Co‐Harvest Phase‐Change Enthalpy and Isomerization Energy for High‐Energy Heat Output by Controlling Crystallization of Alkyl‐Grafted Azobenzene Molecules

Photoisomerization‐induced phase change are important for co‐harvesting the latent heat and isomerization energy of azobenzene molecules. Chemically optimizing heat output and energy delivery at alternating temperatures are challenging because of the differences in crystallizability and isomerization. This article reports two series of asymmetrically alkyl‐grafted azobenzene (Azo‐g), with and without a methyl group, that have an optically triggered phase change. Three exothermic modes were designed to utilize crystallization enthalpy (∆Hc) and photothermal (isomerization) energy (∆Hp) at different temperatures determined by the crystallization. Azo‐g has high heat output (275–303 J g−1) by synchronously releasing ∆Hc and ∆Hp over a wide temperature range (−79 °C to 25 °C). We fabricated a new distributed energy utilization and delivery system to realize a temperature increase of 6.6 °C at a temperature of −8 °C. The findings offer insight into selective utilization of latent heat and isomerization energy by molecular optimization of crystallization and isomerization processes.


Introduction
[7] Unlike conventional phase-change materials, photoresponsive Azo derivatives show a great ability to co-harvest the latent heat and photothermal energy by releasing them at two or three temperature regions, which have a relatively large temperature difference. [8]hus, Azo can potentially utilize or output the crystallization enthalpy (ΔH c ) and photothermal energy (ΔH p ), adapting to environmental temperature on the basis of an isomerizationinduced phase change. [9,10]According to the energy utilization process, the ΔH c and crystallization temperature (T c ) are used to quantify the crystallization performance, where the isomerization kinetics are important for regulating the heat output of Azo. [11,12]Further, this performance is important in the design of advanced energy utilization systems.
The crystallizability and T c of the E-and Zisomers (T c-E and T c-Z , respectively) of Azo determine the ability to release the latent heat and its upper temperature limit, respectively, which determine the heat output ability and temperature region. [13,14][22][23][24] However, chemical modification affects the crystallization of the Azo Z-isomer in particular.The intercalation of flexible chains into Azo assemblies reduces intermolecular π-stacking. [25,26][29] Thus, when the steric conformation changes, the crystallizability of Azo-g depends on the assembly of organic chains on the benzene.According to previous studies, an ordered molecular assembly formed by long flexible alkyl chains favors the crystallization via van der Waals forces for high ΔH c , but results in an increase in T c . [30,31]It remains a great challenge to enhance the crystallinity (and hence, ΔH c ) of Azo-g at a relatively low T c because of the trade-off between molecular interaction and chain motion.[34][35] Energy density is calculated as the sum of ΔH p and ΔH c . [9,36,37]The crystallization of the Eand Z-isomers are the prerequisite to the release of their ΔH c-E and ΔH c-Z , respectively.In general, the E-Azo-g crystal is able to release ΔH c-E at temperatures lower than T c-E , but the poor crystallization of twist Z-isomers affects the utilization of ΔH c-Z .Several organic chaingrafted Azo or arylazopyrazoles are amorphous even at a very low temperature. [38]Furthermore, the grafting of organic chains results in a change in the energy difference (ΔH p ) between Z-and E-isomers because of the change in steric conformation and molecular interaction.As a result, the energy density of heat output can be optimized by controlling the crystallization and energy level of Azo-g.According to the structural transformation and phase change, the basis of the exothermic mode is the light-driven synchronous or asynchronous release of ΔH c and ΔH p .Synchronous release controlled by a photo-induced solid-toliquid phase change at the specific temperature favors a high rate and high-density heat output, but this process is only obtained in the temperature region between T c-Z and T c-E . [13,25,39]Asynchronous release means that Azo-g can selectively output the latent heat (ΔH c-E or ΔH c-Z ) or ΔH p at different temperature regions controlled by temperature or light irradiation. [8]This process can be used for cyclic heat output in an energy delivery system.However, the diversity (length, steric conformation, and molecular interaction) of grafted organic chains onto Azo leads to the complex relationship between energy level and crystallization. [25,26]To the best of our knowledge, there have been few studies on the high-density heat output of Azo-g by optimizing synchronous and/or asynchronous ΔH c (ΔH c-E or ΔH c-Z ) and ΔH p based on controlling crystallization and isomerization processes.
Herein, we designed and prepared two asymmetric structures of alkyl-grafted azobenzenes to control the crystallization and isomerization by adjusting the alkyl chain length and methyl group.The substituent (chain) leads to changes in the steric conformation, molecular interaction, and energy level of E-and Z-isomers.The effects of the substituent on crystallization and photoisomerization were investigated.The crystallizability and T c of E-and Z-isomers determine the ability to release ΔH c and also affect photoisomerization followed by a potential phase change.We proposed three energy utilization modes to achieve selective release of ΔH c and ΔH p at different temperatures according to the crystallization of Z-isomers.Finally, we built a distributed energy utilization and delivery system, including light-driven movement, absorption, and heat release with a phase change.The increasing temperature, tracked by a high-resolution infrared thermal imaging camera, indicated that Azo-g had the ability to utilize both the latent heat and photothermal energy and deliver the energy at an alternating temperature.

Results and Discussion
One of the effective strategies for controlling the crystallization and isomerization of Azo-g is chemically optimizing the substituents or organic chains on the benzene rings. [8,14,15]In this study, we designed and prepared two series of Azo-g derivatives asymmetrically grafted by alkyl chains with carbon atoms from 6 (A6-Azo-g) to 10 (A10-Azo-g) on the benzene ring (Figure 1a).The methyl group was substituted on another benzene at the ortho-position of -N=Nto obtain the corresponding B6-Azo-g, B7-Azo-g, B8-Azo-g, B9-Azo-g, and B10-Azo-g.Fourier transform infrared (FT-IR) spectra, nuclear magnetic resonance (NMR), and high-resolution mass spectroscopy (HRMS) revealed that all Azo-g compounds were successfully prepared (see the Synthesis section, Supporting Information for details).Compared with Azo, the grafting of alkyl chains reduces the molecular planarity of Azo units and affects the regularity of the crystals.However, it favors the selfassembly via van der Waals forces, which might limit the isomerization process because of steric hindrance. [3,40,41]As a result, the crystallizability and isomerization are determined by molecular interactions (π-stacking and van der Waals forces) between Azo-g with different lengths and steric conformation.
The solvent-free photoisomerization of Azo-g was tracked by UV-Vis absorption spectra.As shown in Figure 1b,c and Figure S1, Supporting Information, all Azo-g show a similar time-evolved decrease and increase in the π-π* transition peak (330 nm) under alternate irradiation with UV (365 nm) and blue light (420 nm).The opposite trend in the n-π* transition peak at approximately 440 nm is also observed.The continuous change in absorption arises from reversible E-Z isomerization.At the photostationary, the degree of isomerization (DI = 50-72%, Table S1, Supporting Information) was calculated from the characteristic peaks in 1 H NMR spectra (Figure S2 and Equation S1, Supporting Information).As shown in Figure 1d, all Azo-g show a decrease in DI with increasing length of the alkyl chain.This indicates that the structural transformation is to some extent restricted by steric hindrance in the molecular assembly of alkyl chains.This is supported by the fact that with the same carbon atoms, B-Azo-g has a relatively high DI, high blue light-induced isomerization rate constant (к rev-blue = 5.44-5.93× 10 −3 s −1 ), and short dark half-life (t 1/2 = 52-54 h) compared with A-Azo-g (4.28-4.59× 10 −3 s −1 , 59-62 h; Figure 1e, Figures S3-S5, Tables S1 and S2, Equations S2 and S3, Supporting Information).This is also proved by the low-temperature isomerization performance (Figure S6 and Table S3, Supporting Information).And at low temperatures (−20 °C), Azo-g still has high blue light-induced isomerization rate constant (к rev-blue , 2.04-3.23 × 10 −3 s −1 ), and also has a long dark half-life (321-655 days), which can realize rapid heat release of blue light stimulation at low temperatures and long-term energy storage in the dark.The methyl group further affects the regular molecular assembly.As a result, the increased free volume favors reversible isomerization but potentially affects the crystallization because of the reduced πstacking. [42,43]The increase in structural transformability also leads to a change in ΔH c .Thus, the crystallization performance (ΔH c and T c ) of Azo-g is of essential importance for determining the heat output.
The difference in crystallization among two series of Azo-g determines ΔH c and the exothermic mode.As shown in Figure 2a-h,  Figures S7-S10, Tables S4 and S5, Supporting Information, the ΔH c , T c , and melting temperature (T m ) of Azo-g with different alkyl chains and substituents were measured by differential scanning calorimeter (DSC) and X-ray diffraction (XRD).The T c and T m of Azo-g are determined by the regularity of the crystal structure, which depends on molecular interactions, including π-stacking or van der Waals forces.It can be seen that all E-Azo-g exhibit supercooling (temperature difference between T c and T m ) of approximately 4-31 °C.
E-Azo-g is highly crystalline with several sharp XRD peaks at 2θ of 5-30°(Figure 2g,h and Figure S8, Supporting Information) and a relatively high T c compared with the Z-isomers, except for amorphous B6-Azo-g.Unlike crystallizable E-isomers (T c-E = −9 °C to 25 °C, except for E-B8-Azo-g of −25 °C), only four Z-isomers of A-Azo-g (Z-A7-Azo-g, Z-A8-Azo-g, Z-A9-Azo-g, and Z-A10-Azo-g) can crystallize at low temperature (T c-Z = −25 °C to −4 °C).In comparison, with the same carbon chains, the amorphous Z-isomer of B-Azo-g (n = 7-10) has an extremely low T g (−78 °C to −73 °C) and did not crystallize (Table S5, Supporting Information).These results showing the low T c-E (−25 °C to −2 °C), and favorable isomerization (DI = 62-75%; Figure 1d and Figure S10, Supporting Information) indicate that methyl groups further lower the crystallizability, in particular for twist Zisomers.These findings are corroborated by the fact that E-A-Azo-g has a lower supercooling temperature than that of E-B-Azo-g with the same alkyl chain (Figure S9, Supporting Information).The poor crystallization of the Z-isomer of B-Azo-g arises from the clearly reduced molecular π-stacking, which impedes the formation of the crystal.Interestingly, the crystallization of the Z-isomer of Azo-g mainly depends on the molecular assembly of carbon chains rather than Azo units.Therefore, the Z-isomer of A6-Azo-g with a short chain cannot crystallize with a low T g of −79 °C because of weak van der Waals forces.Importantly, the poor crystallization enables metastable Z-isomer to synchronously release both ΔH c and ΔH p controlled by isomerization-induced crystallization at low temperature.For A-Azo-g (n = 7-10), T c-E and T c-Z are the upper and lower temperature limits for synchronous heat output, respectively.Despite the lack of T c-Z , B-Azo-g (n = 7-10) and A6-Azo-g are able to release high-rate heat at temperatures below T c-E .According to previous studies, it was difficult to lower T c-E of azobenzene derivatives due to a rigid structure, thus limiting the temperature range for photothermal utilization. [21]The low T c-E or even T g endows the Azo-g to release heat for energy utilization at an extremely low temperature, which outperforms the previous azo-based molecules.
Surprisingly, T c-E (T m-E ) of Azo-g (except for B6-Azo-g) has a strong dependence on the odd-even number of carbons in the alkyl chains (Figure S10, Tables S4 and S5, Supporting Information).It can be seen that with increasing even number of carbons, A-Azo-g (n = 6, 8, and 10) and B-Azo-g (n = 6, 8, and 10) have increasing T c-E (A-Azo-g: from 9 °C to 25 °C; B-Azo-g: from amorphous to −2 °C) and T m-E (A-Azo-g: from 17 °C to 30 °C; B-Azo-g: from amorphous to 15 °C).A similar trend is observed for the E-isomer of A-Azo-g (n = 7 and 9) with increasing odd n.The increase in T c-E and T m-E arises from the increasingly high regularity of the E-crystal controlled by self-assembly.However, when the even n increases, the T c-E of B-Azo-g clearly increased (from amorphous to −2 °C) compared with similar T c-E (−9 °C to −7 °C) with an increasing odd carbon number.[46] In addition, the grafting of an alkyl chain with an even n in A-Azo-g resulted in higher T c-E and T m-E values compared to the odd n, which is not true for B-Azo-g.This dependence on the odd-even n is not observed for T c-Z and T m-Z .A-Azo-g exhibits a linear increase in T c-Z with increasing n. Figure 2d also shows that Z-A10-Azo-g crystallizes a second time at −25 °C because of the lower rate of crystallization at lower temperature.The results indicate that the amorphous Z-isomer of B-Azo-g had a larger decrease in crystallization compared with A-Azo-g because of its twist conformation (Figure 4d and Figure S11, Supporting Information).A6-Azo-g has the broadest temperature region (−78 °C to 9 °C) for synchronous release of ΔH p and ΔH c-E .
Compared with traditional phase-change materials, a characteristic of the performance of Azo is selective control of the heat output at different temperatures through the isomerization-induced phase change. [8,14,15]To illustrate this effect on the exothermic mode, we divided all Azo-g into three categories, mainly in terms of the crystallization of the Z-isomer (Tables S4-S8, Supporting Information).I) A-Azo-g (n = 7-10): both the E-and Z-isomers can crystallize at a specific temperature; II) A6-Azo-g, B-Azo-g (n = 7-10): the E-isomer can crystallize, but the Z-isomer is amorphous; extremely low temperature T g ; III) B6-Azo-g: both the E-and Z-isomers are amorphous.
Azo-g has three typical exothermic modes for heat output.Mode (I): A-Azo-g (n = 7-10) are able to release ΔH c-E , ΔH c-Z , and ΔH p by controlling the temperature and photoisomerization conditions.As shown in Figure 3a,b, a closed cycle of energy utilization consists of four phase-change times and three heat output times.Specifically: i) the starting E-isomer with the latent heat is liquid at temperatures above T m-E (16-30 °C).The material releases ΔHc-E through a liquid-tosolid (L-to-S) phase-change at temperatures below T c-E (7-25 °C).ii) Under UV irradiation, the solid E-isomer transforms to liquid Z-isomer to store ΔH p because of a low T c-Z (−24 °C to −4 °C).iii) When the temperature drops further to below T c-Z , the liquid Z-isomer realizes the second heat output (ΔH c-Z ) by crystallization.iv) Under blue light irradiation, the solid Z-isomer further releases ΔH p for the third heat output through Z-to-E isomerization.It is noteworthy that iii) and iv) may occur simultaneously in the temperature range from T c-Z to T c-E .The liquid Z-isomer can synchronously release both ΔH c-E and ΔH p at a high rate under irradiation.v) A-Azo-g returns to the initial liquid state of the E-isomer when the temperature increases to T m-E .
Mode (II): B-Azo-g (n = 7-10) and A6-Azo-g can release ΔH c-E and ΔH p without ΔH c-Z .As shown in Figure 3c,d, the closed cycle consists of four phase-change times and two heat output times.Specifically: i) and ii) are the same as that in Mode (I), where ΔH c-E is released at a relatively low temperature (T c-E = −25 °C to 9 °C).iii) Blue light irradiation enables the Z-isomer to synchronously release ΔH c-E and ΔH p (second heat output) even at an extremely low temperature down to approximately −79 °C.iv) This is the same process as v) in the Mode (I).
Mode (III): As shown in Figure 3e,f, B6-Azo-g can only store and release ΔH p through the isomerization reaction under alternate UV and blue irradiation and is not temperature dependent.
The crystallization of Z-isomer in Mode (I) offers a great potential for selectively releasing heat at different temperatures.According to the exothermic mode, the L-to-S phase change induced by the decrease in temperature or photoisomerization determines the synchronous or asynchronous release of ΔH c and ΔH p .As shown in Figure 4a, the temperature for the synchronous release of A-Azo-g (n = 7-10) is narrow in the range between T c-E and T c-Z .For example, the temperature range of A7-Azo-g is −24 °C to 7 °C.When the temperature decreases to T c-Z , A-Azo-g can only release ΔH p and no phase change occurs.The limited temperature range impedes high-rate heat output at an extremely low temperature.
This problem is solved by B-Azo-g (n = 7-10) and A6-Azo-g in Mode (II) without the crystallization of the Z-isomer (and hence, no ΔH c-Z ).As shown in Figure 4b, the Z-isomer alone with its T g can achieve high-rate light-induced heat release (ΔH c-E and ΔH p synchronously) even at an extremely low temperature of −79 °C.This system remarkably broadens the temperature region of heat output compared to existing materials.In contrast, B6-Azo-g can only output a small amount of heat via photoisomerization, regardless of the temperature.The temperature range of heat output and exothermic modes determine the application in energy utilization and delivery systems.
The energy density is another important performance factor for energy utilization.Therefore, ΔH c and ΔH p were measured by DSC.According to modes (I) and (II), A-Azo-g (n = 7-10) has two kinds of exothermic processes: two or three heat release processes.The total energy density (ΔH total ) was calculated by Equation (1) as ΔH total-1 for the cases with three heat release processes, or by Equation (2) as ΔH total-2 for the cases with two heat release processes In contrast, only ΔH total-2 was calculated for A6-Azo-g and B-Azo-g (n = 7-10).Details are shown in Tables S6-S8, Supporting Information. (1) Equation ( 3) shows that ΔH s is the sum of ΔH c-E and ΔH p , representing the amount of heat released synchronously by light-induced isomerization followed by the phase change.As shown in Figure 4c, we investigated the effect of the alkyl chain length on the ΔH c-E of Azog with different degrees of crystallinity.The ΔH c-E of A-Azo-g and B-Azo-g showed different dependence on the alkyl chain length.Specifically, the ΔH c-E of A-Azo-g clearly increased from 86.6 J g −1 (A6-Azog) to 106.6 J g −1 (A10-Azo-g) when grafted by a long alkyl chain because long chains favor molecular assembly via van der Waals forces to increase the regularity of the E-crystal.In contrast, ΔH c-E of B-Azo-g has little dependence on the alkyl chain length; the ΔH c-E values in the range of 81-84 J g −1 are clearly lower than those of A-Azo-g.Similar We also calculated the photothermal energy of all Azo-g studied here using density functional theory (DFT, ΔG) and compared the values with those measured experimentally by DSC (ΔH p ), to illustrate the effect of the substitution of Azo-g (Figure 4e,f and Tables S6-S9, Supporting Information).As shown in Figure 4d and Figure S11, Supporting Information, the geometric structure of E-and Z-isomers of Azo-g was simulated, and the E-isomer is in a planar structure, while the Zisomer is in a non-planar twisted structure, which also confirmed that the Z-isomer has a more poor regularity.Its makes it have weak crystallization ability.As shown in Figure 4e,f, A6-Azo-g has a maximum calculated ΔG of 200 J g −1 , and B6-Azog has a maximum experimental ΔH p of 132 J g −1 .All Azo-g show a continuously decreasing ΔG and ΔH p with increasing alkyl chain length.The reduction in ΔG is due to the lack of a contribution to the isomerization by the alkyl chains for photothermal storage.However, the calculated ΔG showed a different dependence on the alkyl chain length than the experimental ΔH p .For the same alkyl chain length, B-Azo-g has a lower calculated ΔG but a higher experimental ΔH p than that of A-Azo-g.For a fixed structure, the calculated ΔG is mainly determined by the energy difference between the metastable Z-isomer and stable E-isomer. [32,33]Thus, compared with A-Azo-g, the lower calculated ΔG of B-Azo-g indicates a smaller energy difference.This effect is attributed to a larger increase in the energy level of the Eisomer caused by methyl substitution compared to the case of the Z-isomer due to the reduced regularity of the Eisomer.For example, with the same alkyl chain, the calculated ΔG of B-Azo-g is approximately 15% lower than that of A-Azo-g on average.However, the experimental ΔH p is determined by the combined effect of DI and the energy difference. [32,47]As shown in Figure 1d and Table S1, Supporting Information, B-Azo-g has a much higher DI than that of A-Azo-g because of favorable isomerization.Thus, B-Azo-g has a higher experimental ΔH p than that of A-Azo-g.
The energy density of Azo-g can be maximized by optimizing exothermic modes.As shown for A-Azo-g (n = 7-10) in Figure 4c, ΔH c-E is much higher than ΔH c-Z because of the high regularity of the planar crystals in comparison to that of twist Z-isomers.Thus, as shown in Figure 5a, A-Azo-g (n = 7-10) in Mode (I) can release a higher amount of ΔH total-2 (2ΔH c-E + ΔH s ) than ΔH total-1 (ΔH c-E + ΔH c-Z + ΔH p ).This effect is more obvious for A-Azo-g when grafted by short alkyl chains.That is, A7-Azo-g has a larger difference between ΔH c-E and ΔH c-Z than that of A-Azo-g (n = 8-10).This process for ΔH total-2 is consistent with exothermic Mode (II).Thus, we demonstrate that A10-Azo-g achieves the maximum energy density of 303 J g −1 (Figure 5a and Table S7, Supporting Information) based on a double heat release over −25 °C to 25 °C, which includes the ΔH s of 196 J g −1 by light-driven synchronous release of ΔH c-E and ΔH p (Figure 5c).The ΔH s and temperature range can be further enhanced and broadened for energy utilization, in particular at a much lower temperature.As shown in Figure 5a-c and Table S8, Supporting Information, A6-Azo-g and B7-Azo-g can provide a high rate heat output of ΔH s (208 and 206 J g −1 , respectively) at temperatures above We demonstrate that by optimizing the structure of grafted chains, Azo-g can be designed to tune the crystallization and energy level of isomers based on changes in molecular interactions.The crystallinity and isomerization determine the energy density and exothermic modes of heat output.Azo-g is able to co-harvest the phase-change enthalpy and photothermal energy at different temperatures.5][16][17][18][20][21][22][23][24][25][26] These findings provide a new strategy for the cyclic utilization of energy with high density by optimizing the exothermic mode over a broad temperature range of thermal management.
We designed and fabricated a new distributed energy utilization system with the delivery of the sample controlled by temperature and light irradiation.Energy utilization is based on exothermic Mode (II) with two-time heat output.As shown in Figure 6a, 200 μL of E-A6-Azo-g (T c-E = 9 °C, T g = −79 °C) is sealed in a long quartz tube (length: 25 cm, inner diameter: 2 mm, and outer diameter: 1 mm), and the graphene aerogel (100 mg, 3.38 g cm −3 , diameter: 1.13 cm, and height: 1.64 cm) is placed in the quartz container on both sides of the tube as the light collector.The movement and stop of A6-Azo-g are controlled by alternate irradiation on graphene aerogel in the container based on the temperature difference between container A and container B. A semiconductor cooling plate is shifted along with the tube to offer a low-temperature environment.
Figure 6a and Video S1, Supporting Information show a schematic diagram of distributed energy utilization for the system.A typical cycle contains six steps: 1 Movement at 20 °C: Graphene aerogel (container A) is irradiated by UV light (365 nm, 200 mW cm −2 ) to collect light energy and convert it into heat, leading to an increase in temperature.As a result, the E-A6-Azo-g (orange liquid at region M) moves to region N because of thermal expansion of air in container A. The migration rate is 1.5 cm s −1 . 2 First heat release at 0 °C: When moving to region N at 0 °C, the orange E-A6-Azo-g liquid releases the latent ΔH c-E by crystallization (yellow solid) controlled by a semiconducting cooling plate.3 Heat storage by isomerization at 0 °C: After crystallization, the yellow E-A6-Azo-g isomerizes to red Z-isomer to store ΔH p by absorbing UV light (365 nm, 150 mW cm −2 ) for 40 min.The sample also transforms from solid to liquid spontaneously.4 Movement at −8 °C to 0 °C: After the isomerization, Z-A6-Azog (red liquid) continuously moves toward the low-temperature region O, driven by the irradiating graphene aerogel in container A. 5 Second heat release at −8 °C: When moving to region O at −8 °C, under blue light irradiation (425 nm, 100 mW cm −2 ) for 6.5 min, the red Z-A6-Azo-g liquid changes to a yellow Eisomer solid.It synchronously releases both ΔH c-E and ΔH p due to isomerization-induced recrystallization.6 Heat storage by melting at 20 °C: After isomerization and crystallization, the semiconductor is shifted to container B, and E-A6-Azo-g remelts to reform the initial orange liquid at 20 °C.
This distributed energy utilization cycle contains two heat output steps, and the delivery of the sample from region M to region O. A6-Azo-g also can repeat this cycle to reutilize photothermal energy and the latent heat with the movement from O to N by irradiating graphene We tracked the temperature of A6-Azo-g at regions N(T N ) and O(T O ) as a function of the heat output at 0 °C and −8 °C by a high-resolution infrared (IR) thermal imager.As shown in Figure 6b-e, Videos S2 and S3, Supporting Information, the changes in T N and T O depend on the heat released by A6-Azo-g, including sensible latent heat and photothermal energy.When moving to region N at 0 °C in step ( 2), E-A6-Azo-g gradually releases sensible heat, followed by a small increase in T N from 5.3 °C to 7.6 °C in 30 s (Figure 6b,c and Video S2, Supporting Information).T N then decreases slowly to 6.4 °C in 25 s because the sensible heat is not sufficient to neutralize heat diffusion to the low-temperature environment.In addition, the orange E-A6-Azo-g liquid starts to crystallize, followed by release of ΔH c-E (87 J g −1 ).The latent heat output results in a fast increase in T N from 6.4 °C to 9.6 °C over 75 s.Overall, the ΔT max in this process is 4.3 °C.A6-Azo-g is able to maintain T N above 6.4 °C during the stage from 55 to 177 s.T N starts to decrease again when the crystallization process approaches completion, and returns to 5.6 °C over 80 s.
After E-to-Z isomerization and movement to region O at −8 °C, the red Z-A6-Azo-g liquid was irradiated by blue light in step (5), and the E-isomer was also treated under the same conditions for comparison.As shown in Figure 6d,e, Figure S12 and Video S3, Supporting Information, the Z-and E-isomers produced similar increases in T O from −1.7 °C to 0 °C in approximately 30 s.This indicates that the increase in T O mainly arises from blue light thermal energy absorbed by A6-Azo-g.T O then increases quickly from 0 °C to 5 °C at 135 s, even in a low-temperature environment (−8 °C), because a large degree of Z-to-E isomerization followed by crystallization produces a high-density heat output (ΔH s of 208 J g −1 ).The ΔT max was 6.5 °C, and Z-A6-Azo-g maintained a T O higher than that of the E-isomer after 270 s.This result indicates that Z-A6-Azo-g exhibits a great ability to utilize energy to increase the T O for long-term temperature controllability.After crystallization, both the Z-and E-isomers gave a T O of 0.2 °C by reaching a balance between the blue light thermal effect and heat dissipation to the environment (−8 °C).The energy utilization system showed excellent cycling performance over five cycles with a ΔT max-N of 4.0-4.3°C and ΔT max-O of 6.2-6.6 °C (Figures S13 and S14, Supporting Information).
Z-A6-Azo-g clearly outperforms E-A6-Azo-g in terms of low-temperature heat output due to a high ΔT max and long-time rising T O (Figure S15, Supporting Information).This result is consistent with the fact that Z-A6-Azo-g enables light-driven synchronous release of higher density heat (ΔH c-E and ΔH p ) than that of ΔH c-E .Importantly, heat release is selectively accelerated, slowed, and even stopped by irradiating the material at high or low light density or no light.Moreover, Z-A6-Azo-g is able to release heat at any temperature in the range from T g (−79 °C) to T c-E (9 °C) or in any region by controlling the movement.This unique performance offers great potential for fabricating new energy utilization systems with unique properties compared to previous phase-change-or photothermal-based energy storage systems.The new energy utilization devices can be developed for photo-switchable microfluidic system by optimizing energy delivery.

Conclusion
We prepared two series of optically triggered phase-change Azo-g materials.The crystallization and photoisomerization processes were regulated by adjusting the alkyl chain length and methyl group.All Azo-g showed a decrease in DI (50-72%) and к rev-blue (4.29-6.03× 10 −3 s −1 ) with increasing alkyl chain length.After modification of the methyl substituent, the decrease in π-stacking reduces steric hindrance, which favors the photoisomerization of B-Azo-g and also affects the crystallization.The T c-E and T m-E of A-Azo-g had a high dependence on the odd-even number of carbons in the alkyl chains, unlike B-Azo-g.A-Azo-g had higher crystallizability with increasing T c-E and ΔH c-E compared with B-Azo-g.The crystallization and isomerization processes are also controlled by the alkyl chain length.Therefore, B-Azo-g (n = 7-10) and A6-Azo-g had extremely low T g of −73 °C to −79 °C without crystallization, and also store a relatively high amount of ΔH p .Based on this, we demonstrated three types of exothermic modes to utilize the latent heat and photothermal energy.As a result, Azo-g achieves a high ΔH total of 275-303 J g −1 by co-harvesting ΔH c and ΔH p over a
Synthesis of A-and B-Azo-g: First, Reactant 1 (5 mmol) was added to a 100 mL three-necked flask, and 50 mL of toluene was injected as solvent using a syringe.After dissolving well by sonication, Reactant 2 (5 mmol) and 1.14 mL of glacial acetic acid (CH 3 COOH, 20 mmol) were added.The reaction was carried out under nitrogen (N 2 ) at a reflux temperature of 60 °C for 24 h.The toluene solvent was then removed by rotary evaporation to obtain the crude product, which was dried in a vacuum oven at 60 °C for 12 h.The crude product was purified by column chromatography using a silica gel column with a mixture of petroleum ether and ethyl acetate (Volume ratio, 20:1) as the eluent.The eluent was then removed by rotary evaporation, and the product was dried in a vacuum oven at 60 °C for 24 h.The final orange liquid was obtained.Reactant 1 comprised nitrosobenzene and o-nitrosotoluene.Reactant 2 comprised 4-pentylaniline, hexylaniline, 4-n-4-heptylaniline, 4-octylaniline, p-nonylaniline, and 4-decylaniline.
Computational methods: The single-point energies of the E-and Z-Azo-g were calculated using density functional theory (DFT).The DFT calculations were performed using Gaussian16 at the B3LYP/6-31G(d) level with the Grimme's D3 (BJ) empirical dispersion correction.
Structural characterization: The chemical structure of the Azo-g was characterized using Fourier transform infrared (FTIR) spectroscopy recorded on an IN10 spectrometer (Thermo Scientific Nicolet, USA) with a disc of KBr. 1 H and 13 C NMR and spectra were collected using a AVANCE III HD 400 spectrometer (Bruker, USA).High-resolution mass spectrometry (HRMS) was performed on an UltrafleXtreme MALDI TOF/TOF (Bruker).
E-to-Z isomerization: For E-to-Z isomerization, Azo-g without solvent was irradiated under UV light (HTLD-4II; Shenzhen Height-LED Optoelectronics Tech Ltd., China).The light intensity was measured with a full-spectrum optical power meter (CEL-NP2000-10; Beijing China Education Au-light Ltd.).The 25 μL of samples without solvent were transferred to glass substrates and irradiated by a UV light (365 nm, 150 mW cm −2 ) for different times at room temperature (25.0 °C).The UV irradiation station was covered with a container and an aluminum foil to Energy Environ.Mater.2024, 7, e12607 block ambient light exposure.The samples with different UV irradiation times were then dissolved in 3 mL of ethyl acetate (0.025 mg mL −1 ) and then subjected to UV-Vis spectroscopy using a UV-3600 UV-Visible spectrometer (Shimadzu Ltd., Japan).
Z-to-E isomerization: For Z-to-E isomerization, the 25 μL of samples without solvent after 60 min UV irradiation were irradiated under blue light (420 nm, 100 mW cm −2 ) or dark environment for different times at 25.0 °C.The samples with different blue irradiation times or different storage times in the dark were then dissolved in 3 mL of ethyl acetate (0.025 mg mL −1 ) and then subjected to UV-Vis spectroscopy using UV-Visible spectrometer.The rate of Z-to-E isomerization was calculated based on the intensity change of the transition bands of Azo-g from the UV-Vis absorption spectra.
Degree of isomerization: The degree of isomerization (DI) of the 425 nm PSS was measured based on 1 H NMR spectra.The 25 μL of samples without solvent were transferred to glass substrates and irradiated by a UV light (365 nm, 150 mW cm −2 ) for 60 min at room temperature (25.0 °C).Then, 5 μL of samples were then dissolved in chloroform-d and subjected to 1 H NMR spectra.
Differential scanning calorimetry: Separate differential scanning calorimetry (DSC) tests for E-and Z-isomers of the Azo-g.For the E-isomer, the samples were first heated on a hot table at 80 °C for 30 min, then with the sample cooled to room temperature, 5 μL of samples were taken in a liquid crucible and subjected to DSC testing under a nitrogen atmosphere with the following test procedure: 1) Initial temperature of 40 °C, 2) Cool down to −60 °C at a rate of 2 °C min −1 and hold for 1 min, 3) Rise to 60 °C at a rate of 2 °C min −1 and hold for 1 min, 4) Cool down to −60 °C at a rate of 2 °C min −1 and hold for 1 min, and 5) Rise to 60 °C at a rate of 2 °C min −1 and hold for 1 min.
For the Z-isomer, the samples without solvent first irradiated under UV light for 60 min, then with the sample cooled to room temperature, 5 μL of samples were taken in a liquid crucible and subjected to DSC testing under a nitrogen atmosphere with the following test procedure: 1) Initial temperature of 20 °C, 2) Cool down to −100 °C at a rate of 2 °C min −1 and hold for 1 min, and 3) Rise to 150 °C at a rate of 2 °C min −1 and hold for 1 min.
Distributed energy device: We investigated the feasibility of heat release from A6-Azo-g at different low temperatures.We created a distributed energy device with a quartz tube (length: 25 cm, inner diameter: 2 mm, and outer diameter: 1 mm) in the middle and ground glass bottles on both sides.The two ground glass bottles were used to hold the graphene aerogel.The thermal effect of the graphene aerogel was used to achieve expansion of the gas, which drove the flow of Azo-g in the quartz tube.The recycling of a distributed energy device was divided into the following six steps: 1 Illuminating the graphene aerogel at container A with light so that A6-Azo-g moved from region M of the quartz tube to region N, and then stopping the illumination.2 Cooling down the A6-Azo-g to release heat.3 UV light (365 nm, 150 mW cm −2 ) irradiation causes A6-Azo-g to store latent heat and photothermal heat.4 The graphene in container A was illuminated, A6-Azo-g moved from region N of the quartz tube to region O, and then the illumination was stopped.5 Under blue light (420 nm, 100 mW cm −2 ) irradiation, the A6-Azo-g released latent heat and photothermal heat.6 A6-Azo-g was brought to room temperature by moving it to a cold table, which absorbs ambient heat at room temperature and stores latent heat.
Illuminating the graphene aerogel in container B with light caused A6-Azo-g reverse flow.During step (2), we tracked the time-evolved temperature change of E-A6-Azo-g using a high-resolution infrared thermal imaging camera (Fluke TiX640 Expert HD, USA).During step (4), we tracked the time-evolved temperature change of E-A6-Azo-g and Z-A6-Azo-g using a high-resolution infrared thermal imaging camera.

Figure 3 .
Figure 3. Schematic illustration of (a, c, e) three exothermic modes to utilize the latent heat and photothermal energy in a closed cycle based on reversible phase change and photoisomerization at different temperatures shown in (b, d, f) enthalpy versus temperature plots.a, b) Mode (I), c, d) mode (II), and e, f) mode (III).In (a, c, e), the orange, red segments, and gray lines indicate the planar E-, twist Z-isomer, and alkyl chains of Azo-g, respectively.In (b, d, f), the solid and dashed lines indicate the closed cycle of energy utilization.

Figure 6 .
Figure 6.a) Schematic illustration of a distributed energy utilization system.The cycle contains six steps: 1) movement at 20 °C; 2) first heat release at 0 °C; 3) heat storage by isomerization at 0 °C; 4) movement at −8 °C to 0 °C; 5) second heat release at −8 °C; and 6) heat storage by melting at 20 °C; reverse drive.b) Infrared thermogram and c) time-evolved T N of E-A6-Azo-g at region N in step (2); d) infrared thermogram of Z-A6-Azo-g and e) time-evolved T O of Z-and E-A6-Azo-g at region O in step (5).