Polymer-dispersed liquid crystal with low driving voltage utilising reaction selectivity of two-stage polymerisation

ABSTRACT An optimised approach is presented in this study for the fabrication of a polymer-dispersed liquid crystal (PDLC) device with a low driving voltage and high contrast ratio (CR). This method involves a two-step process utilising Michael addition and radical polymerisation, employing a reaction-selective orthogonal photoinitiated system. The impact of incorporating low surface energy monomers on PDLC performance is thoroughly investigated, and the driving voltage is further reduced through selective reaction-induced interface modification. The polymer morphology can be finely tuned by adjusting the thiol content, leading to various PDLC properties. This paper introduces a novel surface modification technique within the polymer-liquid crystal system, presenting a fresh perspective on regulating PDLC performance. By choosing monomer structures, the sequence of reactions can be precisely controlled, allowing for the introduction of structural units with distinct functionalities at the polymer interface. This approach facilitates interfacial functionalization and enables the design of multi-layer structures for synthetic polymers, realising innovative functionalities. GRAPHICAL ABSTRACT


Introduction
Polymer-dispersed liquid crystal (PDLC) devices have gained significant attention as crucial liquid crystal-based switches for indoor lighting/privacy protection and energyefficient smart windows, contributing to the advancement of flexible smart optical films [1].However, significant polymer content and the strong anchoring effect on liquid crystal molecules in PDLC devices [2] typically necessitate higher operating voltages, surpassing safe voltage limits for human interaction and increased energy consumption.
Previous studies have commonly focused on reducing the drive voltage of PDLCs by increasing the polymer mesh size.However, this approach often resulted in an average mesh size that was too large, leading to a decrease in contrast ratio (CR) and compromising the optoelectronic performance of the device [3,4].Therefore, our approach aims to reduce the anchoring energy of PDLCs to minimise the drive voltage while ensuring a high CR.Among the various materials used for low surface energy modification, silicon, and CONTACT Hongbo Lu bozhilu@hfut.edu.cn;Miao Xu xumiao0711@hfut.eduSupplemental data for this article can be accessed online at https://doi.org/10.1080/02678292.2024.2324464.
fluorine-containing materials have emerged as the two most extensively utilised options [5].Both silicon and fluorine-containing materials have favourable characteristics such as stability, heat resistance, high hydrophobicity, and excellent transparency.
In comparison, fluorine-containing materials exhibit lower surface energy.However, they tend to render the substrate material more brittle, which hinders the fabrication of flexible devices, and their recycling can harm the environment.On the other hand, silicon materials offer advantages such as lower cost and reduced environmental impact and have found extensive research applications in the realm of interface modification [6,7].The incorporation of low surface energy monomers leads to a reduction in the drive voltage [8].It occurs as these monomers become uniformly dispersed within the polymer matrix during the polymerisation process, consequently lowering the overall anchoring energy.However, the key to modulating the driving voltage is the magnitude of the anchoring energy at the polymer-liquid crystal interface [9].Hence, to further reduce the driving voltage, the reaction sequence is controlled where the low surface energy monomers are incorporated during the later stages of the polymerisation process, enriching their presence at the interface between the liquid crystal and polymer domains.
To attain precise control over the reaction sequence, the regulation of the polymerisation rate is implemented through the utilisation of the Michael addition reaction, known for its nucleophilic addition characteristics [10].Thiols exhibit distinct reaction selectivity towards specific acrylate monomers with different structural variations [11,12].For instance, their reactivity towards methacrylate and acrylate structures is notably influenced by the positive induction of the methyl group to the β-carbon, stabilising the electrophilic properties of the carbon atoms and consequently reducing their reactivity [13].During free radical polymerisation, the presence of asymmetric methyl group substitutions impedes the internal rotation of the molecular chain [14].As a consequence, the polymerisation rate of methyl methacrylate monomers is significantly lower compared to acrylates [15].In previous studies, a twostep initiation of polymerisation using heat and light has been used to generate unique polymer interfacial structures, which provides a new idea for regulating the performance of PDLC [16][17][18][19].The two-step preparation of PDLC utilising Michael addition and free radical polymerisation reactions enables the formation of a dual network structure, thereby preserving desirable mechanical properties [20,21].Additionally, this approach leverages the selectivity of the reactions to achieve interfacial functionalization, enhancing the overall electro-optical performance of the PDLC system.
In this study, the mutually non-interfering photoinitiator system comprising the photogenerated basecatalyst NPPOC-TMG and the photoinitiator I2959 was selected (Figure 1).The monomers PETMP, HDDA, and TTPMA were employed as polymerisation monomers to prepare the PDLC using a two-step photoinitiation method.Upon exposure to 405 nm light, the decomposition of NPPOC-TMG resulted in the generation of the base catalyst TMG, facilitating the catalysis of the thiol-ene Michael addition reaction [22].The increased electrophilicity of the double bond in the methyl group led to a significant decrease in the reactivity of TTPMA during the Michael addition reaction, resulting in minimal participation in the reaction.Consequently, the polymerisation process predominantly formed an oligomer composed of HDDA and PETMP.Subsequently, under 365 nm light, I2959 initiated radical polymerisation.Due to the differential reactivity caused by the presence of the methyl group, HDDA underwent a reaction before TTPMA, leading to an enrichment of TTPMA at the polymer interface.This enrichment resulted in a reduction in interfacial anchoring energy, serving the purpose of decreasing the driving voltage.
In contrast to conventional approaches for reducing the driving voltage of PDLCs, this paper introduces a novel method that focuses on surface modification to regulate the polymer liquid crystal (polymer-LC) interface.This approach effectively lowers the interfacial anchoring energy while maintaining a nearly constant domain size.Consequently, a PDLC device with a low driving voltage and high CR can be realised.The proposed regulation method holds significant potential in developing high-performance PDLC optical valves, offering promising prospects.

Materials
The experimental materials comprised the following components: a positive nematic LC E7 (Δn = 0.

Sample preparation
The monomer, catalyst, initiator, and LC mixture were vigorously stirred until a homogeneous system was achieved.Subsequently, the resulting system was filled into cells consisting of two parallel glass substrates with indium-tin-oxide (ITO) transparent electrodes.The cell thickness was carefully controlled using 9 μm spacers.The first stage of the Michael addition reaction was initiated by irradiating the mixture with blue light (~405 nm central wavelength, ~10 mW/cm 2 intensity), followed by UV light irradiation (~365 nm centre wavelength, ~10 mW/cm 2 intensity) to polymerise the monomers.Both irradiation processes were conducted at 35°C.The compositions of the mixtures are presented in Table 1.The experimental samples were divided into four groups.Group A included samples A1-A6, where A1, A3, and A5 were irradiated with both blue and UV light, while A2, A4, and A6 were only subjected to UV light irradiation.In group B, samples B1-B5 varied the molar ratio of HDDA to PETMP in the mixture (1:1, 2:1, 3:1, 4:1, 5:1).The ratios of TTPMA in the mixture were varied from C1-C7 in group C (0 wt%, 5 wt%, 10 wt%, 13 wt%, 15 wt%, 17 wt%, 20 wt%).Group D samples, D1-D5, involved varying the duration of blue light irradiation during the Michael addition reaction (0 min, 5 min, 10 min, 15 min, 20 min), respectively.

Measurements
In order to monitor the curing reaction of acrylates (methacrylates) and thiols, infrared spectra of samples with varying curing times were recorded using an attenuated total reflectance Fourier-transform infrared spectroscopy (ATR-FTIR) instrument (Nicolet, USA).Initially, a homogeneous liquid mixture of sample was uniformly applied to a KBr wafer.Subsequently, the mixture was subjected to light curing, and infrared spectra were collected at different light exposure times.
Scanning electron microscopy (SEM) using a ZEISS Gemini 500 SEM instrument (Germany) was utilised to examine the polymer morphology.To prepare the specimens for SEM analysis, the cells were frozen in liquid nitrogen, fragmented into smaller pieces, and then immersed in hexane to extract the LC molecules.Subsequently, the samples were dried in an oven.Thin layers of gold were sputtered onto the specimens to enhance conductivity, and the morphology of the polymer network was observed.
The electro-optical characteristics of the fabricated LC films were evaluated using a liquid crystal parameter tester (LCT-5046E, North LCD Engineering R&D Center, China).Square-wave modulated electric fields with a frequency of 100.0 Hz were applied as test fields.
The distance between the detector and the samples was approximately 10.0 cm, ensuring accurate measurements.
The peel strength of PDLC films was determined using tensile material testing machines (SH-III-50 N, Nscing Es, China).The sample films were prepared to have dimensions of 10 cm in length and 5 cm in width, ensuring uniformity in testing.

Results and discussion
In this study, the reaction order between the selected monomers HDDA, TTPMA, TTPA, and PETMP was initially confirmed through real-time infrared analysis.Details of the IR experiment are depicted in Figure S1 and S2.Similar results to previous studies were obtained [12,13,23], in mixed monomers HDDA, TTPMA and PETMP, monomer TTPMA was the slowest to polymerise in Michael addition versus free radical polymerisation, attributed to due to the structure of methyl group.
The high driving voltage in PDLC is primarily attributed to the strong anchoring forces.To reduce the driving voltage, it is essential to decrease the anchoring forces through the use of low surface energy materials at the domain interface.In order to investigate the surface energy at the interface, a surface initiation method was employed to prepare polymer films [24].During the polymerisation process, the migration of monomers leads to variations in surface energy on the polymer film surface.The experimental procedure is illustrated in Figure S3.The surface energy of the polymer film was determined by measuring the contact angle, allowing for an examination of the impact of different monomers and polymerisation methods on the surface energy.The composition of the mixture and the liquid employed for spin coating are detailed in Table 2. Samples O1, O3, and O5 were subjected to blue and UV light irradiation, while samples O2, O4, and O6 were solely exposed to UV light.The contact angle of each sample is shown in Table S2.The calculation formula is as follows [25,26].
Here γ S , γ L, and γ SL denote the surface tension of solid, liquid, and the interfacial tension between liquid and solid, respectively.W A is the anchoring energy between polymer and liquid crystal, and θ is the contact angle.As depicted in Figure 3, Samples O1-O4 exhibit lower surface energy compared to samples O5-O6, attributed to the inclusion of siloxane-based monomers.The pronounced reduction in surface energy of sample O2 compared to samples O3-O4 arises from the disparity in reactivity during free radical polymerisation, enabling TTPMA to engage in the later stages of polymerisation and resulting in a noticeable enrichment effect.The lower surface energy of sample O1 relative to sample O2 is attributable to the selective nature of the Michael addition reaction, allowing higher participation of TTPMA in the polymerisation during the subsequent step of the reaction, consequently leading to a more significant enrichment at the polymer interface.
In order to investigate the impact of doped siloxane monomers with selectivity reaction on the photovoltaic properties of PDLC, samples were prepared using both two-step and single-step methods with siloxane monomers TTPMA (with selectivity reaction), TTPA, and monomer M1.As illustrated in Figure 4, consistent with the results from the surface energy experiments, samples A1-A4 exhibited lower driving voltages compared to A5-A6 due to the incorporation of siloxanebased monomers, which resulted in reduced surface energy.Among the one-step method samples, A2 showed lower driving voltages than A4 due to the difference in reactivity of siloxane-based monomers, specifically TTPMA, in free radical polymerisation, allowing for greater enrichment at the polymer interface.Sample A1 exhibited the lowest driving voltage in the two-step method due to the combined effect of the Michael addition reaction and the free radical reaction, resulting in a higher enrichment of TTPMA at the interface.The shorter response time and longer restoration time of sample A1 were also attributed to the low anchoring energy interface, which facilitated easier rotation of the liquid crystal molecules under the influence of the electric field and slower restoration after power withdrawal [27].
The content of thiol monomers affects the molecular weight of oligomers formed by Michael's addition reaction, which in turn affects the system compatibility [21,28,29].By adjusting the ratio of thiol to acrylate, the compatibility of oligomers can be controlled, thereby influencing the phase separation process and regulating the network morphology.As depicted in Figure 6, when the thiol ratio decreases from B1 to B5, the network morphology transitions from spherical to fibrous structures with embedded spheres, ultimately forming a dense mesh-like structure.It can be attributed to the limited compatibility of thiol-alkene oligomers  generated through the Michael addition reaction [30], which tend to precipitate into spherical domains during the phase separation process in free radical polymerisation.As the percentage of thiol decreases, the content of these poorly compatible oligomers diminishes, thereby reducing their influence on subsequent free radical polymerisation and leading to the gradual transformation of the polymer network into a more conventional mesh-like shape.By combining the information from Figures 5 and 6, we can analyse the relationship between the network structure and the performance of PDLC samples [31,32].For instance, let us consider sample B1 as an example.The spherical structure in B1 exhibits a small specific surface area, resulting in weak scattering and consequently leading to low driving voltage and CR.On the other hand, sample B5, which features a mesh structure with a large specific surface area, exhibits strong scattering due to multiple domains.As a result, B5 shows high driving voltage and CR.In sample B3, the fibre structure with embedded spheres also possesses a small specific surface area, while the multi-layer fibre interface contributes to strong scattering.As a result, sample B1 also exhibits low driving voltage and high CR.
The driving voltage of PDLC is influenced by the content of low surface energy monomer.As depicted in Figure 7, an increase in TTPMA content decreases the driving voltage.Furthermore, Figure 8 demonstrates that adding TTPMA as a single functional monomer reduces polymer cross-link density, causing the mesh size to enlarge until it forms a discontinuous spherical shape gradually.Simultaneously, the inclusion of low surface energy monomers reduces the anchoring energy.The combined effect of polymer domains and reduced anchoring energy contributes to the decreased driving voltage.Sample C7 exemplifies the formation of a discontinuous spherical structure resulting from changes in system compatibility due to the excessive addition of mono-functional monomers, which consequently leads to a decrease in CR.
The concentration of the alkali catalyst influences the reaction rate of the Michael addition process.Prolonged exposure to light leads to increased catalyst decomposition, resulting in a higher reaction rate.Therefore, by adjusting the light exposure time during the initial step, the system's alkalinity can be controlled to manipulate the rate of Michael addition polymerisation and, subsequently, the network morphology.The SEM images presented in Figures 9 and 10 demonstrate that the size of the polymer domains initially increases and then decreases as the light exposure time varies from sample D1-D5.Correspondingly, the driving voltage of the device decreases and then increases.Analysing the polymerisation rate and system compatibility, we observe that the polymer's molecular weight is low during the early polymerisation stages, as polymerisation progresses, the concentration of olefin monomers decreases, resulting in a reduction in the rate of free radical polymerisation.This delay in phase separation leads to the gradual enlargement of the domain size.However, as the reaction proceeds to a higher polymerisation degree, compatibility deteriorates and oligomer precipitation leads to an increase in system viscosity up to a certain value, which hinders the flow and growth of the liquid crystal, thereby reducing the size of the liquid crystal domains [33].This observation is further supported by viscosity measurements of the system at different 405 nm light exposure times using a rotational viscometer (Figure S4).
The flexible PDLC film was fabricated using a 7 μm spacer, allowing it to retain excellent electro-optical properties, as depicted in Figure 11(a).However, the current drawback lies in the insufficient mechanical properties, which pose a significant limitation for  flexible PDLC devices.The peel strength (σ γ ) was evaluated by measuring the adhesion force of the sample using the following equation.
Where F is the average value of peel force, and B is the width of the specimen strip.The peel strength test setup is illustrated in Figure 11(b).Initially, a glass substrate is secured, and a 7-μm diameter spacer is introduced to control the thickness of the liquid crystal layer.The mixture is then applied onto the glass substrate and covered with a flexible PET film.Following the curing process, the upper PET film is peeled off from one end at a constant speed.The force-displacement curve obtained during the  peeling process is used to calculate the average peel force (F), and subsequently, the peel strength is determined.The peel strength results are presented in Figure 11(c).It is observed that A1 exhibits higher adhesion compared to A2, indicating that the substrate peeling strength can be improved through the two-step method.Substrate modification is performed based on the two-step method to fulfil application requirements and expand the range of applications.The test results for sample A1-Modified demonstrate that the adhesion is nearly twice as strong as the initial condition, thereby meeting the application conditions of flexible devices.The substrate modification experiment is illustrated in Figure S5.

Figure 1 .
Figure 1.(Colour online) Schematic representation of the two-stage polymerization process.

Figure 2 .
Figure 2. Chemical structures of the utilized materials.

Figure 3 .
Figure 3. (Colour online) Surface energy of polymer films prepared using different monomers and polymerization methods.

Figure 4 .
Figure 4. (Colour online) (a) Transmittance as a function of applied electric field (100 hz).(b) The threshold voltage (V th ) and saturation voltage (V sat ) of group a specimens.(c) Contrast ratio (CR) of group a specimens.(d) Response time τ on and restore time τ off of group a specimens.

Figure 5 .
Figure 5. (Colour online) (a) Transmittance as a function of applied electric field (100 hz).(b) The threshold voltage (V th ) and saturation voltage (V sat ) of group B specimens.(c) Contrast ratio (CR) of group B specimens.(d) Response time τ on and restore time τ off of group B specimens.

Figure 7 .
Figure 7. (Colour online) (a) Transmittance as a function of applied electric field (100 Hz).(b) The threshold voltage (V th ) and saturation voltage (V sat ) of group C specimens.(c) Contrast ratio (CR) of group C specimens.(d) Response time τ on and restore time τ off of group C specimens.

Figure 9 .
Figure 9. (Colour online) (a) Transmittance as a function of applied electric field (100 hz).(b) The threshold voltage (V th ) and saturation voltage (V sat ) of group D specimens.(c) Contrast ratio (CR) of group D specimens (d) response time τ on and restore time τ off of group D specimens.

Table 1 .
Samples and their corresponding compositions.

Table 2 .
Composition of samples and their corresponding mixtures.