Effects of Fatty Acid Anhydride on the Structure and Thermal Properties of Cellulose-g-Polyoxyethylene (2) Hexadecyl Ether

Cellulose was premodified by short-chain fatty acid anhydrides, such as acetic anhydride (CA), propionic anhydride (CP), and butyric anhydride (CB), followed by grafting of polyoxyethylene (2) hexadecyl ether (E2C16) using toluene-2,4-diisocyanate as a coupling agent. The feeding molar ratio of E2C16 and the anhydroglucose unit (AGU) was fixed at 4:1, and then a series of CA-g-E2C16, CP-g-E2C16, and CB-g-E2C16 copolymers were successfully prepared. The structures and properties of the copolymers were characterized using FTIR (fourier transform infrared spectra), 1H-NMR (Proton nuclear magnetic resonance), DSC (Differential scanning calorimeter), POM (polarized light microscopy), TGA (thermogravimetric analysis) and WAXD (wide-angle X-ray diffraction). It was shown that with the anhydride/AGU ratio increasing, the degree of substitution (DS) value of E2C16 showed a trend of up first and then down. With the carbon chain length increasing, the DS value of E2C16 continuously increases. The phase transition temperature and thermal enthalpy of the copolymers increased with an increasing DS value of E2C16. When the ratio of CB/AGU was 1.5:1, the DS of E2C16 was up to the maximum value of 1.02, and the corresponding melting enthalpy and crystallization enthalpy were 32 J/g and 30 J/g, respectively. The copolymers showed solid–solid phase change behavior. The heat resistant temperature of cellulose-based solid–solid phase change materials was always higher than 270 °C. After the grafting reaction, the crystallinity of E2C16 decreased, while the crystal type was still hexagonal.


Introduction
Cellulose is a type of polysaccharide composed of repetitive D-glucose units linked through β(1→4) glycosidic bonds [1]. There are three active -OH groups in each D-glucose unit, and these can easily form strong intramolecular and intermolecular hydrogen bonds, resulting in smaller distances between molecules and rigid chains. As a result, cellulose is insoluble in common organic solvents and unable to be melt-processed before reaching its decomposition temperature [2,3]. Through modification [4][5][6][7][8][9], hydrogen bonding was weakened to a certain extent, while a grafted side chain increased molecular distance, which could further improve the solubility of cellulose. Furthermore, the increased molecular distance is helpful for the grafting reaction [10]. When the degree of substitution reaches a critical point, the cellulose derivatives will be melt-processable.

Preparation of Cellulose-Based Copolymers
The prepolymer solution and DBTDL (0.1 wt % of TDI) were added to cellulose ester solution. The mixture was stirred at 90 °C for 6 h. When the grafting reaction was complete, the reaction mixture was precipitated in distilled water for 12 h. Then, the crude product was filtered and soaked in acetone for another 48 h to dissolve ungrafted prepolymer in precipitation. After that, the product was filtered, dried, and weighed.

FTIR
Fourier transform infrared spectra (FTIR) were collected using a Bruker Tensor-37 Fourier (Germany) transform infrared spectrometer. The dried samples were tested by the KBr disc technique, and infrared spectra were recorded in the range of 4000-400 cm −1 .

1 H-NMR
Proton nuclear magnetic resonance ( 1 H-NMR) spectroscopy was performed using an Avance AV-300 MHz nuclear magnetic resonance spectrometer. The solvent was dimethyl sulfoxide-d 6 .
The degree of substitution (DS) of fatty acid group and E 2 C 16 could be calculated by 1 H-NMR results.
DS 1 of an acetyl group could be calculated by Equation (1) as follows: where I AGU is the integral intensity of the hydrogen atoms in the AGU and C 6 (7H, σ = 3.0~5.5) and I CH3 is the integral intensity of the -CH 3 group (3H, δ = 1.7~2.2) in an acetyl group. DS 1 of a propionyl group could be also calculated by Equation (1), where I CH3 is the integral intensity of the -CH 3 group (3H, δ = 1.1) in a propionyl group. DS 1 of a butyryl group could be calculated by Equation (2) as follows: where I CH2 is the integral intensity of the -CH 2 -group (2H, δ = 1.6) in a butyryl group. DS 2 of E 2 C 16 could be calculated by Equation (3) as follows: where I phenyl is the integral intensity of benzene (3H, δ = 7.0~7.4) in TDI. Differential scanning calorimetry (DSC) was performed using a NETZSCH DSC 200 F3 (Germany). The samples were heated from 20 • C to 250 • C at 30 • C/min to eliminate the thermal history. Then, the samples were cooled to −20 • C and heated from −20 • C to 250 • C at 10 • C/min. All heating and cooling processes were performed under a nitrogen atmosphere. The thermal data were read from the second heating run and the first cooling run.

POM
The type of phase change was carried out with hot-stage polarized light microscopy using a BX51TRF. Samples were firstly heated to 100 • C and then cooled down to room temperature (20 • C) in order to remove the thermal history. Subsequently, they were heated to 100 • C at a heating rate of 10 • C/min to observe the thermal behaviors.

TG
Thermogravimetric analysis (TGA) was carried out in a NETZSCH STA 449F3 from 40 • C to 800 • C at a heating rate of 10 • C/min under a nitrogen atmosphere.

WAXD
Wide-angle X-ray diffraction (WAXD) was recorded on a D/MAX-2500 under conditions of 150 mA and 40 kV at 0 • C. The X-ray source is an 18-kW rotating anode X-ray generator equipped with a Cu target. The incident X-ray beam was monochromated by a pyrolytic graphite, and the data were measured in a range of 2θ = 4 •~4 0 • , λ = 1.54 Å. Figure 1 displays the FTIR spectra of cellulose, E 2 C 16 , CA-g-E 2 C 16 (A3), CP-g-E 2 C 16 (P3), and CB-g-E 2 C 16 (B3) copolymers. In the spectrum of cellulose, the wide peak at approximately 3200 cm −1 is assigned to the absorption band of the hydroxyl groups on the AGU, whose intensity was significantly reduced in the spectra of copolymers. In the spectrum of E 2 C 16 , the peaks at 2852 cm −1 and 2920 cm −1 are assigned to the characteristic absorption bands of the ethylene glycol repeat units and the saturated alkane group of E 2 C 16 , respectively, which also appeared in the spectra of CA-g-E 2 C 16 (A3), CP-g-E 2 C 16 (P3), and CB-g-E 2 C 16 (B3) copolymers. In the spectrum of copolymers, 1385 cm −1 is assigned to the absorption band of C-H, 1230 cm −1 is assigned to the absorption band of C-O, and 1750 cm −1 is assigned to the absorption band of C=O. It is demonstrated that fatty acid groups (acetyl group, propionyl group, and butyryl group) were successfully linked onto the cellulose backbone. We can observe the characteristic absorption band of the conjugated C=C double bond of the benzene ring in TDI at 1603 cm −1 and the amide bands [14] at 1700 cm −1 and 1545 cm −1 in the spectra of copolymers, which indicated that the E 2 C 16 was successfully grafted onto the cellulose backbone through TDI. Based on the analysis above, it could be declared that CA-g-E 2 C 16 , CP-g-E 2 C 16 , and CB-g-E 2 C 16 copolymers were successfully synthesized. groups (acetyl group, propionyl group, and butyryl group) were successfully linked onto the cellulose backbone. We can observe the characteristic absorption band of the conjugated C=C double bond of the benzene ring in TDI at 1603 cm −1 and the amide bands [14] at 1700 cm −1 and 1545 cm −1 in the spectra of copolymers, which indicated that the E2C16 was successfully grafted onto the cellulose backbone through TDI. Based on the analysis above, it could be declared that CA-g-E2C16, CP-g-E2C16, and CB-g-E2C16 copolymers were successfully synthesized. copolymers. Figure 1. FTIR spectra of cellulose, E 2 C 16 , CA-g-E 2 C 16 (A3), CP-g-E 2 C 16 (P3), and CB-g-E 2 C 16 (B3) copolymers. Figure 2 shows the 1 H-NMR spectrum of CA-g-E 2 C 16 (sample A3, a), CP-g-E 2 C 16 (sample P3, b), and CB-g-E 2 C 16 (sample B3, c) copolymers. The peaks at δ = 1.7~2.2 were assigned to the signals of the methyl protons of acetyl group, the peak at δ = 1.1 was assigned to the signal of the methyl protons of propionyl group, the peak at δ = 2.4 was assigned to the signal of the methylene protons of propionyl group, the peak at δ = 0.9 was assigned to the signal of the methyl protons of butyryl group, the peaks at δ = 1.6 and δ = 2.3 were assigned to the signals of the methylene protons of butyryl group [15], and the peaks at σ = 3.0~5.5 were assigned to the signals of the protons in the glucose unit ring and C6. It is also demonstrated that fatty acid groups were successfully linked onto the cellulose backbone. The peak at σ = 0.85 was assigned to the signal of the methyl protons of the alkyl chain in E 2 C 16 , the peaks at σ = 1.1~1.7 were assigned to the signals of the methylene protons of the alkyl chain in E 2 C 16 , the peak at σ = 3.4 was assigned to the signal of the methylene protons adjacent to the ether bond of the alkyl chain in E 2 C 16 , the peaks at σ = 3.5~3.8 were assigned to the signals of the protons of the ethylene glycol repeat units of E 2 C 16 , and the peaks at σ = 7.0~7.4 were assigned to the signals of the protons in the benzene ring of TDI. Based on the analysis above, it is evident that CA-g-E 2 C 16 , CP-g-E 2 C 16 , and CB-g-E 2 C 16 copolymers were successfully synthesized.

1 H-NMR
The DS values of CA-g-E 2 C 16 , CP-g-E 2 C 16 , and CB-g-E 2 C 16 copolymers are shown in Table 2. When the ratio of acid anhydride/AGU increased from 0.5:1 to 1.5:1, both DS 1 and DS 2 increased significantly. When the ratio of acid anhydride/AGU increased to 2:1, DS 1 continued to increase while DS 2 dropped. When the ratio of acid anhydride/AGU increased within a certain range, DS 1 and the distance between cellulose molecules increased due to the anhydride side group. This is helpful for the grafting reaction of E 2 C 16 , so DS 2 increased, as expected. However, when the feeding ratio of anhydride/AGU was higher than 1.5 mol %, DS 1 continued to increase, the number of free hydroxyl groups in AGU decreased, and the intermolecular steric hindrance became stronger, leading to a lower DS 2 value. the cellulose backbone. The peak at σ = 0.85 was assigned to the signal of the methyl protons of the alkyl chain in E2C16, the peaks at σ = 1.1~1.7 were assigned to the signals of the methylene protons of the alkyl chain in E2C16, the peak at σ = 3.4 was assigned to the signal of the methylene protons adjacent to the ether bond of the alkyl chain in E2C16, the peaks at σ = 3.5~3.8 were assigned to the signals of the protons of the ethylene glycol repeat units of E2C16, and the peaks at σ = 7.0~7.4 were assigned to the signals of the protons in the benzene ring of TDI. Based on the analysis above, it is evident that CA-g-E2C16, CP-g-E2C16, and CB-g-E2C16 copolymers were successfully synthesized. The DS values of CA-g-E2C16, CP-g-E2C16, and CB-g-E2C16 copolymers are shown in Table 2. When the ratio of acid anhydride/AGU increased from 0.5:1 to 1.5:1, both DS1 and DS2 increased significantly. When the ratio of acid anhydride/AGU increased to 2:1, DS1 continued to increase while DS2 dropped. When the ratio of acid anhydride/AGU increased within a certain range, DS1 and the distance between cellulose molecules increased due to the anhydride side group. This is helpful for the grafting reaction of E2C16, so DS2 increased, as expected. However, when the feeding  The calculated results showed that DS 2 becomes larger with increasing carbon atom number of acid anhydride because the distance between cellulose molecules increases and the free volume region increases. However, the DS 1 value of cellulose-based copolymers were roughly equal. This is because with the carbon atom number increasing, the effect of increasing distance between cellulose molecules becomes stronger, which is helpful for the grafting reaction of E 2 C 16 . We can conclude that the effect of increasing cellulose molecule spacing is in the order of BC > PC > AC. When the ratio of BC/AGU was 1.5:1, DS 2 of E 2 C 16 in all three types of copolymers reached the maximum value of 1.02. Table 3 show the thermal transition behavior of E 2 C 16 , CA-g-E 2 C 16 , CP-g-E 2 C 16 , and CB-g-E 2 C 16 copolymers with different DS values. Pure E 2 C 16 possessed distinct phase transition peaks with large latent heats (∆H m = 104 J/g; ∆H c = 95 J/g), proving it is a type of phase-change material with good thermal storage capacity. For CA-g-E 2 C 16 , CP-g-E 2 C 16 , and CB-g-E 2 C 16 copolymers, E 2 C 16 is the working phase change material (PCM), while the cellulosic backbone acts as the supporting skeleton for fixing the E 2 C 16 chains and contributes nothing to the enthalpy. Therefore, the thermal behavior of cellulose-based copolymers is in line with the changing trend of DS, as mentioned above.

Figures 3 and 4 and
The enthalpy is a linear relationship with the crystalline d-spacing [10,16]. As shown in Table 3, when the feeding ratio of fatty acid anhydride/AGU increased from 0.5:1 to 2.0:1, the value of DS 1 continuously increased, while the DS 2 reached a maximum value at 1.5:1 of fatty acid anhydride/AGU. It is ascribed to the steric hindrance effect derived from fatty acid anhydride. The reactivity of free radical polymerization is strongly affected by the steric hindrance of side chains, which rises with increase of the DS of side chains. Therefore, with an increasing fatty acid anhydride/AGU ratio, phase change enthalpy (∆H), T mo and T do of the copolymers all exhibited a similar trend of first increasing and then decreasing, and they also reached a maximum value at 1.5:1 of fatty acid anhydride/AGU. Similarly, when the acid anhydride content was the same, the phase change enthalpy of the copolymers also increased with increasing the number of fatty acid anhydride carbon atoms, which is in line with the changing trend of DS 2 . For the CB-g-E 2 C 16 copolymer, when the feeding ratio of BC/AGU was 1.5:1, DS 2 in all copolymers reached the maximum value of 1.02, while ∆H m and ∆H c also showed the maximum values of 32 J/g and 30 J/g, respectively. In addition, compared with pure E 2 C 16 , the phase transition peaks of the copolymers shifted slightly to a lower temperature due to the increase in the side-chain crystal imperfection.
transition peaks with large latent heats (ΔHm = 104 J/g; ΔHc = 95 J/g), proving it is a type of phase-change material with good thermal storage capacity. For CA-g-E2C16, CP-g-E2C16, and CB-g-E2C16 copolymers, E2C16 is the working phase change material (PCM), while the cellulosic backbone acts as the supporting skeleton for fixing the E2C16 chains and contributes nothing to the enthalpy. Therefore, the thermal behavior of cellulose-based copolymers is in line with the changing trend of DS, as mentioned above.   The enthalpy is a linear relationship with the crystalline d-spacing [10,16]. As shown in Table 3, when the feeding ratio of fatty acid anhydride/AGU increased from 0.5:1 to 2.0:1, the value of DS1 continuously increased, while the DS2 reached a maximum value at 1.5:1 of fatty acid anhydride/AGU. It is ascribed to the steric hindrance effect derived from fatty acid anhydride. The reactivity of free radical polymerization is strongly affected by the steric hindrance of side chains, which rises with increase of the DS of side chains. Therefore, with an increasing fatty acid anhydride/AGU ratio, phase change enthalpy (ΔH), Tmo and Tdo of the copolymers all exhibited a similar trend of first increasing and then decreasing, and they also reached a maximum value at 1.5:1 of fatty acid anhydride/AGU. Similarly, when the acid anhydride content was the same, the phase change enthalpy of the copolymers also increased with increasing the number of fatty acid anhydride carbon atoms, which is in line with the changing trend of DS2. For the CB-g-E2C16 copolymer, when the feeding ratio of BC/AGU was 1.5:1, DS2 in all copolymers reached the maximum value of 1.02, while ΔHm and ΔHc also showed the maximum values of 32 J/g and 30 J/g, Figure 4. DSC curves of cellulose, E 2 C 16 , cellulose-g-E 2 C 16 (S-B), CA-g-E 2 C 16 (A3), CP-g-E 2 C 16 (P3), and CB-g-E 2 C 16 (B3) copolymers. Table 3. Thermal transitions of E 2 C 16 , cellulose-g-E 2 C 16 (S-B), CA-g-E 2 C 16 , CP-g-E 2 C 16 , and CB-g-E 2 C 16 copolymers.

Sample
No.  a : the mass percent of E 2 C 16 in copolymer; b : the onset temperature of the melting peak; c : the peak temperature of the melting peak; d : the ∆H m of PCMs if in the condition of blending; e : the onset temperature of the crystallization peak; f : the peak temperature of the crystallization peak.
As shown in Figure 4, with increasing number of fatty acid anhydride carbon atoms, T m and T c of the copolymers follow the order CA-g-E 2 C 16 < CP-g-E 2 C 16 < CB-g-E 2 C 16 . Furthermore, the peak area follows the same order. This means that increasing the length of the carbon chain helps in improving the thermal storage capability.
Compared with ∆H m * based on E 2 C 16 content in the blends, ∆H m based on E 2 C 16 content in the graft copolymers had a higher value. The movement of grafted E 2 C 16 chains is confined so that they cannot freely arrange to crystallize, causing the imperfection of the E 2 C 16 crystals to increase and resulting in a decrease in enthalpy. This also demonstrates that the products were cellulose-based copolymers. Figure 5 shows POM photos of CA-g-E 2 C 16 (A3), CP-g-E 2 C 16 (P3), and CB-g-E 2 C 16 (B3) copolymers heated in the temperature range of 20~80 • C. As listed in Table 3, the melting onset temperatures of A3, P3, and B3 are 8.0 • C, 13.7 • C, and 14.2 • C, respectively. For CA-g-E 2 C 16 (A3), when the environmental temperature rose from 20 • C to 50 • C, the E 2 C 16 side chain of the copolymer experienced melting, and its phase changed from crystalline to amorphous morphology. However, its macroscopic state did not change from the solid state. When the temperature continued to rise to 80 • C, the macroscopic state of the copolymer still was not changed. This is because the melted E 2 C 16 side chain is confined by the cellulosic backbone, and it cannot flow freely; therefore, the CA-g-E 2 C 16 copolymer stayed in a solid state, on the whole. CP-g-E 2 C 16 (P3) and CB-g-E 2 C 16 (B3) copolymers also showed the same phenomenon. This demonstrated that the phase change type of the synthesized copolymers is solid-solid phase change; that is to say, CA-g-E 2 C 16 , CP-g-E 2 C 16 , and CB-g-E 2 C 16 copolymers are a type of solid-solid phase material (SSPCM).

TGA
The thermal stability of cellulose, E2C16, CA-g-E2C16, CP-g-E2C16, and CB-g-E2C16 copolymers were investigated by TG. It can be observed from Figure 6 and Table 4 that all the copolymers are not degradable and that almost no mass loss occurs before 280 °C. The onset decomposition of cellulose occurs at 319 °C [16]. The thermal stability of CA-g-E2C16, CP-g-E2C16, and CB-g-E2C16 copolymers dropped to 294 °C, 288 °C, and 289 °C, respectively. The lower thermal stability mainly is a result of the introduction of the fatty acid anhydride and E2C16, which significantly decrease the crystallinity of cellulose [17]. Furthermore, the presence of grafted side chains on the cellulose backbone was confirmed. It is worthy to note that the Tdo of CA-g-E2C16, CP-g-E2C16, and CB-g-E2C16 copolymers were still much higher than that of the raw E2C16 due to the confinement derived from the cellulose backbone, and were almost independent of the DS value. Compared with pure E2C16, the thermal stabilities of the cellulose-g-E2C16 copolymers were improved by approximately 42~50 °C. It is also shown that the thermal stability was improved gradually with the increase of DS. It was concluded that CA-g-E2C16, CP-g-E2C16, and CB-g-E2C16 copolymers can be applied as a type of SSPCM and show thermal stability in a high-temperature environment.

TGA
The thermal stability of cellulose, E 2 C 16 , CA-g-E 2 C 16 , CP-g-E 2 C 16 , and CB-g-E 2 C 16 copolymers were investigated by TG. It can be observed from Figure 6 and Table 4 that all the copolymers are not degradable and that almost no mass loss occurs before 280 • C. The onset decomposition of cellulose occurs at 319 • C [16]. The thermal stability of CA-g-E 2 C 16 , CP-g-E 2 C 16 , and CB-g-E 2 C 16 copolymers dropped to 294 • C, 288 • C, and 289 • C, respectively. The lower thermal stability mainly is a result of the introduction of the fatty acid anhydride and E 2 C 16 , which significantly decrease the crystallinity of cellulose [17]. Furthermore, the presence of grafted side chains on the cellulose backbone was confirmed. It is worthy to note that the T do of CA-g-E 2 C 16 , CP-g-E 2 C 16 , and CB-g-E 2 C 16 copolymers were still much higher than that of the raw E 2 C 16 due to the confinement derived from the cellulose backbone, and were almost independent of the DS value. Compared with pure E 2 C 16 , the thermal stabilities of the cellulose-g-E 2 C 16 copolymers were improved by approximately 42~50 • C. It is also shown that the thermal stability was improved gradually with the increase of DS. It was concluded that CA-g-E 2 C 16 , CP-g-E 2 C 16 , and CB-g-E 2 C 16 copolymers can be applied as a type of SSPCM and show thermal stability in a high-temperature environment.
were still much higher than that of the raw E2C16 due to the confinement derived from the cellulose backbone, and were almost independent of the DS value. Compared with pure E2C16, the thermal stabilities of the cellulose-g-E2C16 copolymers were improved by approximately 42~50 °C. It is also shown that the thermal stability was improved gradually with the increase of DS. It was concluded that CA-g-E2C16, CP-g-E2C16, and CB-g-E2C16 copolymers can be applied as a type of SSPCM and show thermal stability in a high-temperature environment. . Figure 6. TG analysis of cellulose, E2C16, (a) CA-g-E2C16, (b) CP-g-E2C16, and (c) CB-g-E2C16 copolymers. Table 4. Thermal stabilities of cellulose, E2C16, CA-g-E2C16, CP-g-E2C16, and CB-g-E2C16 copolymers.

Sample
No.    Figure 7 shows the WAXD patterns of cellulose, E 2 C 16 , CA-g-E 2 C 16 , CP-g-E 2 C 16 , and CB-g-E 2 C 16 copolymers that were obtained to investigate the change in the crystal type of E 2 C 16 before and after grafting. There are three diffraction peaks in the WAXD pattern of cellulose, whose locations are 14.8 • , 16.4 • , and 22.6 • , and the d-spacings are 0.60 nm, 0.54 nm, and 0.39 nm, respectively. E 2 C 16 shows two sharp peaks centered at 21.2 • and a small peak centered at 23.7 • , with d-spacings of 0.42 nm and 0.38 nm, and the crystal type of E 2 C 16 is hexagonal [18]. It is reported that the characteristic diffraction peaks (110) and (020) of cellulose II are locate at 19.9 • and 22.1 • [19]. It can be seen that CA-g-E 2 C 16 copolymers show diffraction peaks centered at 20.7 • (d-spacing = 0.43 nm) and 21.6 • (d-spacing = 0.41 nm), which were close to the characteristic diffraction peaks of cellulose II (2θ = 19.9 • ) and E 2 C 16 (2θ = 21.2 • ), respectively. However, the intensity of the three types of cellulose-based copolymers were all weaker than that of E 2 C 16 and cellulose itself, as expected. It is demonstrated that the transformation from cellulose I to cellulose II also occurred after the dissolution and regeneration in AmimCl. The analysis above indicated that the crystalline form of the grafted E 2 C 16 side chain onto the cellulose backbone is not affected by the cellulose molecule. Moreover, CP-g-E 2 C 16 (P3) and CB-g-E 2 C 16 (B3) copolymers also showed the same phenomenon. Figure 8 shows contradistinctive WAXD patterns of CA-g-E 2 C 16 , CP-g-E 2 C 16 , and CB-g-E 2 C 16 copolymers. It can be seen that the diffraction peak locations of these three copolymers are consistent, which indicated that a formed short side chain had no effect on the crystal type of E 2 C 16 .  Figure 7 shows the WAXD patterns of cellulose, E2C16, CA-g-E2C16, CP-g-E2C16, and CB-g-E2C16 copolymers that were obtained to investigate the change in the crystal type of E2C16 before and after grafting. There are three diffraction peaks in the WAXD pattern of cellulose, whose locations are 14.8°, 16.4°, and 22.6°, and the d-spacings are 0.60 nm, 0.54 nm, and 0.39 nm, respectively. E2C16 shows two sharp peaks centered at 21.2° and a small peak centered at 23.7°, with d-spacings of 0.42 nm and 0.38 nm, and the crystal type of E2C16 is hexagonal [18]. It is reported that the characteristic diffraction peaks (110) and (020) of cellulose II are locate at 19.9° and 22.1° [19]. It can be seen that CA-g-E2C16 copolymers show diffraction peaks centered at 20.7° (d-spacing = 0.43 nm) and 21.6° (d-spacing = 0.41 nm), which were close to the characteristic diffraction peaks of cellulose II (2θ = 19.9°) and E2C16 (2θ = 21.2°), respectively. However, the intensity of the three types of cellulose-based copolymers were all weaker than that of E2C16 and cellulose itself, as expected. It is demonstrated that the transformation from cellulose I to cellulose II also occurred after the dissolution and regeneration in AmimCl. The analysis above indicated that the crystalline form of the grafted E2C16 side chain onto the cellulose backbone is not affected by the cellulose molecule. Moreover, CP-g-E2C16 (P3) and CB-g-E2C16 (B3) copolymers also showed the same phenomenon. Figure 8 shows contradistinctive WAXD patterns of CA-g-E2C16, CP-g-E2C16, and CB-g-E2C16 copolymers. It can be seen that the diffraction peak locations of these three copolymers are consistent, which indicated that a formed short side chain had no effect on the crystal type of E2C16.

Conclusions
Cellulose was esterified using short-chain fatty acid anhydrides (acetic anhydride, propionic anhydride, and butyric anhydride), followed by grafting of polyoxyethylene (2) hexadecyl ether (E2C16) using toluene-2,4-diisocyanate as a coupling agent and with the feeding molar ratio of E2C16/AGU fixed at 4:1, and then a series of CA-g-E2C16, CP-g-E2C16, and CB-g-E2C16 copolymers were successfully obtained. It is shown that with the anhydride/AGU ratio increasing, the DS value of E2C16 was first increased and then decreased. With the carbon chain length increasing, the DS value WAXD patterns of cellulose, E 2 C 16 , (a) CA-g-E 2 C 16 , (b) CP-g-E 2 C 16 , and (c) CB-g-E 2 C 16 copolymers. Figure 7.

Conclusions
Cellulose was esterified using short-chain fatty acid anhydrides (acetic anhydride, propionic anhydride, and butyric anhydride), followed by grafting of polyoxyethylene (2) hexadecyl ether (E 2 C 16 ) using toluene-2,4-diisocyanate as a coupling agent and with the feeding molar ratio of E 2 C 16 /AGU fixed at 4:1, and then a series of CA-g-E 2 C 16 , CP-g-E 2 C 16 , and CB-g-E 2 C 16 copolymers were successfully obtained. It is shown that with the anhydride/AGU ratio increasing, the DS value of E 2 C 16 was first increased and then decreased. With the carbon chain length increasing, the DS value of E 2 C 16 showed a trend of increasing continuously. The phase transition temperature and enthalpy of the copolymers increased with increasing DS value of E 2 C 16 . When the ratio of butyric anhydride/AGU was 1.5:1, the DS of E 2 C 16 was up to the maximum value of 1.02, and the corresponding melting enthalpy and crystallization enthalpy were 32 J/g and 30 J/g, respectively. All copolymers showed solid-solid phase change behaviors. The heat resistant temperature of cellulose-based solid-solid phase change materials is always higher than 270 • C. After the grafting reaction, the crystallinity of E 2 C 16 is decreased, but the crystal type is still hexagonal.
Author Contributions: For research articles with several authors, a short paragraph specifying their individual contributions must be provided. The following statements should be used "Na Han and Xingxiang Zhang conceived and designed the experiments; Wanyong Yu and Na Han performed the experiments; Na Han and Yongqiang Qian analyzed the data; Na Han, Xingxiang Zhang and Wei Li contributed reagents/materials/analysis tools; Wanyong Yu and Na Han wrote the paper". Authorship must be limited to those who have contributed substantially to the work reported.