A 3D printing UV-curing resin that enables continuous color change of elastomers

Summary

This study introduces a UV-curing resin leveraging the Christiansen effect to achieve dynamic color modulation in phase-separated elastomers (PSE) through solvent composition and temperature control. By matching refractive indices between solvent (dimethyl phthalate/tributyl citrate) and polymer phases, PSE selectively transmits-specific wavelengths, enabling continuous color shifts from red to purple. The elastomers exhibit robust mechanical properties (>600% elongation, >300 kPa modulus) and function across a broad temperature range (−25°C–130°C). Compatible with 3D printing, the resin enables real-time solvent adjustments during fabrication, supporting spatially continuous color transitions without parameter modifications. This approach advances applications in adaptive optical devices, temperature-responsive displays, and multi-material 3D printing, offering a dye-free strategy for structural coloration in smart materials.

Graphical abstract

Highlights

• •Solvent-tuned Christiansen effect enables dynamic coloration without dyes
• •Elastomers exhibit >600% elongation, >300 kPa modulus, −25°C–130°C stability
• •3D-printable with real-time solvent control for spatial color gradients
• •Dye-free structural coloration advances adaptive optics and multi-material printing

Chemistry; Materials science; Physics

Introduction

Light-curing resins have been increasingly used in batteries,^1,2,3 inks,^4,5,6,7 biomedical devices,^8,9,10,11,12 3D printing,^{13,14,15,16,17} and other fields due to the advantages of easy availability of raw materials and ease of use. Commercially available light-curing resins are usually transparent, with dyes added to the resin to impart a rich color expression. More importantly, the dyes absorb UV light to improve curing accuracy.¹⁸ However, commercially available dyes are often toxic, while bio-based dyes with lower toxicity generally exhibit poor stability, are prone to fading, and come with high production costs.^19,20,21 It makes alternative methods to impart colors an attractive filed of research. For example, photonic crystals are made with periodic arrangements of media having different refractive index, enabling precise screening and presentation of specific light bands,^22,23,24,25 while birefringence is harnessed in the photoelastic effect to create colors through interfering polarized light.^{26,27,28,29,30,31} Nevertheless, these methods have their limitations, with the preparation of photonic crystals or ordered micro-nano arrays generally complex and time consuming. In addition, the interfering colors generated by the photoelastic effect have to be observed through a polarizer, which limits its practicality.

An alternative method to impart colors to a material is based upon the Christiansen effect. In particular, modulation of the refractive indices of the solvent and of the polymer phases enables the differentiation of different wavelengths of light, making specific colors visible to the naked eye.^32,33,34 This method makes preparation of materials with different colors very simple and effective, their optical characterization does not need additional instrumentation, in turn greatly improving practical implementation and expanding applications in the field of UV-curing resins. In this context, our team has developed a new light-curing resin and successfully prepared bicontinuous phase-separated elastomers (PSE) through the unique agglomeration behavior of polymers in poor solvents.^35,36,37,38 In our team’s previous work, hydroxyethyl methacrylate (HEMA) was used as the monomer, and 4′-pentyl-4-cyanobiphenyl (5CB) and tributyl citrate (TBC) were used as solvents to prepare dielectric gels capable of achieving tunable optical properties.³⁹ However, 5CB is an expensive liquid crystal only necessary if the product is to be used as a dielectric actuator. In addition, the low activity of HEMA makes it difficult to be adapted to light-curing 3D printers, and it is difficult to cure properly even with the addition of photoinitiators with very high initiation capacity.⁴⁰

In the present work, we optimized the formulation of the photosensitive resin by including hydroxyethyl acrylate (HEA) as the monomer, and replacing 5CB with dimethyl phthalate (DMP), leading to the preparation of a bicontinuous PSE that achieved similar rainbow-colored transition as the previous material, and with selective transmittance of monochromatic light of different wavelengths of around 90%. Coincidentally, the mechanical properties of the elastomer obtained are also substantially improved compared to the previous work, and the resin ink is also able to be adapted to UV-curing on conventional 3D printers. A systematic comparison with our team’s previous work can be found in the Supporting Information. Continuous color changes across the entire visible spectrum (from red to purple) were obtained in 3D printed objects by modulating the composition of the solvent, from pure DMP to 25% TBC in DMP. Interestingly, 3D printing parameters remained unaffected despite the change in solvent composition, enabling continuous printing of multi-colored objects.

Results and discussions

Construction and characterization of phase-separation systems

The conformational change of a polymer in a solvent is significantly dependent on the compatibility of the polymer chain segments and the solvent. If in a good solvent, the polymer chain segments are able to stretch sufficiently in the good solvent environment to form a homogeneous structure, a transparent continuous phase similar to a hydrogel, allowing unobstructed passage of light. However, as the solvent gradually changes to a poor solvent for the polymer, the polymer chain segments begin to undergo a conformational transition from a stretched state to a curled state, resulting in the formation of a distinct two-phase separation between the solvent and the polymer. This phase separation affects the optical and mechanical properties. The Christensen effect is the selective transmission of monochromatic light that occurs when the refractive indices of the two phases are the same. When the solvent and the polymer are very poorly compatible, the polymer chain segments will further aggregate to form micrometer-sized particles, at which time the transmittance of the polymer to light is further reduced. At the same time, the modulus of the material increases significantly and the elongation at break decreases sharply. The ability of the polymer network to conserve solvents decreases, making it susceptible to leakage behavior. Therefore, a phase-separated structure intermediate between this conformation and a homogeneous structure, i.e., a bicontinuous polymer-solvent structure, maintains the elasticity of the polymer without loss and ensures the color selectivity due to phase separation. This property has potential applications in the fields of optics, display, and sensing.

As shown in Figure 1A, polymers polymerized without solvents exhibit transparent and homogeneous qualities in which light travels in a straight line without scattering. The addition of poor solvents causes bending of the polymer chain segments, resulting in dispersion of natural light within the material. And by adjusting the refractive index of the solvent phase to be the same as that of the polymer phase, the phase-separated elastomer will be selectively transmissive for monochromatic light at particular wavelengths. Further, with the increase of the poor solvent, the proportion of the two phases in the system tends to be balanced, and this phenomenon enhances the color selective transmittance, as shown in Figure 1B. To achieve the aforementioned properties, HEA is selected as the monomer and TBC and DMP as the poor solvents in this study (Figure 1C). Both solvents are poor solvents for poly(HEA) (PHEA), but they have a moderate degree of phase separation, which avoids excessive agglomeration of the polymer chains to form particles. The final prepared phase-separated elastomer (PSE) spontaneously forms the unique structure shown in Figure 1D. It contains micrometer-scale folds inside, but still maintains an overall continuous phase state. In addition, the solvent itself acts as a free liquid and constitutes another continuous phase. The specific experimental procedures are detailed in the Supporting Information. Figure S1 demonstrates SEM photographs of PSE with different solvent ratios, with the sample prepared with pure DMP on the top and the sample prepared with a 4:1 mixture of DMP and TBC on the bottom. As can be seen from these figures, the PSE prepared by both poor solvents showed similar structural features, confirming the bicontinuous phase state of the PSE system. To verify this conclusion on a more microscopic scale, we performed atomic force microscopy (AFM) test on the PSE. Figure S2 is small-angle X-ray scattering (SAXS) test. It shows a smooth curve, which proves that PSE did not form a granular structure at the microscale and nanoscale. And Figures S3 and S4 show the elevation and phase diagrams taken by AFM, respectively. And the morphology coincides with the morphology revealed by SEM photographs, which further confirms that the polymer chain segments have a folded structure at the microscopic level, and the height of these folds does not exceed 1 μm.

Figure 1.

Polymer synthesis, phase separation, and structural characterization of PSEs

(A) Schematic and physical photograph of polymerization of HEA into a transparent homogeneous PHEA elastomer.

(B) Schematic and physical photograph of the polymerization of PSE after addition of poor solvents to HEA.

(C) Chemical formulas of HEA and solvents used in this work.

(D) Scanning electron micrograph of PSE.

Optical and mechanical characterization of solvent ratio-dependent PSEs

As shown in Figure 2A, the refractive index of HEA was accurately determined to be 1.4450 using an Abbe refractometer at a constant temperature of 29°C, whereas the refractive index of PHEA formed after the polymerization reaction was significantly increased to 1.4965. This change is attributed to the breaking of the C=C double bond and the formation of the C-C single bond during the polymerization process, which is accompanied by a volume shrinkage effect, thus contributing to the increase in refractive index. The refractive index of TBC is 1.4450 and the refractive index of DMP is 1.5168, the refractive index of PHEA is exactly in between. Therefore, these two solvents form an ideal refractive index adjustment interval. By finely tuning the solvent ratios, we were able to construct specific structures with matching refractive indices of the two phases. However, it is worth noting that the refractive index of a mixed-solvent system is not a simple weighted average of the refractive indices of two pure solvents to match PHEA. This is because the poor solvents cause PHEA to curl, which can lead to still slight changes in the refractive index of the polymer phase. Based on these conclusions, we designed and prepared a series of PSE with a monomer-to-solvent mass ratio of 7:3, where the content of TBC was differentiated as a key variable. For example, PSE-0 indicates that DMP was used exclusively as the solvent, and its refractive index was 1.5048 (as shown in Figure 2A). PES-0.6 indicates that the 30% solvent mixture consisted of 6% TBC and 24% DMP, and the refractive index of this sample was adjusted to 1.4990. Due to the presence of the Christiansen effect, the two elastomers exhibited red and violet structural colors, respectively, under natural light which precisely calibrates the refractive index range of the red to violet transition in the rainbow color scheme (1.4990–1.5048). This range is slightly larger than the PHEA refractive index of 1.4965, which also proves that the microscopic state of the polymer chain segments is not completely stretched, but rather curled, which is why the refractive index of the polymer phase in PSE is slightly larger than that of the PHEA elastomer.

Figure 2.

Optical and mechanical characterization of solvent ratio-dependent PSEs: refractive indices, tensile behavior, modulus, and transmittance performance

(A) Refractive indices of HEA, PHEA, PSE-0.6, and PSE-0.

(B) Tensile photographs of PSE.

(C) Stress-strain curves of PSE for a range of different solvent ratios.

(D) Young’s modulus of PSE for a range of different solvent ratios.

(D) Comparison photos before and after PSE stretching.

(E) Transmittance of PSE-(-0.2), −0.1, 0, 0.1, and 0.2 in the visible range.

(F) Transmittance of PSE-0.2, 0.3, 0.4, 0.5, and 0.6 in the visible range.

Figure 2B visualizes the morphology of PSE-0.6 before and after stretching, clearly demonstrating the excellent deformation capability of PSE. It can easily stretch to several times its original length, demonstrating its excellent elasticity. Further, a series of PSE samples were subjected to stress-strain tests. Figure 2C presents their stress-strain curves. Despite the fact that no cross-linking agent was introduced into the precursor solution of PSE, these samples exhibited far more elastic characteristics than the viscoelastic behavior common to conventional linear polymers. Regardless of the solvent ratio adjustment, PSE showed excellent elastic response, with elongation at break generally exceeding 600% and reaching 850% for PSE-0.6. In addition, PSE has good strength with tensile strengths up to 0.5 MPa (PSE-0.2). The excellent mechanical properties can be attributed to the undesirable solvent-induced phase separation phenomenon, in which the chain segments within the polymer phase are compactly arranged, which reduces the distance between polymer chains and enhances the effect of van der Waals forces. Van der Waals forces essentially act as physical cross-linking, which significantly enhances the elasticity as well as the mechanical strength of the PSEs.

We prepared a homogeneous transparent gel system (named as PSE-PC) by blending the highly polar solvent propylene carbonate (PC) with HEA. Mechanical property tests revealed that PSE-PC exhibited a fracture elongation of only 250% along with significantly lower tensile strength compared to the PSE-0 sample (Figure S5). This mechanical performance disparity originates from distinct structural characteristics: The PSE-PC system lacks a physical crosslinking network, resulting in reduced material strength, while its polymer chains remain fully extended in the homogeneous system, leading to direct chemical bond breakage upon stress application. In contrast, the PSE-0 system developed in this work demonstrates coiled chain conformations due to poor solvent effects. During stretching, these chains first undergo extension and physical interaction disruption before eventual chemical bond fracture, thereby achieving superior fracture elongation (>600%). Moreover, we conducted a cyclic stretching test on the PSE-0 sample at a strain level of 100%, repeated for 100 cycles. The stress-strain curves indicate that as the number of cycles increased, the sample progressively exhibited creep behavior, with its tensile strength decreasing to 70% of the initial value after 100 cycles (Figure S6). Although the fatigue performance is not yet outstanding, it is sufficient to meet the requirements of routine applications. Additionally, none of the samples in this study were treated with any chemical crosslinking agents. The fact that such mechanical properties were achieved solely through physical interactions is already quite satisfactory. Finally, we conducted tensile tests on PSE-0 at 40°C and 80°C to evaluate its stress-strain behavior (Figure S7). The experimental results demonstrate that although PSE retains a certain degree of mechanical strength and flexibility at elevated temperatures, its overall mechanical performance exhibits a significant decline. Specifically, the elongation at break decreases to 106% at 80°C. This degradation is attributed to the disruption of the material’s physical crosslinking network under high-temperature conditions. The measured Young’s modulus at 80°C drops below 200 kPa, representing an approximately 33% reduction compared to its value at room temperature. These findings indicate that while PSE remains functional in extreme thermal environments, its mechanical properties inevitably deteriorate, leading to a corresponding reduction in service life.

Figure 2D shows the Young’s modulus data of these PSEs. The Young’s modulus of all the samples is stable above 300 kPa with little difference among them. This result indicates that despite the variation in the proportion of poor solvents, it has little effect on the overall mechanical properties of PSEs, and PSE samples of different colors (i.e., different solvent ratios) exhibit similar mechanical strength. It should be particularly noted that although the percentage of DMP in the solvent composition of the PSE-(-0.2) sample mentioned in Figures 2C and 2D is 28%, the remaining 2% is 5CB, which can further enhance the refractive index. It is intended to be comparatively analyzed with the PSE system in the red-light region in this study. Similarly, the depiction of PSE-(-0.2) and PSE-(-0.1) in Figure 2E follows this solvent ration.

Figures 2E and 2F show in detail the transmittance curves of PSE in the visible range, which are arranged according to the gradual decrease of the refractive index of the solvent phase (from PSE-(-0.2) to PSE-0.6, the thickness is 2 mm). In order to present the change rule clearly, we place the transmittance curves in Figures 2E and 2F, respectively, so as to avoid the observation difficulty caused by too many curves in the same figure. Figure 2E shows the transmittance curves of PSE-(-0.2), −0.1, 0, 0.1, and 0.2 for five elastomers. As can be seen from the figure, the transmittance of PSE-(-0.2) is lower than 10% in the full visible wavelength band, which macroscopically shows a white opaque state similar to conventional phase-separated materials. With the gradual decrease of the refractive index of solvent phase, its transmittance starts to increase in the high wavelength region, and macroscopically shows a trend of gradual enhancement of red light. And Figure 2F shows the transmittance curves of five elastomers of PSE-0.2, 0.3, 0.4, 0.5, and 0.6. When the refractive index of solvent phase is further reduced, the visible light transmittance of PSE at high wavelengths starts to decrease, while the transmittance at low wavelengths gradually increases. Specifically, the peak transmittance wavelengths of PSE-0.2, 0.3, 0.4, and 0.5 are located at 623 nm, 504 nm, 446 nm, and 405 nm, and the corresponding transmittances reach 89.6%, 90.1%, 90.3%, and 87.8%, respectively. Combined with Figures 2E and 2F, it can be found that the wavelengths of PSE with the maximum transmittance are blue-shifted with the decrease of the refractive index of solvent phase, which is macroscopically manifested in the gradual transition of the color of PSE from red to violet and the formation of a rainbow-like color change. In addition, the peak transmittance of different PSEs is stabilized at about 90%. Compared with the previous work of our team, the optical performance has realized significant improvement. The transmittance of PSE-0.6 in the visible wavelength is increasing with the decrease of wavelength, and has not reached the peak value at 400 nm. Based on the curve trend and the aforementioned law, we can infer that the peak transmittance should be located in the ultraviolet region.

Thermal analysis and thermoresponsive optical behavior of PSEs

By finely tuning the solvent ratio, we are able to modulate the refractive index of the solvent phase, which in turn effectively changes the color properties of PSE. In addition, temperature, another key parameter, induces a similar color shift effect due to the physical phenomenon that the refractive index of a liquid decreases with increasing temperature. Therefore, the intrinsic link between PSE color and temperature has been explored in depth. The first step was to clarify the temperature tolerance range of PSE in practical applications. As shown in Figure 3A, we tested the DSC curves of PSE-0 and PSE-0.6, and found that they have two significant temperature points of heat flow change: −25°C and −82°C. Since the melting points of DMP and TBC are 2°C and −20°C, respectively, which are much closer to −25°C, it can be judged that −25°C marks the critical temperature for the beginning of solvent solidification in PSE. And −82°C corresponds to the characteristic temperature of the polymer chain from the elastic state to the glassy state. Therefore, we concluded that the PSE use temperature should not be lower than −25°C. In order to define the maximum use temperature of PSE, thermogravimetric analysis (TGA) was performed. Figure 3B shows the thermogravimetric curve of PSE-0.6. The curve reveals two significant mass loss temperatures: 130°C and 340°C. The boiling points of DMP and TBC are 280°C and 170°C, respectively, which are much lower than 340°C. Therefore, we judged that 130°C is the temperature at which the solvents in PSE-0.6 begin to volatilize significantly, and 340°C is the temperature at which decomposition of the polymer skeleton begins to occur. In summary, the normal use temperature range of PSE is −25°C–130°C. This wide temperature range is sufficient to cover the temperature conditions in daily life and most industrial applications, which demonstrates the broad applicability and high utility of this light-curing resin material.

Figure 3.

Thermal analysis and thermoresponsive optical behavior of PSEs with composition- and temperature-dependent variations

(A) Differential DSC curves of PSE-0 and PSE-0.6.

(B) Thermogravimetric curve of PSE-0.6.

(C) Visible light transmittance of PSE-0 at different temperatures.

(D) Visible light transmittance of PSE-0.3 at different temperatures.

(E) Color comparison of PSE-0, PSE-0.3, and PSE-0.6 with different solid content at 25°C and 60°C.

Based on the aforementioned study, we designed and implemented a series of variable-temperature visible light transmittance tests to investigate the optical properties of PSE materials at different temperatures. The specific test temperatures were 0°C, 25°C, 50°C, and 75°C. Figure 3C visualizes the transmittance trend of PSE-0 with increasing temperature. Especially in the red-light band (about 800 nm), the transmittance increases significantly from 17% at 0°C to 53% at 75°C. This change is manifested macroscopically by the gradual transformation of the color of PSE-0 from a nearly opaque state to a sharp red color, which fully verifies the significant effect of temperature increase on the optical properties of PSE. Figure 3D shows the transmittance curve of PSE-0.3. At 0°C, the maximum transmittance of PSE-0.3 is 59% at 800 nm, and as the temperature increases up to 50°C, the maximum transmittance is 89% at 800 nm. However, when the temperature is further increased to 75°C, the transmittance of the red band starts to decrease and the peak value shifts to low wavelengths. 81% of the transmittance at 800 nm at 75°C and the peak value is 94% at 558 nm, which indicates that the color of the PSE-0.3 is gradually transitioning from red to green, showing the fineness of the temperature regulation of the PSE color. In summary, increasing the temperature can blue shift the color of PSE, showing consistency in effect with increasing the TBC content. Figure 3E shows the color comparison between PSE samples with different solid contents at room temperature and heated up to 60°C in the form of real photographs. Each PSE exhibits a significant color change, further reinforcing the experimental finding of a blue shift in color due to elevated temperature. In addition, the saturation of the colors increased with increasing solvent content, confirming the findings in Figure 1B. However, it is interesting to note that there is a tendency for the color of PSE to redshift as the solid content decreases. At first, we thought it was due to the fact that the density of the poor solvent is slightly higher than that of the monomer HEA. The same mass of PSE will have a larger volume if the solid content in the system is higher. This causes the particle arrangement of the polymer phase to become looser and therefore the refractive index is slightly reduced. We have supplemented the test with scanning electron microscopy (SEM) images of homogeneous elastomers and PSE with varying degrees of phase separation. The homogeneous elastomer was prepared using the highly polar solvent PC. PSE with different degrees of phase separation were achieved by adjusting the ratio of monomer to solvent (DMP). The SEM images (Figure S8) clearly reveal that as the degree of phase separation increases, the internal structure of the elastomer exhibits more pronounced undulations, which may ultimately lead to material aggregation. Based on these findings, the redshift in color is not only influenced by the polymer density but is also closely related to the solid content of the system. When the solid content of the system is higher, the degree of phase separation is relatively weaker, thereby reducing the occurrence of aggregation and resulting in a more loosely arranged polymer phase. The effect of the lower refractive index of the polymer phase is equivalent to an increase in the refractive index of the poor solvent, which ultimately manifests itself as a red-shift effect of the color. This series of findings not only deepens our understanding of the optical properties of PSE, but also provides valuable experimental basis and theoretical guidance for the optimized design and application of subsequent resins.

After deeply dissecting the property of PE color change, we designed a series of demonstration experiments, the results of which are presented in Figures 4A and 4B. Figure 4A shows the fluid system with the Christiansen effect. Specifically, we added 20% mass fraction of PMMA microspheres with similar refractive indices to a mixed solution of TBC and DMP, which was green at room temperature by finely tuning the ratio of the mixed solution. We then cooled the solution to −20°C and used a hair dryer for localized top heating. As the temperature increased, the color of the solution experienced a transition from orange to purple starting from the top (Video S1). Figure 4B shows PSE-0.4, which was also cooled to −20°C but then heated uniformly, and the color of PSE-0.4 shows a very different color transition path, with the overall color changing uniformly from orange to purple (Video S2). These two graphs illustrate that the generation of color under the Christiansen effect is not limited to the solid state, but is based on the presence of phase separation and the matching of the refractive indices of the multiple phases. The difference in the heating method is able to precisely direct the regional and directional nature of the color change, enabling selective manipulation of the color distribution. This extraordinary property opens up a novel pathway for encoding and transmitting complex information through color changes. Figure S9 shows a simple archery target we prepared using this light-curable resin. We adjusted the resin components to prepare different colored rings and finally assembled the archery target.

Figure 4.

Thermochromic 3D-printed gradient elastomers via Christiansen effect and automated fabrication

(A) Photograph of the color change of the solution with the Christiansen effect when it is heated from the top.

(B) Photograph of the color change of PSE-0.3 when it is heated uniformly.

(C) Photograph of the internal structure of a 3D printer with an automatic feeding system.

(D) Photograph of a tower model for 3D printing.

(E) Tower elastomer with continuous color change obtained by 3D printing (from purple to red).

(F) Tower elastomer with continuous color change obtained by 3D printing (from red to purple).

Thermochromic 3D-printed gradient elastomers via Christiansen effect

In addition, given the properties of PSE, we cleverly used 3D printing technology for the light-curing reaction. After the optimization of the reaction system, we preferred a liquid crystal display (LCD) 3D printer with a wavelength of 385 nm, and added 0.2% mass fraction of photoinitiator 819 and 0.1% mass fraction of Omnistab 326 (bumetrizole), the molecular structures of which are shown in Figure S10. Although photoinitiator 184 and diphenyl(2,4,6-trimethylbenzoyl)phosphine oxide (TPO) were also attempted to be used in this process, we finally decided to use photoinitiator 819 as the initiator for this work due to its higher initiation efficiency. The introduction of bumetrizole is a key step to improve the printing accuracy, which can effectively regulate the exposure level and prevent overexposure phenomenon, showing significant advantages for the accurate printing of fine structures.⁴¹ However, the amount of its addition needs to be strictly controlled so as not to over-absorb low-wavelength light and weaken the blue-violet color produced by the Christiansen effect. This is supported by the significant weakening of the blue color of the PSE-0.5 after the addition of bumetrizole, as shown in Figure S11. Figure 4C shows the 3D printer we used, which is equipped with an automatic feeding system on the right front to continuously replenish the liquid feedstock for the printing process. This design allows us to pre-position the PSE-0 precursor solution in the trough, while the supply tank is responsible for supplying the TBC-based solvent as the replenishment solution. During the printing process, we simply control the flow rate of the replenishment solution. As the replenishment solution is added, the color of the printed product transitions from red to purple and vice versa. This innovative technology greatly enriches the color level of the printed product. This study employed standardized 3D printing parameters: Layer thickness: 100 μm, Base layer exposure: 10 s (3 layers), subsequent layer exposure: 5.8 s, light source: 385 nm UV at constant intensity of 6 mW cm⁻². In order to visualize this color change effect, we chose a tower model with a high sense of hierarchy as the print object, as shown in Figure 4D. Figures 4E and 4F present the printing results of transition from purple to red and from red to purple, respectively, which fully verifies the feasibility and artistry of this technology and opens up a new way of color regulation and expression in the field of 3D printing. On this basis, we can still change the color by changing the temperature. As shown in Figure S12, the tower structure is an intuitive result of the sample shown in Figure 4E after cryoprocessing. From the Figure S12, we can see that the tip position of the tower has changed from red to opaque, and the blue color at the bottom of the tower has also changed to yellow, which shows an overall red shift in color. Besides, we conducted systematic stability studies on the sample shown in Figure 4E. As we can see from Figure S13, the middle image displays the sample prepared in July 2024, which has been 8 months of air exposure under ambient conditions. Long-term observation revealed that while the sample exhibited slight fading in color saturation—attributable to the kinetic process of gradual solvent evaporation—its structural coloration remained clearly discernible. In high-temperature stability testing, after six hours of thermal treatment at 100°C, the Christiansen effect-induced structural colors nearly disappeared while the sample turned yellowish. This failure mechanism stems from the disruption of microphase separation caused by substantial solvent loss. The residual yellowish appearance primarily originates from the intrinsic color of photoinitiator 819. Our findings demonstrate that the material maintains excellent long-term stability under normal environmental conditions. For practical applications in integrated devices, proper encapsulation would effectively address solvent evaporation concerns.

In particular, comparing Figures 4E with 4F, it is easy to see that the top of the tower shows lower saturation, while the bottom of the tower is relatively full of color. This difference can be attributed to the fact that the top of the tower is thinner than the bottom of the tower, and the light path through the phase-separated structure is shorter, so the selective transmission of monochromatic light is weaker. To further visualize the comparison, a solid-color tower sample was fabricated, as shown in Figure S14 Although no changes were made to the precursor solution during the printing process, the blue color at the base of the tower is significantly more saturated compared to the top of the tower. However, this phenomenon does not change the basic principle of color gradation, which is simply an indication of the correlation between material thickness and optical properties. Finally, using this parameter set, we successfully printed four kinds of pagoda-structured samples with different solvent ratios (PSE-0, PSE-0.2, PSE-0.4, and PSE-0.6). All four printing trials maintained strict parameter consistency and completed successfully, demonstrating the robustness and reproducibility of this printing protocol. As shown in Figure S15, the samples with different solvent ratios exhibit distinct optical coloration characteristics, demonstrating that the solvent ratio effectively modulates PSEs’ optical properties.

In summary, the precursor solutions exhibit similar physicochemical properties regardless of whether the finished print is red or purple. The fine-tuning of the resin composition does not affect our established system of printing parameters, i.e., key parameters such as exposure time and substrate running rate are kept constant and do not fluctuate with subtle changes in resin composition. This feature ensures the stability of the printing process and the continuity of color changes, and lays a solid foundation for a multi-material printing strategy in three-dimensional space. Therefore, the introduction of this type of resin not only enriches the diversity of 3D printing materials, but also opens up completely new possibilities in this field, signaling that the manufacture of more complex 3D printing products will become a reality.

Conclusion

Based on the Christiansen effect, we have designed a light-curing resin that imparts continuous and controlled color change properties to the cured polymer. The resin system uses HEA as the monomer and DMP and TBC as the poor solvents to regulate the phase separation, resulting in an elastomer with a bicontinuous phase-separated structure (PSE), which is unique in that the solvent and polymer phases have similar refractive indices, a characteristic that allows the PSE to exhibit excellent selectivity for monochromatic light at specific wavelengths. By finely tuning the ratio of the two solvents and the environmental temperature, we were able to achieve a full-spectrum iridescent gradation of PSE colors from red to violet, providing unprecedented flexibility and precision in material color design. In addition, the mechanical properties of PSE are outstanding, with elongation at break exceeding 600% of the original length and Young’s modulus exceeding 300 kPa, ensuring the structural integrity and durability of the material in a wide range of application scenarios. PSE also excels in temperature adaptability, with an operating temperature range of −25°C–130°C, meeting the needs of a wide range of extreme environments. Most importantly, the resin is perfectly compatible with 3D printing technology and allows the composition of poor solvents to be adjusted in real time during the printing process without any modification to the printing parameters. This innovative breakthrough not only achieves the feat of continuous multi-material printing in three-dimensional space, but also ensures the continuity and controllability of color changes during the printing process, which greatly broadens the application of 3D printing technology in the fields of material science and product design.

Limitations of the study

While this work demonstrates a solvent-mediated structural coloration strategy with promising adaptability, several limitations warrant consideration. First, the color modulation relies on precise refractive index matching between solvents and polymers, which restrict solvent selection to systems with compatible physicochemical properties. The volatility and potential toxicity of current solvents (e.g., DMP) may limit biocompatible applications. Second, although the elastomers operate across a wide thermal range (−25°C–130°C), prolonged exposure to extreme temperatures could accelerate solvent evaporation, potentially compromising color stability. Addressing these challenges could enhance the material’s versatility for industrial and biomedical applications.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Qihao Jiang.

Materials availability

This study did not generate new unique reagents.

Data and code availability

All data reported in this paper will be shared by the lead contact upon request. This paper does not report original code. Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Acknowledgments

We would like to acknowledge Dr. Simone Dimartino (School of Engineering, University of Edinburgh) for providing editing support. This research was supported by the China Postdoctoral Science Foundation (2024M752514) and Natural Science Basic Research Program of Shaanxi (program no. 2024JC-YBQN-0137).

Author contributions

Y.G. and Q.J. designed and conducted the experiments. Y.G. and Q.J. performed sample testing. Y.G. wrote the manuscript. Q.J. helped write the manuscript. Y.L. supervised the research. All authors contributed to the general discussion.

Declaration of interests

The authors declare no conflict of interest.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Chemicals
Tributyl citrate	Aladdin	CAS: 77-94-1
Hydroxyethyl acrylate	Aladdin	CAS: 818-61-1
Dimethyl phthalate	Aladdin	CAS: 131-11-3
4′-pentyl-4-cyanobiphenyl	Aladdin	CAS: 40817-08-1
Irgacure 819	Aladdin	CAS: 162881-26-7
Omnistab 326	Rhawn	CAS: 3896-11-5
Deposited data
All data reported in this paper will be shared by the lead contact upon request

Method details

Materials

All the reagents in this work were purchased from Aladdin and Rhawn, they are analytically pure and can be used directly without further purification.

Synthesis of PSEs

All the chemicals used in this research were of analytical grade. For the typical synthesis of PSE, HEA and poor solvents (DMP and TBC) were mixed under magnetic stirring for 1 min to form a homogeneous solvent mixture. Photo-initiator 819 were then added to the mixture and continuously stirred for another 2 min at room temperature to develop a transparent precursor solution. The molar percentages of photo-initiator 819 was controlled at 0.2%, respectively. Finally, the as-prepared precursor solution was injected into a glassy mould and cured by ultraviolet light irradiation (365 nm, 400 W power) for 2 min to obtain the PSE samples.

Mechanical tests

PSEs were cut into dumbbell-shaped tensile samples with the dimensions of 12 mm × 2 mm × 2mm (length × width × thickness). Then, these tensile samples were stretched on an electronic tensile machine (CMT6503, MTS) with a 50 N sensor to evaluate their mechanical performance under the stretching speed of 50 mm min^-1.

Glass temperature tests

The glass temperature measurements were performed on a differential scanning calorimeter (DSC822E). The temperature range was from -100°C to 30°C under liquid nitrogen and a heating rate was 10°C min^-1.

Thermogravimetric analysis (TGA) measurements

The TGA measurements were performed on a TG209C via scanning a temperature range from room temperature up to 800°C (10°C min^-1) under flowing N₂.

Transmittance tests

Transmittance testing is accomplished with a UV-visible spectrophotometer (Lambda 950), and the test wavelength range is 400-800nm. The thickness of PSEs was 1mm.

Scanning electron microscope (SEM) tests: The morphologies of samples were examined by scanning electron microscopy (SEM, Zeiss, Gemini 500).

Small-angle x-ray scattering (SAXS) tests

SAXS measurements on PSDG films were performed on a SAXS point 2.0 instrument (Anton Paar, Austria) equipped with a Cu/Mo dual microfocal X-ray source and a two-dimensional hybrid photon counting detector (EIGER R 1M)

Atomic Force Microscope (AFM) tests

The microphase structure of PSDGs on microsubstrates were measured by tapping-mode atomic force microscopy (AFM, Bruker, INNOVA) using tapping MPP-rotated cantilevers with silicon probes. The amplitude set point was adjusted above 250 mV, and the AFM phase images were collected.

Quantification and statistical analysis

No statistical analyses or significance tests were conducted. All the data in this manuscript were plotted using Origin software (Version: Origin 2018 64-bit).

Published: June 9, 2025