blending a fluorinated high-k polymer (27) with poly(methyl methacrylate) (PMMA) (28) as a...