中文   |   English   |   Email Alert
Fig/Tab
Advanced Search
image

Solution Fabrication and Morphology Control of Organic Semiconductor Single-Crystal Thin Films

Shifeng Ke , Zheyuan Wang , Boyu Peng , Hanying Li

Progress in Chemistry ›› 2026, Vol. 38 ›› Issue (5) : 849 -869.

PDF
Progress in Chemistry ›› 2026, Vol. 38 ›› Issue (5) : 849-869 DOI: 10.7536/PC20250918
research-article
Scan to read & Share
Solution Fabrication and Morphology Control of Organic Semiconductor Single-Crystal Thin Films
Author information +
History +
PDF

Abstract

Organic semiconductor single-crystals are promising materials for application in organic electronic devices， such as organic field-effect transistors， due to their superior charge carrier transport properties. Solution-based methods， with the advantages of low cost and scalability for mass production， have emerged as important approaches for the fabrication of organic semiconductor single-crystal thin films. A thorough understanding of the solution crystallization mechanisms of organic semiconductors is crucial for achieving precise control over thin film morphology. However， the understanding of solution crystallization mechanisms remains elusive， and thin film fabrication often relies on a trial-and-error approach. This review summarizes recent progress in the solution fabrication of organic semiconductor single-crystal thin films and the control of thin film morphology. We introduce commonly used solution processing techniques， including spin coating， drop casting， meniscus-guided coating， and liquid-surface-mediated growth， and elaborate strategies for morphology control， such as substrate surface modification， process parameter optimization， and solvent engineering. Additionally， the application of molecular dynamics simulations in exploring film deposition and microscopic crystallization mechanisms is discussed. Through a deeper understanding and precise manipulation of the crystallization mechanisms， the large-area fabrication and application of organic semiconductor single-crystal thin films can be realized in the near future， ultimately promoting the commercialization of organic semiconductors in areas such as flexible electronics.

Contents

1 Introduction

2 Structure and performance of organic semiconductor single crystals

3 Solution fabrication of organic semiconductor single-crystal thin films

3.1 Spin coating

3.2 Drop casting

3.3 Meniscus-guided coating

3.4 Liquid-surface-mediated growth

4 Control of solution crystallization and optimization of film morphology

4.1 Substrate surface modification

4.2 Process parameter optimization

4.3 Solvent properties and solute-solvent interaetions Ideal single-crystal film morphology and general regulation strategies

5 Molecular dynamics simulation of thin film deposition and crystallization

6 Conclusion and outlook

Graphical abstract

Keywords

organic semiconductor / single crystal / solution fabrication / thin-film morphology

Cite this article

Download citation ▾
Shifeng Ke, Zheyuan Wang, Boyu Peng, Hanying Li. Solution Fabrication and Morphology Control of Organic Semiconductor Single-Crystal Thin Films. Progress in Chemistry, 2026, 38(5): 849-869 DOI:10.7536/PC20250918

登录浏览全文

4963

注册一个新账户 忘记密码

1 Introduction

Compared with traditional inorganic semiconductors (such as silicon and gallium arsenide), organic semiconductors offer advantages including low-temperature processing, low-cost manufacturing, and flexibility. Meanwhile, the diversity of organic molecular structures enables the combination of multiple functions and characteristics through molecular design, thereby opening up broad avenues for the development of organic electronic devices. Currently, a large number of organic semiconductor materials have been discovered or synthesized and are widely used in the fabrication of organic electronic devices, including organic field-effect transistors (OFETs), organic light-emitting diodes (OLEDs), and organic photovoltaics (OPVs).[1-3]The disorder in traditional amorphous or polycrystalline organic films not only disrupts strong intermolecular π-π stacking interactions but also complicates carrier transport and exciton diffusion processes. In contrast, organic single crystals possess advantages such as long-range ordered structures, high purity, absence of grain boundaries, and fewer defects and traps, which can effectively improve carrier transport efficiency and better reflect the intrinsic properties of organic semiconductors. Therefore, organic single crystals have become important tools for enhancing the performance of organic devices, exploring the relationship between the structure and properties of organic semiconductors, and studying fundamental organic physics, attracting widespread attention from the academic community.[4-7].
However, preparing large-area, low-grain-boundary thin films with uniform thickness and orientation is a prerequisite for fully leveraging the long-range ordered characteristics of organic single crystals and ultimately achieving large-scale industrial applications. As a low-cost and easily scalable method for thin film preparation, solution processing has received widespread attention over the past two decades[8-9]. Droplet-pinned crystallization (DPC)[10], meniscus-guided coating (MGC)[11], and liquid-interface-mediated growth[12-17]and other such technologies have fully demonstrated the feasibility of preparing high-quality organic semiconductor single-crystal thin films via solution methods. It should be noted that the term "single-crystal thin films"[18-20]here does not refer to perfect single crystals completely free of grain boundaries, but rather to thin films composed of large-sized (typically greater than several hundred micrometers), highly oriented crystalline domains. Compared to polycrystalline thin films composed of randomly oriented microcrystals, devices fabricated from such films ensure that charge carriers transport within the same crystalline domain, almost unaffected by grain boundaries, thereby exhibiting the charge transport characteristics of single crystals. Some literature also refers to them as "single-crystalline thin films"[21-22], "single crystal arrays"[23-24], or "single-crystalline arrays"[25], among others. The crystallization process of thin films in solution methods is subject to the complex influence of many factors, including substrate properties[26], process parameters[27-29], and solvent characteristics[30-33]. These factors are coupled with each other, making precise control of the crystallization process extremely challenging; consequently, practice often relies on continuous trial and error to optimize film morphology. Therefore, deeply understanding the solution crystallization mechanism of organic semiconductors to provide theoretical guidance for regulating crystallization behavior and optimizing film morphology has always been a focus of research in this field[4,34]. This article aims to systematically review recent research progress in the solution-based preparation and morphology control of organic semiconductor single-crystal thin films, introduce common solution preparation methods and strategies for regulating film morphology, discuss the application of molecular dynamics simulations in exploring film deposition and microscopic crystallization mechanisms, and provide an outlook on the large-area preparation and application of organic semiconductor single-crystal thin films.

2 Structure and Properties of Organic Semiconductor Single Crystals

As early as 1948, the intrinsic semiconductor properties of organic single crystals were discovered in metal-free phthalocyanine single crystals[35]. Until 1996, Horowitz et al.[36]reported the first organic single crystal-based field-effect transistor. α-Sexithiophene single crystals prepared by physical vapor transport (PVT) as the semiconductor active material exhibited a mobility of 0.075 cm2/(V·s). Since then, the mobility of organic semiconductor single crystals has been significantly improved in the early 21st century. For example, Podzorov et al.[37]achieved a mobility as high as 8 cm for p-type field-effect transistors based on Rubrene2/(V·s). Takeya et al.[38]further increased the mobility to 18 cm by improving the interface between the crystal and the insulating layer2/(V·s), and suggested that the intrinsic hole mobility of Rubrene could reach 40 cm2/(V·s). In the same year, Jurchescu et al.[39]also achieved a maximum mobility of 40 cm in Pentacene single crystal field-effect transistors2/(V·s). This value far exceeds the initial polythiophene thin films[40]of 10-5 cm2/(V·s), and surpasses amorphous silicon (mobility approximately 1 cm2/(V·s)), meeting the basic requirements for industrial applications. These results demonstrate the great application potential of organic single crystals as semiconductor active materials.
The high intrinsic mobility of organic single crystals has attracted increased attention and research over the past two decades. In-depth studies on the structure-property relationships of organic single crystals provide theoretical guidance for designing high-performance organic semiconductor molecules. Watanabe et al.[41]found that, under identical fabrication and testing conditions, hexacene single crystals exhibit higher mobility than pentacene. Theoretical calculations indicate that extending the π-conjugated system in the hexacene structure effectively enhances intermolecular interactions without altering the molecular packing arrangement, thereby increasing transfer integrals and reducing reorganization energy, ultimately leading to improved mobility. The consistency between theoretical and experimental results demonstrates that extending the π-conjugated backbone is an effective approach to enhancing carrier mobility. Besides extending the π-conjugated backbone, introducing substituents is another common method for tuning the electrical properties of organic single crystals. Rubrene is a typical example. The introduction of four ortho-substituted phenyl groups transforms the molecular packing of rubrene single crystals from the original herringbone arrangement of tetracene into a coplanar herringbone arrangement with enhanced π-π stacking, achieving effective electronic coupling and larger transfer integrals[42]. Furthermore, several molecular design strategies have been employed to improve other properties of organic single crystals. For instance, introducing heterocycles (such as thiophene) can enhance the air stability of π-conjugated molecules[43-44]; introducing alkyl side chains can improve their solubility[45-46]. Based on the aforementioned molecular design strategies, numerous acene and heteroacene derivative single crystals have been synthesized[47]. Many of these organic single crystals exhibit hole mobilities greater than 1 cm2/(V·s) while maintaining good solubility and air stability, for example: 6,13-bis(triisopropylsilylethynyl)pentacene (TIPS-PEN)[45], 5,11-bis(triethylsilylethynyl)anthradithiophene (TES-ADT)[48], difluoro tri-sec-butylsilylethynyl pentacenedithiophene (F-TSBS-PDT)[49], dithieno[2,3-d;2′,3′-d′]benzo[1,2-b;4,5-b′]dithiophene (DTBDT)[50], among others. Some even approach polycrystalline silicon levels (> 10 cm2/(V·s)), for example: 2,7-dioctyl[1]benzothieno[3,2-b][1]benzothiophene (C8-BTBT)[19], 2,9-didecyldinaphtho[2,3-b:2′,3′-f]thieno[3,2-b]thiophene (C10-DNTT)[51], 3,11-didecyldinaphtho[2,3-d:2′,3′-d′]benzo[1,2-b:4,5-b′]dithiophene (C10-DNBDT)[22], among others. The emergence of these high-mobility, easily processable organic single crystals lays the foundation for large-area, low-cost fabrication of high-performance organic electronic devices.

3 Solution-processed organic semiconductor single-crystal thin films

The growth of organic semiconductor single crystals is a prerequisite for studying their structure and properties, as well as for fabricating organic devices.[52]Obtaining single crystals with few impurities, few defects, and high quality is essential to reflect the intrinsic properties of organic semiconductors as accurately as possible, and is also key to improving device performance. Currently, preparation methods for organic semiconductor single crystals are mainly divided into two categories: vapor-phase methods and solution-phase methods. Vapor-phase methods (such as the PVT method) can produce bulk single crystals with structures close to ideal, high quality, and high purity; many single crystals used for theoretical research on structure and properties are prepared using this method. Furthermore, the highest carrier mobilities in organic semiconductors have mostly been measured in single crystals prepared by vapor-phase methods.[37-39]However, the actual large-scale production and application of organic devices often require the semiconductor layer to exist in the form of a thin film.[53]Random nucleation and island growth on substrates lead to non-uniform thickness and orientation of single crystals prepared by vapor-phase methods, making it difficult to form continuous and uniform organic single-crystal thin films.[54-55]Due to the lack of suitable growth substrates, traditional epitaxial growth techniques for inorganic single-crystal thin films (which utilize lattice-matched substrates) are difficult to apply to organic single crystals.[56]Therefore, vapor-phase methods are difficult to use for preparing large-area organic semiconductor single-crystal thin films and are only suitable for fundamental theoretical research on individual devices.[57]In addition, vapor-phase methods typically require high vacuum and high-temperature conditions, which not only consume significant time and cost but also impose higher requirements on the thermal stability of organic molecules.
Compared to the vapor phase method, the solution method offers the advantage of being conducted under ambient pressure and low-temperature conditions. It is a simple and low-cost preparation technique, conducive to large-scale production and commercial applications. Furthermore, the solution method has been widely applied in the fabrication of organic semiconductor thin films.[8-9]. With continuous improvements to traditional solution methods and the proposal of new techniques, the crystal quality (mainly reflected in grain size and orientation) and device performance of solution-processed organic semiconductor thin films have significantly improved. This section will focus on introducing common solution processing techniques and improvement methods for fabricating organic semiconductor single-crystal thin films.

3.1 Spin coating method

Spin coating is one of the most common solution methods for preparing organic semiconductor thin films. In this method, a solution is dropped onto a substrate, which is then rotated at high speed to rapidly evaporate the solvent and form a thin film; the film thickness is influenced by the rotation speed, solution concentration, and viscosity.[58]. Spin coating can efficiently and conveniently prepare organic semiconductor thin films with uniform thickness, and it has been widely used in both laboratory research and the actual production of organic electronic devices.[59]. However, due to the difficulty in controlling crystal nucleation and growth during the rapid spinning process, spin coating typically yields amorphous or polycrystalline thin films.[18,46,60]. The low crystallinity, small crystallite size, and abundant grain boundaries in the thin films are detrimental to carrier transport.
To improve the crystal quality of thin films and device performance, researchers have made many improvements to the spin-coating method. A common strategy is to pre-mix polymer dielectric materials into the organic semiconductor solution, utilizing the self-assembly behavior during the spin-coating process to achieve vertical phase separation between the organic semiconductor and the polymer.[60]The polymer dielectric layer can improve the wettability during the film deposition process and reduce interface defects. In addition, selecting specific polymer materials can also serve as a template to control the crystallization of organic semiconductors. Kwon et al.[18]incorporated poly(4-styrenesulfonate) (PSS) into a C8-BTBT solution, utilizing the electrostatic interaction between the periodic sulfonic acid side groups of PSS and C8-BTBT to induce nucleation and growth, ultimately obtaining C8-BTBT films with single-crystal domain sizes at the millimeter level. Since the single-crystal size is sufficient to cover organic field-effect transistor devices, the film exhibits an intrinsic mobility of 18 cm2/(V·s). To improve the uniformity of crystal domain orientation in C8-BTBT films, Zhenan Bao et al.[61]employed eccentric spin-coating to prepare blend solution films of C8-BTBT and polystyrene. These films possess a highly oriented metastable crystal structure, with the corresponding transistors achieving an average hole mobility of 25 cm2/(V·s). Unlike central spin-coating (where the substrate is located at the center of rotation and the solution diffuses radially outward), eccentric spin-coating places the substrate 20–40 mm away from the center of rotation, utilizing centrifugal force to induce film orientation.

3.2 Drip method

Drop casting is a very simple method that involves dripping an organic semiconductor solution onto a target substrate, allowing crystals to precipitate as the solvent evaporates. In this relatively static equilibrium process, organic semiconductor molecules form ordered crystal structures through self-assembly driven by intermolecular forces. The key to improving crystal quality lies in maintaining the stability of the solvent evaporation process; feasible approaches include using a sealed chamber[62-63], patterned substrates[64], saturated solvent atmosphere[65], inert gas flow[66], etc. However, since droplets tend to move on the substrate during shrinkage, causing unstable and uneven deposition, drop casting typically yields discrete small crystallites or polycrystalline films with chaotic orientation. Fabricating transistor devices based on such morphologies requires manual alignment of the electrodes with the position and direction of the crystals to ensure effective contact, making it difficult to scale up for actual production.
To address this issue, Li Hanying et al.[10,67]proposed the droplet-pinned crystallization (DPC) method. Asshown in Figure 1a, the basic principle of DPC is to use a silicon wafer (or alternatively a needle[68], capillary tube[69], or other fixed objects) to pin the droplet, achieving stable contraction of the three-phase contact line (substrate-solution-air) at the droplet edge. As the solvent evaporates, nuclei formed near the contact line can grow stably and continuously in the direction of the contact line contraction (toward the droplet center), yielding highly oriented stripe-like single-crystal arrays. The length of these stripe single crystals can reach hundreds of micrometers (Figure 1d, e) or even millimeter levels[68]. Since charge carriers can transport along the long axis of the stripes within a single crystal domain, field-effect transistors based on stripe-like TIPS-PEN and C60single-crystal arrays exhibited average hole mobilities of 2.4 cm2/(V·s) and electron mobilities of 3.0 cm2/(V·s), with maximum values reaching 3.8 and 11 cm2/(V·s), respectively[10,70]. Another significant advantage of the DPC method is that the growth area can be easily expanded by changing the shape, size, and position of the pinning objects. Asshown in Figure 1b, oriented crystal growth was achieved on a 4-inch wafer using multiple elongated pinning objects[10]. Additionally, the DPC method can even grow two types of organic semiconductor single crystals simultaneously at different interfaces, enabling the one-step fabrication of organic single-crystal heterojunctions from mixed solutions (Figure 1c)[71]. Therefore, the DPC method is not only a simple and efficient approach for growing organic semiconductor single crystals but also provides a stable and controllable platform for studying solution crystallization processes and fabricating more complex organic device structures.
Takeya et al.[72]also adopted a similar strategy, using a silicon wafer on an inclined substrate to fix the droplet, causing the contact line to contract toward the edge of the wafer, thereby achieving oriented growth of organic single crystals.

3.3 Meniscus-guided coating method

Meniscus-guided coating (MGC) is a general term for various solution coating methods, holding great potential for the large-area, cost-effective fabrication of high-performance organic semiconductor films and organic electronic devices.[11]These methods utilize the translation of the substrate or coating tool to induce oriented crystal growth during film deposition. The meniscus refers to the curved gas-liquid interface formed between the coating tool and the substrate. As the solvent evaporates, the solution reaches supersaturation, leading to crystallization at the three-phase contact line along the meniscus edge, thereby forming a film. During the MGC process, the coating tool pulls the meniscus to move directionally across the substrate surface, guiding the oriented growth of crystals and the alignment of molecules within the film. Additionally, MGC is compatible with roll-to-roll technology in industrial production, promising large-area continuous fabrication of organic semiconductor films. Therefore, MGC is currently one of the mainstream methods for preparing organic semiconductor single-crystal films and has been applied to various organic semiconductor systems. Depending on the coating tool used, MGC can be primarily categorized into dip-coating, blade-coating, slot-die coating, and zone-casting (Figure 2).
Dip coating equipment is simple and does not require specific coating tools. After immersing the substrate in the solution, the solution climbs onto the substrate under the action of surface tension via capillary forces, forming a meniscus. When the substrate is pulled upward, the meniscus is stretched due to the pinning effect at the wetting interface, resulting in films with significant anisotropy. By selecting appropriate solvent types, solution concentrations, and pulling speeds, uniform organic semiconductor films can be obtained.[29]. Rogowski et al.[73]utilized mixed azeotropic solvents and suitable pulling speeds to prepare oriented ribbon-like TIPS-PEN single crystal arrays, achieving a mobility of 1 cm2/(V·s). Chai et al.[74]also prepared self-oriented ribbon-like C8-BTBT films by controlling the pulling speed and solution concentration. They pointed out that when the substrate is withdrawn from the solution, the continuous line of C8-BTBT molecules at the contact line undergoes segregation due to "fingering instability"[75], forming periodically distributed domains. These domains act as nucleation sites, guiding C8-BTBT crystallization to form ribbons parallel to the pulling direction. Pulling speeds that are too fast or too slow lead to uneven nucleation at the contact line, resulting in non-uniform film thickness and morphology. Solution concentration affects the supply of solution at the contact line; only sufficiently high concentrations can form continuous ribbon crystals. However, dip coating requires immersing the substrate in a solution bath, which not only consumes a large amount of solution but also has limited operational space, presenting certain limitations.
Compared to the dip-coating method, the blade coating method (Blading) requires only injecting a small amount of organic semiconductor solution into the gap between the coating tool and the substrate to begin coating, significantly reducing the consumption of the organic semiconductor solution. Blade coating methods include blade coating, solution shearing, and bar coating; their coating processes are similar, with the main difference lying in the morphology of the coating tools. The first two methods use blade-shaped coating tools (typically silicon wafers). In blade coating, the blade is fixed vertically hundreds of micrometers above the substrate, whereas in solution shearing, the blade is placed at an incline. Bar coating uses cylindrical coating tools, such as metal rods.
To prevent adsorption of organic semiconductors, the surface of coating tools usually needs to be modified with hydrophobic self-assembled monolayers (SAM)[76]. In addition, more complex microstructure modifications can also be used to improve mass transfer. Zhenan Bao et al.[77]controlled the shape and spacing of micropillars by modifying patterned micropillars on the blade surface (Fluid-enhanced crystal engineering, FLUENCE) to enhance lateral mass transfer (perpendicular to the blade coating direction), thereby reducing the formation of voids and dendritic growth during thin film deposition. Asshown in Figure 3e, TIPS-PEN films without patterned micropillars exhibit a very high grain boundary density, with dendritic domains only at the micrometer scale; whereas after using patterned micropillars, grain boundaries are significantly reduced, domain sizes reach the centimeter scale, and there is almost no dendritic growth. Lei Zhang et al.[25]combined the above two strategies by modifying superhydrophobic SAMs on patterned micropillars, obtaining oriented TIPS-PEN single crystal arrays via blade coating, with mobility as high as 6.8 cm2/(V·s).
Similarly, Chan et al.[33]utilized the blade coating method to prepare C8-BTBT films with a thickness of only a few monolayers by selecting appropriate solvents, temperatures, and coating speeds, achieving crystal domain sizes at the millimeter scale. The OFET devices fabricated based on these films exhibited an average mobility of 13.7 cm2/(V·s) and demonstrated good uniformity on cm-sized substrates. Li Hanying et al.[78]prepared C8-DNTT monolayer films using the solution shearing method, where the crystal domain sizes could easily exceed 5 mm2(actually reaching up to 16 mm2). The OFET arrays (containing 95 devices) fabricated within a single crystal domain achieved a mobility as high as 11 cm2/(V·s), with a coefficient of variation (CV) of only 2.57%. These results indicate that the blade coating method holds great potential for preparing large-area uniform organic semiconductor single-crystal films and organic electronic devices.
Slot-die coating is sometimes referred to as zone casting[79-80], and can be regarded as an improvement of the doctor blade coating method. In the traditional doctor blade coating method, a fixed amount of solution is dropped only before the coating begins; therefore, in the case of large-area continuous preparation, problems such as insufficient solution in the later stage and unstable meniscus shape occur. The hollow die used in slot-die coating can not only serve as a coating tool to pull the meniscus but also allow solution to flow through, solving the problem of continuous liquid supply. Takeya et al.[81]introduced a continuous liquid supply channel into the doctor blade to invent the continuous edge casting method. By precisely controlling the solution concentration, substrate moving speed, and substrate temperature, they obtained C8-DNBDT-NW ultra-thin films with large-area coverage, featuring continuous single-crystal domains several centimeters long and a thickness of only 1~2 monolayers (Figure 4).
Similarly, Park et al.[82]invented a meniscus-guided coating (MGC) method based on continuous flow microfluidic channels (CoMiC). Compared with the traditional solution shearing method, CoMiC introduces a liquid supply channel into the blade, where the continuous solution supply allows the meniscus to maintain a stable shape throughout the coating process. Furthermore, the movement of the contact line at the edge of the meniscus is also influenced by the fluid behavior within the supply channel. Through hydrodynamic simulations, they found that introducing staggered herringbone grooves (SHM) within the channel can effectively enhance convection, thereby increasing the mass transfer rate and ensuring synchronized movement of the crystal growth front and the contact line. Consequently, compared with the traditional solution shearing method, the TIPS-PEN ribbon-like single crystals grown via SHM-CoMiC exhibit more regular orientation, and the performance uniformity of the in-situ fabricated thin-film transistor devices is significantly improved (the coefficient of variation (CV) of mobility decreased from 61.2% to 7.3%).

3.4 Liquid-mediated growth method

Liquid-surface-mediated growth is an emerging method for preparing two-dimensional organic semiconductor single crystals. Unlike other solution-based methods, this approach utilizes the liquid surface as a growth substrate. The liquid surface possesses molecular-level flatness, avoiding the adverse effects of surface defects and roughness of traditional solid substrates on crystal nucleation and growth. Selecting an appropriate liquid can significantly enhance the two-dimensional spreading of the organic semiconductor solution on the liquid surface, effectively suppressing three-dimensional molecular assembly in the vertical direction, ultimately forming organic single crystals with a thickness of only a few or even a single molecular layer.
Hu Wenping et al.[16]reported a "solution epitaxy" method for obtaining large-area two-dimensional organic semiconductor single crystals on water surfaces. Asshown in Figure 5a, this method consists of two steps: (1) Several tens of microliters of organic semiconductor solution are dropped onto the water surface, where the droplets spread out under the high surface tension of water. Organic semiconductor molecules aggregate via π-π interactions and evolve into micrometer-scale two-dimensional organic single crystals, which serve as seeds for the next step; (2) Continue dropping the solution to enable epitaxial growth of the two-dimensional organic single crystals, ultimately yielding millimeter- or even centimeter-scale two-dimensional organic single crystals. Subsequent studies have shown that adding surfactants can further promote the spreading of organic semiconductor solutions on the water surface, enhancing space-confined self-assembly (SCS), thereby obtaining large-area two-dimensional organic single crystals in a single step (Figure 5b)[14]. Additionally, using glycerol with higher viscosity instead of water can prevent droplet splitting and displacement (Figure 5c, d)[17]. Therefore, simultaneously using surfactants and glycerol can maximally promote the spreading of organic semiconductor solutions on the liquid surface (Figure 5e), thereby preparing various large-area two-dimensional organic semiconductor single crystals (Figure 5f~i). Similarly, Jie Jiansheng, Zhang Xiujuan, et al.[15]used glycerol as a liquid substrate and obtained wafer-scale 2,8-difluoro-5,11-bis(triethylsilyl ethynyl) anthradithiophene (diF-TES-ADT) two-dimensional single-crystal films through anchored crystal-seed epitaxial growth (ACSEG). OFET arrays (7×8 devices) fabricated based on these films exhibited a reliable mobility of up to 7.35 cm2/(V·s) with a CV of only 8.9%.

4 Solution crystallization control and morphology optimization strategies

The emergence of various solution preparation techniques has promoted the application of organic semiconductor single crystals in low-cost, high-throughput, and large-area organic electronic devices. The morphology of organic semiconductor single-crystal thin films is crucial for their device performance; for instance, the performance of OFET devices is influenced by factors such as the interface quality between the film and the dielectric layer, the contact between the film surface and the electrodes, grain boundaries, and thickness.[83-84]Therefore, precisely controlling the thin-film morphology during the solution deposition process is key to improving device performance. However, actual film morphology remains affected by fabrication process conditions. Compared with vapor-phase methods, solution methods offer more adjustable parameters, thereby providing greater scope for optimizing film morphology. To optimize the morphology of solution-processed organic semiconductor single-crystal thin films, it is necessary to deeply understand and control their solution crystallization process. In recent years, numerous strategies for regulating and optimizing film morphology have been proposed, such as fluid flow control, solvent composition adjustment, evaporation rate control, meniscus engineering, and contact line engineering. This section will introduce the fundamental principles of these strategies and summarize universal regulation approaches applicable to various solution preparation techniques, providing guidance for controlling the solution crystallization process of organic semiconductors and optimizing film morphology.

4.1 Substrate surface modification

The substrate is an indispensable condition in the preparation of organic semiconductor single crystals and serves as the primary site for crystallization. During vapor deposition, organic semiconductor molecules nucleate and grow directly on the substrate surface; the roughness of the substrate surface and the interaction between the substrate and semiconductor molecules significantly influence crystallization behavior. SAMs or polymers are commonly used to modify the substrate surface to reduce the number of heterogeneous nuclei.[84]. Studies have found that surface treatments with low surface energy (e.g., modification with octadecyltrichlorosilane (OTS)) often increase the grain size of polycrystalline films and device mobility.[85-87]. However, during solution deposition, organic semiconductor molecules are first dissolved in a solvent and then dropped onto the substrate. The surface properties of the substrate indirectly affect the crystallization of organic semiconductor molecules by influencing the wetting of organic droplets and the contraction of the contact line.
Janneck et al.[26]systematically investigated the effects of different substrate surface treatments on solution-processed organic semiconductor crystal films. Asshown in Figure 6, continuous and uniform striped single-crystal films can only be obtained when the substrate exhibits good wettability to the solvent, with a contact angle less than 5°. This is because a smaller contact angle increases the specific surface area at the tip of the meniscus, enhancing solvent evaporation, which leads to supersaturation and nucleation of the solute at the contact line along the meniscus edge. Conversely, when the contact angle is larger, solvent evaporation at the meniscus tip is suppressed, preventing solute precipitation. In addition to high-surface-energy treatment of solid substrates, the liquid-interface-mediated method introduced earlier also typically requires using a liquid substrate with high surface tension or adding surfactants to improve the wetting and spreading of organic droplets on the liquid substrate surface. Thus, ensuring good wettability of the organic solution on the substrate is key to preparing organic semiconductor single-crystal films; high-surface-energy surface treatments are more conducive to the wetting and spreading of organic droplets on the substrate.
Based on the phenomenon that thin film deposition occurs only on wetted surfaces, precise control over the location of crystal nucleation and growth can be achieved by designing reasonable wetted and non-wetted regions on the substrate. Bao Zhenan et al.[8]modified patterned wetted and non-wetted regions on a silicon substrate using phenyltrichlorosilane (PTS) and OTS, respectively (Figure 7a). During the initial coating stage, since the droplet wets only the triangular region, the droplet contact line is pinned at the triangle's boundary, causing its shape to change accordingly. The curvature of the contact line increases at the triangle's tip, accelerating solvent evaporation, which leads to nucleation preferentially occurring at the tip. Subsequently, continuous slits prevent further growth of randomly nucleated crystals at other locations. Ultimately, organic semiconductor films prepared using this patterned wetted substrate exhibit highly oriented crystalline domains. Furthermore, the asymmetric triangular design can also prevent the formation of twin grain boundaries. Similarly, Jie Jiansheng et al.[88]designed a funnel-shaped wetted surface such that misoriented crystals in the Crystal Orientation Filtering Region (COFR) stop growing at the wetting/dewetting boundary during the initial stage, thereby selecting crystals growing parallel to the blade-coating direction. These selected crystals then serve as seeds to continue growing in the Seed Crystal Expansion Region (SCER) in the middle section, ultimately yielding single-oriented organic semiconductor single-crystal films in the Crystal Growth Region (CGR) (Figures 7b–f).
Patterned wettability modification primarily regulates crystal nucleation and growth by altering the shape of the contact line (or meniscus). The shape of the contact line can also be controlled via templated substrates. Jiansheng Jie et al.[23]proposed the Channel-restricted meniscus self-assembly (CRMS) strategy. This strategy utilizes photoresist to fabricate microchannels with a width of 5 μm and a depth of 800 nm on silicon substrates. These channels divide the larger meniscus front into a series of smaller meniscus fronts and confine them within the channels, thereby guiding crystals to nucleate and grow uniformly along the photoresist sidewalls, avoiding uneven nucleation and dendritic growth that occur at larger meniscus fronts.

4.2 Optimization of Solution Preparation Process Parameters

The solution preparation of organic semiconductor single-crystal thin films is essentially a process where solvent evaporation leads to supersaturation, subsequently causing crystal precipitation. Due to the faster solvent evaporation rate at the three-phase contact line (substrate-solution-air) at the droplet edge, solute precipitates in this region, and crystals continuously deposit along the direction of the moving contact line to form a thin film. To obtain oriented crystals, Li Hanying et al.[10]pointed out that, on one hand, a high nucleation density is required to promote the formation of solute depletion zones between nuclei; on the other hand, stable contraction of the contact line must be ensured. To increase nucleation density, a sufficiently high solution concentration (typically greater than 10-1 mg/mL) is needed, and the substrate should be heated to accelerate solvent evaporation. However, excessively high concentrations can lead to the coffee ring effect[89], while excessively high substrate temperatures can cause overly rapid solvent evaporation[90]. These factors all contribute to unstable contact line contraction. Therefore, in methods involving static evaporation of droplets on a substrate, such as the drop-casting method and the DPC method, the primary process parameters requiring adjustment are solution concentration and substrate temperature.
In the MGC method, droplets are dragged by the coating tool, and the movement of the contact line is also influenced by the coating speed (substrate moving speed), requiring additional adjustment of this parameter. Due to differences in solubility, solvent boiling point, and other aspects among different organic semiconductor solution systems, the optimal ranges for corresponding process parameters usually vary. Furthermore, different parameters often require coordinated adjustment, leading to a situation where repeated trial and error is frequently necessary during actual film preparation to determine the optimal combination of process parameters. To reduce the cost of parameter adjustment and accelerate film morphology optimization, it is essential to deeply understand the impact of different process parameters on the film deposition process and morphology, thereby providing theoretical guidance for determining the optimal process parameters in advance. Given that the MGC method offers more adjustable parameters and holds greater promise for industrial application, current research on process parameter optimization is primarily based on this method.[27-29,91-94].
Coating speed is a key parameter affecting film deposition in the MGC method[93-94]. Jang et al.[28]argued that in dip coating, the coating speed should match the supply of solute near the contact line and the evaporation of the solvent, with the latter being related to the substrate temperature and the solvent boiling point. Janneck et al.[92]provided a general method for quantitatively predicting the optimal coating speed under different substrate temperatures and solvent boiling points. They first proposed a model to calculate the retraction speed of the edge of a pure solvent droplet:
${v}_{\mathrm{e}\mathrm{f}\mathrm{e}}=A{V}_{\mathrm{M}}{T}^{0.7}\mathrm{e}\mathrm{x}\mathrm{p}\left(-\frac{\mathrm{\Delta }{S}_{\mathrm{v}\mathrm{a}\mathrm{p}}}{R}\frac{{T}_{\mathrm{b}}}{T}\right)$
among them, VMis the molar volume of the solvent molecule, Tbis the boiling point of the solvent, Aand ΔSvap/Rare fixed constants. Using this model, the droplet shrinkage velocity at different substrate temperatures Tcan be conveniently calculated as vefe. As shown in Figures 8a~h, they further investigated the effect of the ratio between coating speed vcand vefe(${\stackrel{-}{v}}_{\mathrm{c}}={v}_{\mathrm{c}}/{v}_{\mathrm{e}\mathrm{f}\mathrm{e}}$) on the morphology of C8-BTBT crystals. When ${\stackrel{-}{v}}_{\mathrm{c}}<1$, the coating speed is too slow, causing the solute to nucleate in the bulk solution and form microcrystals; when ${\stackrel{-}{v}}_{\mathrm{c}}>1$, the coating speed is too fast, leading to narrower ribbon crystals that eventually form discontinuous isolated islands. Only when ${\stackrel{-}{v}}_{\mathrm{c}}=1$can the optimal film morphology be obtained, where the organic semiconductor single-crystal film almost completely covers the substrate without any microcrystal precipitation. In addition, the film thickness decreases as the blade-coating speed increases. Under various substrate temperatures, solvents, and organic semiconductor systems, ${\stackrel{-}{v}}_{\mathrm{c}}=1$films prepared under the condition of Figures 8i~k).
The initial concentration of the solution also determines the solution supply near the contact line[29,91]. Chan et al.[27]systematically investigated the effects of the above four basic parameters (coating speedv, solute concentrationc, substrate temperatureT, and solvent boiling pointTb) on crystal growth and film morphology, and established their quantitative relationships. TheTbof different solvents can serve as an indicator of the crystal growth rate. The closer the deposition temperature is toTb, the faster the crystal growth rate. They separately adjusted the four coating parameters to prepare films of different thicknesses and correlated crystal growth with these four parameters using the following equation:
$\frac{\mathrm{d}m}{\mathrm{d}t}={\rho }_{\mathrm{o}\mathrm{s}\mathrm{c}}av{t}_{\mathrm{e}\mathrm{q}}=c\frac{\mathrm{d}\forall }{\mathrm{d}t}=F\mathrm{e}\mathrm{x}\mathrm{p}\left(-\frac{\mathrm{\Delta }{S}_{\mathrm{v}\mathrm{a}\mathrm{p}}{T}_{\mathrm{b}}}{RT}\right)$
among them, dm/dtis the mass transfer rate, ρoscis the density of the organic semiconductor, ais the width of the substrate, teqis the equivalent film thickness, $\mathrm{d}\forall /\mathrm{d}t$is the solvent evaporation rate, Fis a collection of solvent properties, geometric factors, and solute concentration. They pointed out that vmainly affects the film thickness and is inversely proportional to it, cis directly proportional to the crystal growth rate. Furthermore, crystal growth is a temperature-dependent thermodynamic process, whose energy barrier is related to Tb. By adjusting these four parameters, different film morphologies can be obtained. Although current research findings cannot yet directly provide complete optimal coating parameters, they are sufficient to help us narrow down the parameter adjustment range and provide direction for optimizing different morphologies.

4.3 Solvent properties and solvent-molecule interactions

Selecting an appropriate solvent is a prerequisite for preparing high-quality organic semiconductor thin films. When preparing organic semiconductor solutions, the primary condition for the solvent is good solubility for specific semiconductor molecules. The concentration of semiconductor solutions used for solution-processed thin films is typically around 10-1~10 mg/mL. Commonly used organic solvents include phenyl-containing toluene and xylene, as well as alkanes such as hexane and heptane, along with their halogenated derivatives. As mentioned earlier, the surface tension of the solvent affects the wetting performance of droplets on the substrate. Janneck et al.[26]believed that the polar component of the surface tension of common organic solvents is relatively small, and wettability is mainly determined by the dispersion component; solvents with a small dispersion component of surface tension are more conducive to droplet wetting, thereby forming uniform thin films. Additionally, Section 4.2 of this paper also discusses the influence of the solvent's boiling point on droplet evaporation and contact line retraction. Besides solubility, surface tension, and boiling point, other characteristics of the solvent also affect the crystallization of organic semiconductors and the thin film deposition process, and the same characteristic may simultaneously influence multiple kinetic and thermodynamic processes.
Park et al.[31]research indicates that the geometric shape of solvent molecules determines the C60crystal shape obtained after droplet evaporation and drying. Asshown in Figure 9a, the crystal morphologies obtained from pseudo-one-dimensional solvents (p1D), pseudo-two-dimensional solvents (p2D), and pseudo-three-dimensional solvents (p3D) exhibit zero-dimensional dot-like, one-dimensional linear, and two-dimensional sheet-like shapes, respectively. Li Hanying et al.[10]used the DPC method with m-xylene as the solvent to obtain oriented needle-like C60crystals with a similar one-dimensional shape (Figure 9b, c). TEM and SAED results (Figure 9d) indicate that m-xylene molecules are embedded in the C60lattice[95]. They further used a mixture of m-xylene and carbon tetrachloride as the solvent to obtain ribbon-like C60crystals intermediate between one-dimensional linear and two-dimensional sheet shapes (Figure 9e~g).
The flat surface and high coverage of these ribbon-like crystals facilitate larger contact areas for OFET devices. The influence of solvents on the geometry of organic semiconductor crystals in C60systems is particularly significant. This is because solvent molecules easily adsorb onto the surface of icosahedral spherical molecules, affecting the self-assembly process[95-97]. Jianian Yao, Hongbing Fu, et al.[98]found that carbon disulfide solvent molecules easily insert into the C60lattice, forming a cage-like structure and guiding its unidirectional growth. However, other organic semiconductor molecules typically possess planar symmetric structures with strong anisotropic π-π interactions; their self-assembly behavior is primarily determined by their own π-π stacking. Therefore, common semiconductor molecules such as TIPS-PEN and C8-BTBT can form ribbon-like crystals in both one-dimensional alkanes and two-dimensional aromatic solvents, with no solvent molecules present in the lattice[27,77,99]. Nevertheless, Zhenan Bao et al.[30]also discovered that the molecular volume of the solvent is related to the formation of the TIPS-PEN metastable crystal polymorph. The larger the solvent molecular volume, the more the crystal polymorph deviates from the equilibrium state. This is mainly because larger solvent molecules promote vertical spatial confinement effects. They speculated that the increase in solvent molecular volume may, on one hand, alter the Gibbs free energy of the system due to reduced surface tension, and on the other hand, may force organic semiconductor molecules to undergo larger displacements or conformational adjustments, thereby directly affecting the crystallization kinetics. However, the specific microscopic mechanism remains unclear.
Vladimirov et al.[32]found thatN,N′-di((S)-1-methylpentyl)-1,7(6)-dicyano-perylene-3,4:9,10-bis(dicarboximide) (PDI1MPCN2) growth in droplets is not only related to the solvent surface tensionγLA, but also influenced by the solvent viscosity η. Asshown in Figures 10a~c, when using highγLAand high η solvents, PDI1MPCN2 crystals grow sheet-like at the gas-liquid interface, forming large-area thin films; conversely, when using lowγLAand low η solvents, discrete blocky three-dimensional crystals are formed. By adding polystyrene to the solution to adjust η alone, they found that neither highγLAnor high η can independently induce crystallization at the gas-liquid interface. They believe that highγLAis the main driving force for heterogeneous nucleation and growth of crystals at the gas-liquid interface, while high η can slow down the Marangoni flow caused by surface tension gradients, thereby facilitating controlled crystallization at the gas-liquid interface.
Marangoni flow is a fluid mechanics phenomenon referring to liquid flow caused by surface tension gradients at the gas-liquid interface due to uneven heating of droplets.[100].As shown in Figure 11a, the solution temperature in regions far from the substrate heat source is lower, and its surface tension is higher than that of the solution near the contact line (the surface tension of solvents is typically inversely proportional to temperature). The Marangoni effect causes the solution to flow from low surface tension regions to high surface tension regions, carrying away solute molecules near the contact line, thereby inhibiting the formation of continuous films. To overcome the negative Marangoni effect caused by temperature differences in single-solvent systems, Chan et al.[33]proposed using mixed solvents (m-xylene and o-dichlorobenzene) to counteract this. Asshown in Figure 11d, since o-dichlorobenzene is less volatile than m-xylene, its concentration near the contact line is higher. Meanwhile, due to the higher surface tension of o-dichlorobenzene, it pulls organic solute molecules toward the contact line, thereby counteracting the thermally driven Marangoni flow. Therefore, using mixed solvents can achieve higher coating speeds, smaller film thicknesses, and better film coverage.
In summary, the various properties of solvents (such as geometric shape, molar volume, viscosity, and surface tension) exert a highly complex influence on the crystallization behavior of organic semiconductor solutions, encompassing both kinetic and thermodynamic effects. Therefore, appropriate solvent selection requires comprehensive consideration from multiple perspectives. However, due to the mutual coupling of different solvent properties, it is experimentally difficult to independently adjust a specific property. This not only complicates solvent selection but also makes investigating the mechanisms by which solvents affect the solution crystallization of organic semiconductor single crystals and thin film deposition extremely challenging.

4.4 Ideal Single Crystal Morphology and Common Control Strategies

To maximize the performance potential of organic semiconductors, it is crucial to prepare single-crystal thin films with specific ideal morphologies. Typically, ideal organic semiconductor single-crystal thin films must possess three key characteristics: uniform orientation, continuous coverage, and essentially consistent thickness. First, uniform orientation directly determines the consistency of device performance. Due to the significant anisotropy of organic crystals, carrier mobility is highest along specific crystallographic axes. By achieving high orientational consistency of all crystalline domains on the substrate, grain boundary scattering can be effectively eliminated and internal traps reduced. Second, large-area, continuous, and crack-free single-crystal thin films form the foundation for scalable manufacturing, significantly reducing the risk of leakage current. Continuous single-crystal thin films are not only applicable to lateral structure devices such as organic field-effect transistors but also hold promise for vertical structure devices including organic light-emitting diodes, organic photovoltaics, organic photodetectors, and organic spintronic devices. Finally, thickness uniformity, especially in the fabrication of monolayer or few-layer molecular crystals (MMCs), can effectively suppress bulk resistance effects, improve electrode contact, and enhance charge injection efficiency.[101]. To achieve the aforementioned ideal morphologies, researchers have developed various innovative regulation strategies over the past decade, focusing on the precise control of nucleation kinetics, fluid dynamics, and intermolecular interactions.
For organic semiconductor molecules with specific fast-growing crystal axes, such as TIPS-PEN, meniscus-guided coating can effectively screen for dominant orientations and terminate nuclei with slow-growing orientations, thereby obtaining organic single-crystal films with high overall orientation uniformity[77]. However, for molecules like C8-BTBT that possess multiple in-plane dominant growth crystal axes, their uniformity is often poor. Chen Guoliang et al.[102]found that C8-BTBT films prepared by single-pass bar coating exhibit botha-axis andb-axis orientations, and both orientations can stably grow to centimeter-scale lengths at the same coating speed; however, due to the anisotropy of the crystal structure, the intrinsic mobilities of thea-axis andb-axis differ significantly, resulting in macroscopic mobility non-uniformity (Fig. 12a–e). By rotating 90° and selecting a strategy where the crystallization starting line of the second coating lies within the same orientation-uniform region, the team achieved secondary growth of existing crystals and successfully obtained inch-scale orientation-uniform C8-BTBT films (Fig. 12f–j), with a mobility as high as 14.9 cm2/(V·s) and a relative standard deviation of mobility (i.e., coefficient of variation) as low as 8.7%[102].
Li Rongjin et al.[103]proposed a stepwise strategy combining high supersaturation nucleation with low supersaturation growth, developing a novel in-situ seeded dip coating (SDC) technique. Stable C10-DNTT nuclei were first obtained through residence at the growth front and the classical Ostwald ripening effect, subsequently enabling precise control of supersaturation to achieve continuous growth of existing nuclei (Figure 13). The wafer-scale single crystals prepared via this method exhibit monolayer molecular thickness (3.3 nm), smooth surfaces (0.3 nm), and low trap density (1.3 × 10-10 cm-3). Organic field-effect transistor arrays fabricated based on these wafer-scale single crystals demonstrate an average mobility as high as 14.5 cm2/(V·s), with a relative standard deviation of mobility of only 4.5%, representing a leading level among large-area devices prepared by solution methods[103].
Organic semiconductor molecules designed for efficient carrier transport generally exhibit high in-plane carrier mobility, but suffer from poor interlayer transport in the out-of-plane direction, especially for solution-processable organic small-molecule semiconductors with unilateral or symmetric alkyl substitution. On this basis, the thickness of single-crystal thin films, or more precisely, the number of molecular layers, will have a very significant impact on OFET performance. Takeya et al.[81], Chen Guoliang et al.[104], Wang Xinran et al.[56], Hu Wenping and Li Rongjin et al.[14,17]have all reported that monolayer or bilayer organic single-crystal thin films possess the most ideal transport characteristics, simultaneously satisfying high mobility and low contact resistance. If monolayer single crystals have lattice constants and molecular conformations similar to those of multilayer molecular single crystals, then the monolayer exhibits the best overall performance; conversely, if factors such as interface dipole induction, interface trap states, or lattice distortion caused by out-of-plane symmetry breaking lead to a decrease in mobility in monolayer single crystals, then the bilayer single crystal exhibits the best overall performance. However, regardless of the situation, organic single-crystal thin films with uniform molecular layer counts over large areas will exhibit ideal mobility uniformity. Among these, Jiang Lang et al.[105]reported that monolayer single-crystal OFET arrays patterned via screen-printed steel mesh nodes achieved a relative standard deviation of mobility as low as 4.4%. Li Hanying, Peng Boyu et al.[78]reported that OFET arrays based on single-domain monolayer single crystals exhibited a relative standard deviation of mobility as low as 2.57%, both approaching or reaching the performance uniformity of inorganic transistors.
Theoretically, organic single-crystal thin films with uniform orientation and thickness have largely met the performance and uniformity requirements for organic field-effect transistors or other planar transport devices. However, to further exploit organic single crystals in micro/nano-scale planar devices (such as integrated circuits and sensor arrays) or vertical devices (such as light-emitting diodes and photovoltaics), organic single-crystal thin films with 100% coverage have become a critical requirement. Consequently, the anisotropy of crystallization and its regulation mechanisms have emerged as key scientific issues. In reported literature, the coverage of organic single-crystal thin films generally depends on the intrinsic crystalline anisotropy of the organic small-molecule semiconductors themselves. For instance, alkyl-substituted molecules such as C8-BTBT, C10-DNTT, and NDI-8CN2 often form large-area fully covered single-crystal films, or even large-area single-molecular-layer single-crystal films. In contrast, a class of organic small-molecule semiconductors represented by TIPS-PEN typically forms ribbon-like single-crystal films; the specific determining mechanisms and regulation methods remain unclear. Recently, Li Hanying et al.[106]proposed a regulation mechanism wherein solvent surface tension controls the anisotropy of single-crystal growth. They found that high surface tension solvents (such as anisole) can effectively promote the transformation of TIPS-PEN-like single-crystal morphology from ribbon-like single crystals into two-dimensional continuous single-crystal thin films. Conversely, in low surface tension solvents (such as m-xylene), C8-BTBT forms ribbon-like single-crystal thin films. Compared to ribbon-like single-crystal thin films, two-dimensional continuous TIPS-PEN single-crystal thin films exhibit uniformly high mobility in both parallel and perpendicular coating directions. The proposal of this mechanism establishes a theoretical foundation for the universal regulation of organic single-crystal thin film morphology and the enhancement of coverage, holding promise for further application in vertical organic electronic devices based on organic single-crystal thin films.
Based on the research progress in advanced preparation methods and morphology control strategies for organic single crystals mentioned above, some common control strategies for solution-based preparation of organic single crystal films can be summarized: (1) A stable crystallization region provided collectively by substrates with high surface energy and smoothness, solvents with high surface tension and high boiling points, a stable contact line movement velocity, and minimized flow field instabilities caused by effects such as temperature differences, surface viscosity, capillary phenomena, and Marangoni flows; (2) Nucleation and growth processes precisely controlled based on solution supersaturation, along with control of the crystallization interface and crystallization anisotropy based on solvent-small molecule interactions, enabling organic single crystal films to preferentially achieve an in-plane two-dimensional continuous and out-of-plane layer-count controllable growth mode; (3) Ultra-large domain single single-crystal or regularly oriented single-crystal films combining strategies such as surface microstructures, nucleus orientation screening, and multi-step or secondary growth, to minimize mobility non-uniformity caused by inconsistent orientations and structural or performance instabilities caused by grain boundary defects. Among these, (1) hydrodynamic-related factors can generally be studied via finite element simulations of computational fluid dynamics; (3) single-crystal growth behaviors are typically studied through post-mortem optical and electrical characterization; however, (2) the study of nucleation behavior and crystallization anisotropy based on solvent-molecule interactions faces a dilemma in experiments due to the difficulty of in-situ observation and in simulations due to the lack of multi-scale research tools.

5 Molecular dynamics simulation of thin film deposition and crystallization

Currently, research on the crystallization behavior during solution deposition of organic semiconductor thin films has mostly focused on observing the impact of initial growth conditions on the final crystal morphology, or on in situ observation of the droplet shrinkage process at the macroscopic level using techniques such as optical microscopy. Although these studies have provided ideas and strategies for regulating crystal growth and thin film deposition processes, they struggle to elucidate the microscopic crystallization mechanisms at the molecular level. Although AFM technology has been used to observe the crystallization behavior of organic molecules in real time at the nanoscale[54,107-108], the information provided is limited and difficult to combine with the solution coating process. To overcome the limitations of experimental methods and gain a deep understanding of the microscopic crystallization process, molecular dynamics (MD) simulations, as a powerful complementary tool, can provide a microscopic research perspective with atomic-level precision[109-110].
In recent years, some studies[111-113]have established solution models containing explicit solvent molecules and simulated the formation process of organic semiconductor films during solvent evaporation using MD. Other studies[114]have simulated the effects of evaporation rates in real drop-casting methods and shear flow in MGC methods on film morphology by controlling solvent evaporation at the gas-liquid interface and introducing velocity gradients. However, due to significant limitations in the time and spatial scales of current MD simulations, these non-equilibrium simulation studies often require conditions such as solution concentrations, evaporation rates, and shear rates that are far higher than actual situations. This results in final simulated films with relatively low crystallinity, differing significantly from the morphology of actual organic semiconductor single-crystal films. Therefore, the information provided by studying the non-equilibrium kinetic processes of crystallization via MD simulations remains very limited. Meanwhile, with the rapid development of enhanced sampling techniques, MD has been widely applied to study and predict the thermodynamic behavior during the crystallization process of organic molecules through free energy calculations. By decomposing the complex crystallization process into several key steps and establishing appropriate equilibrium models, it is possible to provide thermodynamic information that is difficult to obtain experimentally. In response to the aforementioned research background and challenges, this paper aims to combine MD simulations with experimental methods to systematically investigate the microscopic mechanisms of organic semiconductor solution crystallization based on solvent surface tension and the laws governing macroscopic morphology regulation, thereby providing theoretical guidance for solvent selection and film optimization.
With improvements in algorithms and the development of computer hardware and software, studies using MD simulations for systems containing hundreds of thousands of atoms have become commonplace in recent years, with time scales even reaching the microsecond level. Therefore, MD is highly suitable for studying nanoscale crystallization and thin film growth of organic semiconductor molecules, processes dominated by weak intermolecular interactions and insensitive to electronic structure.
MD simulations are commonly used to study the vapor deposition behavior of pentacene[115-119]. A key issue in the vapor-phase growth of pentacene films is the tendency to form three-dimensional stepped terraces on the surface. This is mainly due to the Ehrlich-Schwoebel barrier[120-121], which hinders surface-adsorbed molecules from crossing step edges to migrate downward, thereby forming pyramid-like structures. In 2007, Goose et al.[115]used MD simulations to investigate the effects of molecular orientation, incident kinetic energy, and angle on the film growth mechanism during collisions between incident molecules and step edges of pentacene films. The study showed that molecules with high kinetic energy and vertical incidence favor the two-dimensional growth of pentacene; when molecular orientations in the beam are random, higher incident kinetic energy is required. These results validate the experimental principle of using inert carrier gases to accelerate incident molecules to promote the two-dimensional film growth of pentacene[122]. With the continuous development of computational power, MD simulations can provide increasingly deep research perspectives. In 2018, Roscioni et al.[118]directly simulated the nucleation and growth processes of two complete monolayers of pentacene on silica. They found that the growth of pentacene crystalline films occurs in two stages: first, a small number of lying-down molecules aggregate to form long-lived, partially diffusive clusters; second, when the cluster size reaches a critical value, a transition from lying-down to standing-up occurs, accompanied by the emergence of an ordered crystalline structure. Thereafter, the growth of crystalline islands follows the standard nucleation and growth model via the attachment of new molecules to the island edges. Furthermore, they obtained the Ehrlich-Schwoebel barriers experienced by pentacene at various step edges through free energy calculations, confirming the layer-dependent dependence observed in experiments[123], and additionally elucidated the influence of step orderliness on the magnitude of the Ehrlich-Schwoebel barrier. Thus, MD simulations can not only observe the growth process of organic semiconductor molecules in situ from a microscopic perspective and explain the underlying crystallization mechanisms but also directly calculate thermodynamic numerical changes that are difficult to observe experimentally. Some literature has also investigated the effects of substrate temperature and surface modification on the morphology of vapor-deposited organic semiconductor films[124-125]. The mutual validation between these simulation results and experimental phenomena demonstrates the feasibility of MD simulations in studying the preparation processes of organic semiconductor films.
Compared to vapor deposition processes, simulating solution deposition processes is more challenging for the following reasons: on one hand, the large number of solvent molecules leads to a surge in computational cost; on the other hand, the simulation scales currently achievable with computing power are insufficient to reproduce kinetic processes such as solvent evaporation and mass transport in real droplets. Nevertheless, a small number of studies have indicated that investigating solvent evaporation processes via MD simulations is feasible. For example, Alessandri et al.[111]employed a coarse-grained model to simulate the formation process of poly(3-hexylthiophene)-phenyl-C61-butyric acid methyl ester (P3HT-PCBM) bulk heterojunction films in chlorobenzene solution. Compared to all-atom models (which treat atoms as the smallest moving units), coarse-grained models treat groups of atoms as single entities for calculation, thereby significantly reducing the number of particles required for computation and enabling the extension of simulation time scales up to 14 μs. Meanwhile, they employed a method of periodically and randomly removing a portion of solvent molecules to simulate the solvent evaporation process. Gertsen et al.[112]also adopted a similar artificially assisted evaporation scheme in all-atom MD simulations to study the film formation process of the non-fullerene acceptor IDTBR during solvent evaporation. However, Alessandri et al. also pointed out that such assisted evaporation leads to a uniform distribution of the molecules to be evaporated within the system, which contradicts the actual situation where solvent evaporation occurs only at the gas-liquid interface.
To avoid this issue, Lee et al.[113]introduced a vacuum layer above the solution surface, allowing solvent molecules to naturally escape at the gas-liquid interface. By periodically removing solvent molecules within 2 nm of the escaping gas-liquid interface, the volatilization and drying process of the solution can be simulated more realistically. In addition to solvent volatilization at the gas-liquid interface, Guthrie et al.[114]also applied an additional velocity to the solution in the surface layer to simulate the shear flow caused by the directional movement of the coating tool in the MGC method. They believe that the slower volatilization rate in the drop-casting method is beneficial for molecular orientation in the film (Figure 14c); whereas in the MGC method, rapid shear flow alters the molecular orientation in the film (Figure 14f).
However, limited by current computational power, traditional MD simulations struggle to fully capture and reproduce the nucleation and growth processes of crystals within their typical timescales (nanoseconds to microseconds). This is because high energy barriers often exist between different structures or conformations during crystallization, causing the system to remain trapped in local energy minima for extended periods, leading to insufficient sampling of the entire conformational space (especially high-energy regions). To improve the sampling efficiency of MD simulations, various enhanced sampling methods have been developed in recent years[126]. Enhanced sampling algorithms introduce bias potentials to modify the system's potential energy surface, thereby lowering energy barriers and increasing the sampling of these high-energy transition events. Common enhanced sampling methods include, but are not limited to: Umbrella sampling, Adaptive biasing force (ABF), Metadynamics (MetaD), and Variationally enhanced sampling (VES). Among these, many enhanced sampling methods require the pre-definition of suitable Collective variables, also known as Reaction coordinates, to parameterize and effectively guide the evolution process of interest. Collective variables are typically functions of atomic coordinates and must be capable of effectively describing the evolution process of concern. For example, as mentioned earlier, Roscioni et al.[118]used the three-dimensional coordinates of the center of mass of a pentacene molecule as collective variables to calculate the free energy changes associated with its movement along and across step edges using the ABF algorithm, thereby obtaining the magnitude of the Ehrlich-Schwoebel barrier. To describe the crystallization process of organic molecules more comprehensively, the definition of collective variables must consider not only intermolecular assembly but also intramolecular orientation. However, overly complex collective variables not only significantly increase computation time but also impair the effectiveness of enhanced sampling[127]. Therefore, enhanced sampling simulations based on collective variables still require simplification of the crystallization process. For instance, Bjelobrk et al.[128]defined the growth of different crystal faces of Naphthalene in solution as Surface layer crystallinity:
${s}_{\mathrm{s}\mathrm{l}\mathrm{c}}=\sum _{i=1}^{N}{s}_{\mathrm{s}\mathrm{l}\mathrm{c},i}=\sum _{i=1}^{N}{\varphi }_{i}{\rho }_{i}{\mathrm{\Theta }}_{i}$
among them, ${\varphi }_{i}$, ρi, Θirepresent whether the i-th solute molecule is within the surface layer, the local density within the surface layer, and the local order, respectively; they are nonlinear functions of the solute atomic coordinates. When the surface layer is completely disordered, sslc=0; whereas when solute molecules continuously self-assemble within the surface layer to form an ordered crystal structure, sslcbecomes larger, approaching the number of molecules in a complete crystal layer. Therefore, the free energy change of sslcfrom 0 to Nrepresents the free energy change experienced during crystal face growth. Using sslcfor enhanced sampling calculations yields the free energy surface (Free energy surface, Figure 15a), from which the energy barriers that must be overcome for the growth of different naphthalene crystal faces are obtained. Based on the magnitude of these barriers, the relative growth rates of each crystal face are calculated, and they predicted the final crystal shape of naphthalene growing in bulk ethanol solution (Figure 15c), which is consistent with experimental results. It is evident that through reasonable definition of collective variables and system simplification, MD simulations and enhanced sampling techniques can provide quantitative statistical thermodynamic information for studying the solution crystallization of organic semiconductors, offering deep insights for understanding crystallization mechanisms and predicting crystallization behavior.
Based on the aforementioned development of molecular dynamics simulations, these studies have significantly advanced the microscopic understanding of solution crystallization processes. However, they still face a substantial gap between the system size and time scales in simulations and those in actual experiments. Consequently, molecular dynamics simulations often serve as a new dimension for interpreting existing experimental crystallization results, rather than playing a dominant role in morphology regulation. Currently, given the complex multiphysics and multiscale processes involved in solution crystallization, relying solely on molecular dynamics simulation tools makes it difficult to achieve breakthroughs. New research paradigms, such as artificial intelligence methods and high-throughput experiments, hold promise for bridging the gaps between experiments and simulations, offering new potential solutions for collaborative optimization problems in such complex systems.

6 Conclusion and Outlook

In summary, this article reviews recent advances in the solution-based preparation and morphology control of organic semiconductor single-crystal thin films. Various simple and low-cost solution methods have been successfully applied to fabricate high-quality, uniformly oriented organic semiconductor single-crystal thin films. In particular, solution methods represented by meniscus-guided coating demonstrate significant potential for large-area fabrication and device applications. However, due to an incomplete understanding of the solution crystallization mechanisms of organic semiconductors, optimization of film morphology often relies on trial and error. Currently, effective guidance strategies have been proposed regarding substrate surface modification, process parameter optimization, and solvent selection. In the future, to establish more universal guidelines, it is urgent to gain a deeper understanding of the solution crystallization mechanisms of organic semiconductors. Molecular dynamics simulations, as an important auxiliary tool, face limitations in spatiotemporal scales; nevertheless, their microscopic research perspective provides strong support for understanding the solution crystallization mechanisms of organic semiconductors. With the continuous development of computational hardware and software, the accuracy and efficiency of multi-scale simulations will significantly improve, offering more robust support for the in-depth understanding and precise control of the organic semiconductor solution crystallization process. Addressing these challenges will ultimately enable precise control over film morphology and promote the commercialization of organic semiconductors in fields such as flexible electronics. Promising research directions expected to achieve breakthroughs in recent years include: (1) Scalable continuous preparation methods for organic single-crystal thin film materials, such as combining slot-die coating with roll-to-roll printing, along with precise and synergistic optimization of various parameters in the crystallization process, to realize meter-scale or larger fabrication of high-performance organic single-crystal thin films; (2) Investigating the full-scale physicochemical processes of organic single-crystal solution crystallization, namely nucleation, ripening, growth, reversible/irreversible phase transitions, etc., occurring from molecules in free solution until solidification into crystals, and their corresponding thermodynamic and kinetic influence mechanisms; (3) Based on ideal organic single-crystal thin film morphologies, further achieving optimal optoelectronic properties by combining strategies such as defect passivation, physical/chemical doping, co-crystallization, and blending.

References

Publishing order | Descend order by publishing year | Descend order by cited within
[1]

Bronstein H, Nielsen C B, Schroeder B C, McCulloch I. Nat. Rev. Chem., 2020, 4(2): 66.
[2]

Wang C L, Dong H L, Hu W P, Liu Y Q, Zhu D B. Chem. Rev., 2012, 112(4): 2208.
[3]

Facchetti A. Mater. Today, 2007, 10(3): 28.
[4]

Podzorov V. MRS Bull., 2013, 38(1): 15.
[5]

Wang C L, Dong H L, Jiang L, Hu W P. Chem. Soc. Rev., 2018, 47(2): 422.
[6]

Zhang X, Dong H, Hu W. Adv. Mater., 2018, 30(44): 1801048.
[7]

Jiang H, Hu W P. Angew. Chem. Int. Ed., 2020, 59(4): 1408.
[8]

Diao Y, Shaw L, Bao Z N, Mannsfeld S C B. Energy Environ. Sci., 2014, 7(7): 2145.
[9]

Wang S H, Peng L, Sun H B, Huang W. J. Mater. Chem. C, 2022, 10(35): 12468.
[10]

Li H Y, Tee B C, Cha J J, Cui Y, Chung J W, Lee S Y, Bao Z N. J. Am. Chem. Soc., 2012, 134(5): 2760.
[11]

Ren C X, Cao L, Wu T. Small, 2023, 19(22): 2300151.
[12]

Tian X Z, Yao J R, Zhang L J, Han B, Shi J W, Su J W, Liu J, Li C L, Liu X F, Zhai T Y, Jiang L, Yang F X, Zhang X T, Zou Y, Li R J, Hu W P. Nano Res., 2022, 15(3): 2667.
[13]

Wang J, Wu X, Pan J, Feng T, Wu D, Zhang X, Yang B, Zhang X, Jie J. Adv. Mater., 2020, 32(38): 2003315.
[14]

Wang Q, Yang F, Zhang Y, Chen M, Zhang X, Lei S, Li R, Hu W. J. Am. Chem. Soc., 2018, 140(16): 5339.
[15]

Wang J, Ren Z, Pan J, Wu X, Jie J, Zhang X, Zhang X. Adv. Mater., 2023, 35(36): 2301017.
[16]

Xu C H, He P, Liu J, Cui A J, Dong H L, Zhen Y G, Chen W, Hu W P. Angew. Chem. Int. Ed., 2016, 55(33): 9519.
[17]

Yao J R, Zhang Y, Tian X Z, Zhang X L, Zhao H J, Zhang X T, Jie J S, Wang X R, Li R J, Hu W P. Angew. Chem., 2019, 131(45): 16228.
[18]

Kwon S, Kim J, Kim G, Yu K, Jo Y R, Kim B J, Kim J, Kang H, Park B, Lee K. Adv. Mater., 2015, 27(43): 6870.
[19]

Minemawari H, Yamada T, Matsui H, Tsutsumi J, Haas S, Chiba R, Kumai R, Hasegawa T. Nature, 2011, 475(7356): 364.
[20]

Zhao Y J, Sheng Q Y, Ke S F, Wu R H, He L H, Ren X C, Peng B Y, Li H Y. Small, 2024, 20(46): 2404770.
[21]

Kumagai S, Yamamura A, Makita T, Tsurumi J, Lim Y Y, Wakimoto T, Isahaya N, Nozawa H, Sato K, Mitani M, Okamoto T, Watanabe S, Takeya J. Sci. Rep., 2019, 9: 15897.
[22]

Soeda J, Uemura T, Okamoto T, Mitsui C, Yamagishi M, Takeya J. Appl. Phys. Express, 2013, 6(7): 076503.
[23]

Deng W, Zhang X J, Dong H L, Jie J S, Xu X Z, Liu J, He L, Xu L, Hu W P, Zhang X H. Mater. Today, 2019, 24: 17.
[24]

Zhao X, Zhang H, Zhang J, Liu J, Lei M, Jiang L. Adv. Sci., 2023, 10(15): 2300483.
[25]

Xiao C Y, Kan X N, Liu C M, Jiang W, Zhao G Y, Zhao Q, Zhang L, Hu W P, Wang Z H, Jiang L. J. Mater. Chem. C, 2017, 5(10): 2702.
[26]

Janneck R, Heremans P, Genoe J, Rolin C. Adv. Mater. Interfaces, 2018, 5(12): 1800147.
[27]

Chen M, Peng B Y, Huang S Y, Chan P K L. Adv. Funct. Mater., 2020, 30: 1905963.
[28]

Jang J, Nam S, Im K, Hur J, Cha S N, Kim J, Bin Son H, Suh H, Loth M A, Anthony J E, Park J J, Park C E, Kim J M, Kim K. Adv. Funct. Mater., 2012, 22(5): 1005.
[29]

Wu K J, Li H W, Li L Q, Zhang S N, Chen X S, Xu Z Y, Zhang X, Hu W P, Chi L F, Gao X K, Meng Y C. Langmuir, 2016, 32(25): 6246.
[30]

Giri G, Li R P, Smilgies D M, Li E Q, Diao Y, Lenn K M, Chiu M, Lin D W, Allen R, Reinspach J, Mannsfeld S C B, Thoroddsen S T, Clancy P, Bao Z N, Amassian A. Nat. Commun., 2014, 5: 3573.
[31]

Park C, Song H J, Choi H C. Chem. Commun., 2009, 32: 4803.
[32]

Vladimirov I, Kellermeier M, Geßner T, Molla Z, Grigorian S, Pietsch U, Schaffroth L S, Kühn M, May F, Weitz R T. Nano Lett., 2018, 18(1): 9.
[33]

Zhang Z C, Peng B Y, Ji X D, Pei K, Chan P K L. Adv. Funct. Mater., 2017, 27(37): 1703443.
[34]

Chen S N, Li Z, Qiao Y L, Song Y L. J. Mater. Chem. C, 2021, 9(4): 1126.
[35]

Eley D D. Nature, 1948, 162(4125): 819.
[36]

Horowitz G, Garnier F, Yassar A, Hajlaoui R, Kouki F. Adv. Mater., 1996, 8(1): 52.
[37]

Podzorov V, Sysoev S E, Loginova E, Pudalov V M, Gershenson M E. Appl. Phys. Lett., 2003, 83(17): 3504.
[38]

Takeya J, Yamagishi M, Tominari Y, Hirahara R, Nakazawa Y, Nishikawa T, Kawase T, Shimoda T, Ogawa S. Appl. Phys. Lett., 2007, 90(10): 102120.
[39]

Jurchescu O D, Popinciuc M, van Wees B J, Palstra T T M. Adv. Mater., 2007, 19(5): 688.
[40]

Tsumura A, Koezuka H, Ando T. Appl. Phys. Lett., 1986, 49(18): 1210.
[41]

Watanabe M, Chang Y J, Liu S W, Chao T H, Goto K, Islam M M, Yuan C H, Tao Y T, Shinmyozu T, Chow T J. Nat. Chem., 2012, 4(7): 574.
[42]

da Silva Filho D A, Kim E G, Brédas J L. Adv. Mater., 2005, 17(8): 1072.
[43]

Anthony J E. Chem. Rev., 2006, 106(12): 5028.
[44]

Gao J H, Li R J, Li L Q, Meng Q, Jiang H, Li H X, Hu W P. Adv. Mater., 2007, 19(19): 3008.
[45]

Anthony J E, Brooks J S, Eaton D L, Parkin S R. J. Am. Chem. Soc., 2001, 123(38): 9482.
[46]

Park S K, Jackson T N, Anthony J E, Mourey D A. Appl. Phys. Lett., 2007, 91(6): 063514.
[47]

Dong H, Fu X, Liu J, Wang Z, Hu W. Adv. Mater., 2013, 25(43): 6158.
[48]

Payne M M, Parkin S R, Anthony J E, Kuo C C, Jackson T N. J. Am. Chem. Soc., 2005, 127(14): 4986.
[49]

Goetz K P, Li Z, Ward J W, Bougher C, Rivnay J, Smith J, Conrad B R, Parkin S R, Anthopoulos T D, Salleo A, Anthony J E, Jurchescu O D. Adv. Mater., 2011, 23(32): 3698.
[50]

Gao P, Beckmann D, Tsao H N, Feng X, Enkelmann V, Baumgarten M, Pisula W, Müllen K. Adv. Mater., 2009, 21(2): 213.
[51]

Nakayama K, Hirose Y, Soeda J, Yoshizumi M, Uemura T, Uno M, Li W Y, Kang M J, Yamagishi M, Okada Y, Miyazaki E, Nakazawa Y, Nakao A, Takimiya K, Takeya J. Adv. Mater., 2011, 23(14): 1626.
[52]

Jiang H, Kloc C. MRS Bull., 2013, 38(1): 28.
[53]

Reese C, Roberts M, Ling M M, Bao Z N. Mater. Today, 2004, 7(9): 20.
[54]

Chen J J, Zhu E B, Liu J, Zhang S, Lin Z Y, Duan X F, Heinz H, Huang Y, De Yoreo J J. Science, 2018, 362(6419): 1135.
[55]

Hlawacek G, Puschnig P, Frank P, Winkler A, Ambrosch-Draxl C, Teichert C. Science, 2008, 321(5885): 108.
[56]

He D W, Zhang Y H, Wu Q S, Xu R, Nan H Y, Liu J F, Yao J J, Wang Z L, Yuan S J, Li Y, Shi Y, Wang J L, Ni Z H, He L, Miao F, Song F Q, Xu H X, Watanabe K, Taniguchi T, Xu J B, Wang X R. Nat. Commun., 2014, 5: 5162.
[57]

de Boer R W I, Gershenson M E, Morpurgo A F, Podzorov V. Phys. Stat. Sol. (a), 2004, 201(6): 1302.
[58]

Zhang F J, Di C A, Berdunov N, Hu Y Y, Hu Y B, Gao X K, Meng Q, Sirringhaus H, Zhu D B. Adv. Mater., 2013, 25(10): 1401.
[59]

Forrest S R. Nature, 2004, 428(6986): 911.
[60]

Liu C, Li Y, Lee M V, Kumatani A, Tsukagoshi K. Phys. Chem. Chem. Phys., 2013, 15(21): 7917.
[61]

Yuan Y B, Giri G, Ayzner A L, Zoombelt A P, Mannsfeld S C B, Chen J H, Nordlund D, Toney M F, Huang J S, Bao Z N. Nat. Commun., 2014, 5: 3005.
[62]

Kim D H, Han J T, Park Y D, Jang Y, Cho J H, Hwang M, Cho K. Adv. Mater., 2006, 18(6): 719.
[63]

Kim D H, Lee D Y, Lee H S, Lee W H, Kim Y H, Han J I, Cho K. Adv. Mater., 2007, 19(5): 678.
[64]

Goto O, Tomiya S, Murakami Y, Shinozaki A, Toda A, Kasahara J, Hobara D. Adv. Mater., 2012, 24(8): 1117.
[65]

Wang S H, Kappl M, Liebewirth I, Müller M, Kirchhoff K, Pisula W, Müllen K. Adv. Mater., 2012, 24(3): 417.
[66]

Chen J H, Tee C K, Shtein M, Martin D C, Anthony J. Org. Electron., 2009, 10(4): 696.
[67]

Peng B Y, Wu R H, Li H Y. Acc. Chem. Res., 2021, 54(24): 4498.
[68]

Liu S, Wu J K, Fan C C, Xue G B, Chen H Z, Xin H L, Li H Y. Sci. Bull., 2015, 60(12): 1122.
[69]

Hsieh C T, Chen C Y, Lin H Y, Yang C J, Chen T J, Wu K Y, Wang C L. J. Phys. Chem. C, 2018, 122(28): 16242.
[70]

Li H Y, Tee B C, Giri G, Chung J W, Lee S Y, Bao Z N. Adv. Mater., 2012, 24(19): 2588.
[71]

Li H B, Wu J K, Takahashi K, Ren J, Wu R H, Cai H Y, Wang J R, Xin H L, Miao Q, Yamada H, Chen H Z, Li H Y. J. Am. Chem. Soc., 2019, 141(25): 10007.
[72]

Uemura T, Hirose Y, Uno M, Takimiya K, Takeya J. Appl. Phys. Express, 2009, 2(11): 111501.
[73]

Rogowski R Z, Dzwilewski A, Kemerink M, Darhuber A A. J. Phys. Chem. C, 2011, 115(23): 11758.
[74]

Chai Z M, Abbasi S A, Busnaina A A. ACS Appl. Mater. Interfaces, 2018, 10(21): 18123.
[75]

Fitzgerald S D, Woods A W. Nature, 1994, 367(6462): 450.
[76]

Becerril H A, Roberts M E, Liu Z H, Locklin J, Bao Z N. Adv. Mater., 2008, 20(13): 2588.
[77]

Diao Y, Tee B C, Giri G, Xu J, Kim D H, Becerril H A, Stoltenberg R M, Lee T H, Xue G, Mannsfeld S C B, Bao Z N. Nat. Mater., 2013, 12(7): 665.
[78]

Sheng Q Y, Peng B Y, Ji C, Li H Y. Adv. Mater., 2023, 35(52): 2370373.
[79]

Su Y J, Gao X, Liu J G, Xing R B, Han Y C. Phys. Chem. Chem. Phys., 2013, 15(34): 14396.
[80]

Wade J, Steiner F, Niedzialek D, James D T, Jung Y, Yun D J, Bradley D D C, Nelson J, Kim J S. J. Mater. Chem. C, 2014, 2(47): 10110.
[81]

Yamamura A, Watanabe S, Uno M, Mitani M, Mitsui C, Tsurumi J, Isahaya N, Kanaoka Y, Okamoto T, Takeya J. Sci. Adv., 2018, 4(2): eaao5758.
[82]

Lee J C, Lee M, Lee H J, Ahn K, Nam J, Park S. Adv. Mater., 2020, 32(48): 2004864.
[83]

Lim J A, Lee H S, Lee W H, Cho K. Adv. Funct. Mater., 2009, 19(10): 1515.
[84]

Virkar A A, Mannsfeld S, Bao Z N, Stingelin N. Adv. Mater., 2010, 22(34): 3857.
[85]

Don Park Y, Lim J A, Lee H S, Cho K. Mater. Today, 2007, 10(3): 46.
[86]

Kline R J, DeLongchamp D M, Fischer D A, Lin E K, Heeney M, McCulloch I, Toney M F. Appl. Phys. Lett., 2007, 90(6): 062117.
[87]

Lim S C, Kim S H, Lee J H, Kim M K, Kim D J, Zyung T. Synth. Met., 2005, 148(1): 75.
[88]

Deng W, Lei H M, Zhang X J, Sheng F M, Shi J L, Zhang X L, Liu X Y, Grigorian S, Zhang X H, Jie J S. Adv. Mater., 2022, 34(13): 2109818.
[89]

Deegan R D, Bakajin O, Dupont T F, Huber G, Nagel S R, Witten T A. Nature, 1997, 389(6653): 827.
[90]

Jiang L F, Liu J, Lu X Q, Fu L L, Shi Y J, Zhang J, Zhang X, Geng H, Hu Y Y, Dong H L, Jiang L, Yu J S, Hu W P. J. Mater. Chem. C, 2018, 6(10): 2419.
[91]

Corrales T P, Bai M J, del Campo V, Homm P, Ferrari P, Diama A, Wagner C, Taub H, Knorr K, Deutsch M, Retamal M J, Volkmann U G, Huber P. ACS Nano, 2014, 8(10): 9954.
[92]

Janneck R, Vercesi F, Heremans P, Genoe J, Rolin C. Adv. Mater., 2016, 28(36): 8007.
[93]

Li L Q, Gao P, Schuermann K C, Ostendorp S, Wang W C, Du C, Lei Y, Fuchs H, De Cola L, Müllen K, Chi L F. J. Am. Chem. Soc., 2010, 132(26): 8807.
[94]

Sele C W, Kjellander B K C, Niesen B, Thornton M J, van der Putten J B P H Myny K, Wondergem H J, Moser A, Resel R, van Breemen A J J M, van Aerle N, Heremans P, Anthony J E, Gelinck G H. Adv. Mater., 2009, 21(48): 4926.
[95]

Wang L, Liu B B, Yu S D, Yao M G, Liu D D, Hou Y Y, Cui T, Zou G T, Sundqvist B, You H, Zhang D K, Ma D G. Chem. Mater., 2006, 18(17): 4190.
[96]

Sathish M, Miyazawa K, Hill J P, Ariga K. J. Am. Chem. Soc., 2009, 131(18): 6372.
[97]

Zheng S S, Cuong N T, Okada S, Xu T, Shen W Q, Lu X, Tsukagoshi K. Chem. Mater., 2018, 30(20): 7146.
[98]

Wei L, Yao J N, Fu H B. ACS Nano, 2013, 7(9): 7573.
[99]

Jiang H H, Peng B Y, Liu S, Ren J, Yang W T, Lin C C, Wu R H, Chen H Z, Li H Y. J. Mater. Chem. C, 2021, 9(17): 5621.
[100]

Hu H, Larson R G. J. Phys. Chem. B, 2006, 110(14): 7090.
[101]

Sun L, Li Y T, Xie J C, Zhou L Q, Wang P, Xu J B, Shi Y, Wang X R, He D W. Electron, 2023, 1(2): e23.
[102]

Zhou Z W, Zhang Z C, Wu Q S, Ji X D, Wang J L, Zeng X C, Feng S P, Chan P K L. ACS Appl. Mater. Interfaces, 2018, 10(41): 35395.
[103]

Shen X F, Wu X S, Yao X M, Ma C L, Sun L J, Wang Z F, Gao P C, Zeng D, Deng W, Jie J S, Yang F X, Li R J, Hu W P. Proc. Natl. Acad. Sci. U. S. A., 2026, 123(2): e2522805123.
[104]

Peng B Y, Cao K, Lau A H Y, Chen M, Lu Y, Chan P K L. Adv. Mater., 2020, 32(34): 2002281.
[105]

Liu J, Yu Y M, Liu J, Li T, Li C L, Zhang J, Hu W P, Liu Y Q, Jiang L. Adv. Mater., 2022, 34(7): 2107574.
[106]

Ke S F, Shen W, Zhao Y J, Peng B Y, Wang X R, Li H Y. Nano Res., 2025, https://doi.org/10.26599/NR.2025.94908218.
[107]

Chen H L, Li M L, Lu Z Y, Wang X G, Yang J S, Wang Z, Zhang F, Gu C H, Zhang W N, Sun Y J, Sun J L, Zhu W G, Guo X F. Nat. Commun., 2019, 10: 3872.
[108]

Jiang Y, Kellermeier M, Gebauer D, Lu Z H, Rosenberg R, Moise A, Przybylski M, Cölfen H. Nat. Commun., 2017, 8: 15933.
[109]

Bhat V, Callaway C P, Risko C. Chem. Rev., 2023, 123(12): 7498.
[110]

Clancy P. Chem. Mater., 2011, 23(3): 522.
[111]

Alessandri R, Uusitalo J J, de Vries A H, Havenith R W A, Marrink S J. J. Am. Chem. Soc., 2017, 139(10): 3697.
[112]

Gertsen A S, Sørensen M K, Andreasen J W. Phys. Rev. Materials, 2020, 4(7): 075405.
[113]

Lee T, Sanzogni A V, Burn P L, Mark A E. ACS Appl. Mater. Interfaces, 2020, 12(36): 40548.
[114]

Guthrie S M, Gao Y, Stone K H, Xu B X, Giri G. J. Phys. Chem. C, 2021, 125(11): 6269.
[115]

Goose J E, Clancy P. J. Phys. Chem. C, 2007, 111(43): 15653.
[116]

Goose J E, Killampalli A, Clancy P, Engstrom J R. J. Phys. Chem. C, 2009, 113(15): 6068.
[117]

Muccioli L, D’Avino G, Zannoni C. Adv. Mater., 2011, 23(39): 4532.
[118]

Roscioni O M, D’Avino G, Muccioli L, Zannoni C. J. Phys. Chem. Lett., 2018, 9(23): 6900.
[119]

Zeng Y Q, Tao B, Yin Z P. J. Cryst. Growth, 2014, 405: 73.
[120]

Ehrlich G, Hudda F G. J. Chem. Phys., 1966, 44(3): 1039.
[121]

Schwoebel R L, Shipsey E J. J. Appl. Phys., 1966, 37(10): 3682.
[122]

Wu Y, Toccoli T, Koch N, Iacob E, Pallaoro A, Rudolf P, Iannotta S. Phys. Rev. Lett., 2007, 98(7): 076601.
[123]

Zhang X N, Barrena E, Goswami D, de Oteyza D G, Weis C, Dosch H. Phys. Rev. Lett., 2009, 103(13): 136101.
[124]

Zeng Y Q, Tao B, Chen J K, Yin Z P. J. Cryst. Growth, 2015, 429: 35.
[125]

Roscioni O M, Muccioli L, Mityashin A, Cornil J, Zannoni C. J. Phys. Chem. C, 2016, 120(27): 14652.
[126]

Yang Y I, Shao Q, Zhang J, Yang L J, Gao Y Q. J. Chem. Phys., 2019, 151(7): 070902.
[127]

Neha, Tiwari V, Mondal S, Kumari N, Karmakar T. ACS Omega, 2023, 8(1): 127.
[128]

Bjelobrk Z, Piaggi P M, Weber T, Karmakar T, Mazzotti M, Parrinello M. CrystEngComm, 2019, 21(21): 3280.

Funding

Zhejiang Provincial Natural Science Foundation of China(LZ24E030003)

National Natural Science Foundation of China(U23A20592)

National Natural Science Foundation of China(52394274)

PDF (11048KB)

266

Accesses

0

Citation

Detail
Sections
Recommended

/

〈 〉

Address of Editorial Office: No. 33 Beisihuan Xilu, Zhongguancun , Beijing 100190, China
Tel: +86 10 8262 7757
Fax: +86 10 8262 7757
E-mail: progchem@mail.las.ac.cn
http://www.progchem.ac.cn
    • 　
    Copyright © Editorial Office of Progress in Chemistry