Hydrogen bond producers in powerful protic ionic liquids for enhancing dissolution of natural cellulose

Abstract

The manipulation of hydrogen bonding within protic ionic liquids is conducive to conquering the robust hydrogen bonding interactions in cellulose for its effective dissolution, but it is a great challenge to establish the delicate balance of hydrogen bonding network between solvent and cellulose. Herein, we proposed the concept of “hydrogen bond producers” for urea molecules in 1,1,3,3‐tetramethylguanidinium methoxyacetate acid ([TMGH][MAA]) to enhance the dissolution of cellulose. The optimization of physicochemical properties for [TMGH][MAA] solvent as a function of urea concentration revealed a remarkable increase in cellulose solubility from 13% to 17% (w/w) by adding only 0.25 wt% urea, highlighting the efficiency of [TMGH][MAA] as a powerful solvent for the dissolution of cellulose. The experimental and simulation results verified that the significant improvement on dissolution of cellulose was attributed to the hydrogen bonding interaction of urea molecules with ion pairs and part of free ions, reducing the interference with the active ions bonded to cellulose. Furthermore, the considerable enhancement on comprehensive properties of regenerated cellulose films demonstrated the effectiveness of [TMGH][MAA]/urea solvent. The concept of “hydrogen bond producers” presented here opens a new avenue for significantly enhancing the dissolution of natural cellulose, promoting the sustainable development in large‐scale processing of cellulose.

INTRODUCTION

The escalating global concerns regarding the rapid resource depletion, climate change, and species extinctions have promoted a transition from a traditional linear economy to a circular bioeconomy,^1,2 which seems to be highly promising to realize the resource sustainability. Biomass resource, the largest renewable energy reservoir on Earth,³ presents vast potential to be processed into a broader perspective of valuable products as alternatives to petroleum-based sources. As the inexhaustible biomass resource, cellulose has been extensively utilized in paper, textiles, films, coatings, medical products, and polymeric materials due to its attractive mechanical properties and excellent chemistry stability.⁴ Regrettably, owing to the dense intrinsic hydrogen bonding network, it is a very difficult task for processing cellulose in the form of melt or solution in common solvents. This directly leads to a dilemma for natural cellulose in rapid prototyping, customizable shapes, and homogeneous modification,⁵ thereby limiting its application prospects for large-scale processing. Therefore, it has been a persistent challenge to conquer the robust hydrogen bonding interactions for the solution processing of natural cellulose.

Encouragingly, ionic liquids (ILs) are considered promising media for the dissolution of cellulose due to their strong capability in accepting and donating protons.^6,7 A groundbreaking work on the 1-butyl-3-imidazolium chloride has been carried out by Rogers Group for the dissolution of cellulose in ILss.⁸ Subsequently, the imidazolium-based ILs (1-allyl-3-methylimidazolium chloride and 1-ethyl-3-methylimidazolium acetate) reported by Zhang,^9,10 along with a range of protic ILs investigated by King,¹¹ have been demonstrated to hold the impressive abilities in dissolution of cellulose. Specifically, the protic IL-based Ioncell-F is presented as a promising candidate to Lyocell processes using viscose and N-methylmorpholine-N-oxide.¹² It is well accepted that the capacity of ILs for dissolving cellulose is closely related to their chemical components containing cations, anions, and ion complexes or oligomeric ions. Greaves et al.¹³ highlighted that the strong hydrogen bonding interactions between cations and anions led to the formation of abundant ion pairs. Rogers also pointed out that the hydrogen bonding interaction between cation–cation and anion–anion is responsible for the formation of oligomeric ions.¹⁴ In light of the well-established dissolution mechanism of ILs for cellulose, the dissociated ions or ion pairs are the adverse factors to interfere with the hydrogen bond formation between ions and cellulose, thus compromising the dissolution ability of ILs. To tackle this issue, it is a promising strategy for the structural design of ILs to enhance the solubility of cellulose. In our previous work,¹⁵ the configuration of 1,1,3,3-tetramethylguanidinium methoxyacetate acid ([TMGH][MAA]) with a Y-shaped conjugation in the cations and highly acidic anions effectively facilitated the transfer of protons, giving rise to the dissociation of ion pairs for interacting with cellulose via hydrogen bonding. An impressive solubility of 13.0% (w/w) was achieved at an appropriate 1,1,3,3-tetramethylguanidine/methoxyacetic acid (TMG/MAA) ratio of 7:3. However, in that case, excessive base also inevitably resulted in the generation of oligomeric ions. Thereby, inhibiting the formation of ion complexes in ILs is essential to ensure the high activity of cations and anions, destroying the intra- and intermolecular hydrogen bonds of cellulose.

It is well known that the protein denaturation in water is frequently accompanied by the conformational change from dense and highly ordered structure to loose and random structure due to the destruction of hydrogen bonds in protein. One of the proposed mechanisms for protein denaturation is that the hydrogen bonding structure between water molecules is manipulated via hydrophilic molecules, promoting the arrangement for interfacial water molecules at the protein surface with the aid of the strengthened interaction between water and protein for its denaturation.^16,17 Inspired by this, we proposed the concept of “hydrogen bond producers” to manipulate the hydrogen bonding structure between [TMGH] and [MAA], thus enhancing the dissolution of cellulose. As a typical hydrogen bonding dominated organic molecule,¹⁸ urea molecule is featured with the Y-shaped conjugation, exhibiting a large amount of π-electron delocalization.¹⁹ This allowed urea molecules to participate in hydrogen bond formation without excessively disrupting the original hydrogen bonds in [TMGH][MAA],^20,21 exhibiting the great potential as “hydrogen bond producers.” After optimizing the physicochemical properties of [TMGH][MAA] solvent as a function of urea concentrations, the cellulose solubility was significantly increased from the original 13% (w/w) up to 17% (w/w) with the addition of only 0.25 wt% urea at 80°C, strengthening the great competitiveness of [TMGH][MAA] as a powerful solvent compared to other cellulose solvent systems. The mechanism for the enhanced solubility was that a small number of urea molecules exclusively occupied the hydrogen bonding sites of ion pairs and part of free ions as “hydrogen bond producers,” thereby reducing the interference with the active ions bonded to cellulose. Moreover, the enhanced dissolution of cellulose improved the thermostability and mechanical performances of regenerated cellulose films with high transparency, exhibiting the satisfactory effectiveness of “hydrogen bond producers” for processing cellulose products on a large scale. This is the first report on the concept of “hydrogen bond producers” in solvents to significantly enhance the capacity for protic ILs on dissolving cellulose. This simple and efficient method for enhancing the solubility of cellulose in protic ILs has a great significance for developing high-performance cellulose products.

RESULTS AND DISCUSSION

The dissolution of cellulose is greatly dependent on the physicochemical properties of the [TMGH][MAA] solvent with the addition of urea, which are elaborated in detail by measuring the electrical conductivity and viscosity and using nuclear magnetic resonance (NMR), Fourier transform infrared (FTIR), differential scanning calorimetry (DSC), and thermogravimetric analysis (TGA) characterizations. As shown in Figure 1A, the conductivity of solvent system is plotted with the concentration of urea and temperature. It is anticipated that the conductivity increases with an increase of temperature at the same concentration of urea. At the fixed temperature, there is no dependence of the conductivity on the concentration of urea below 70°C. However, the conductivity above 70°C decreases first and then increases with the increase of the urea concentration. The lowest conductivity is obtained with the addition of 0.25 wt% urea. Figure 1B presents the viscosity of various [TMGH][MAA]/urea solvents as a function of temperature. As expected, it can be found that viscosity decreases with increasing temperature at a fixed concentration of urea. At the same temperature, the viscosity is observed to first decrease and then increase with an increase of urea concentration, reaching the lowest value with the addition of 0.1 wt% urea. These results reveal that only small amounts of urea molecules can tailor the conductivity and viscosity of [TMGH][MAA] solvent, which is possibly related to the interaction between urea and [TMGH][MAA].

[IMAGE OMITTED. SEE PDF]

As a highly effective method for probing diverse non-covalent interactions at a molecular scale, NMR spectroscopy is employed to discuss the mechanism of the interaction between urea and [TMGH][MAA].²² The ¹³C NMR spectra of pure [TMGH][MAA], urea, and [TMGH][MAA]-urea mixtures were collected, as shown in Figure 1C. As for [TMGH][MAA]-urea mixture, the V-C peak assigned to [MAA] shows a downfield shift from 172.54 to 173.21 ppm with the addition of urea. This indicates that urea molecules form hydrogen bonds with the –COO⁻, resulting in the decreased electron cloud density around V-C. By contrast, the II-C peak of [TMGH] is figured out to shift upfield from 164.61 to 164.56 ppm. It exhibits an increase in the electron cloud density around II-C because of the hydrogen bond information between urea molecules and the protons of [TMGH]. To further investigate the hydrogen bonding sites of urea, the ¹H NMR (Figure S1) and ¹⁵N NMR (Figure 1D) spectra were characterized. It can be observed in Figure S1 that the amino H atom peak of urea shifts downfield from 5.42 to 5.72 ppm after mixing the [TMGH][MAA], suggesting that [TMGH][MAA] interacts with amino H atoms via hydrogen bonding. Besides, the downfield shift of amino N atom peak in urea from 76.36 to 76.98 ppm (Figure 1D) indicates that the amino N atoms of urea participate in the hydrogen bonding interaction. This is in-line with the ¹³C NMR in Figure 1C, where the VI-C peak of urea shifts downfield. Therefore, the detailed interaction between urea and [TMGH][MAA] is as follows: the amino H atoms of urea form hydrogen bonds with the –COO⁻ groups of [MAA], whereas the amino N atoms of urea interact with the protons of [TMGH] by hydrogen bonding.

To further explore the dependence of hydrogen bonding interaction between urea and [TMGH][MAA] on the concentration of urea, the ¹³C NMR spectra of [TMGH][MAA] with various concentrations of urea were characterized. Figure 1E shows the impact of urea concentration on the ¹³C chemical shifts of II-C and V-C peaks of [TMGH][MAA]. The II-C peak initially shifts upfield (Δδ = −0.03 ppm) and then moves downfield as the concentration of urea increases. The largest upfield shift occurs at the urea concentration of 0.1 wt%. It indicates that urea molecules mainly interact with [TMGH] cations via hydrogen bonding, whereas the hydrogen bonds between ion pairs are not affected. The V-C peak shows a first downfield shift (Δδ = 0.07 ppm), followed by a upfield shift with the increase of urea concentration. The most significant downfield shift is observed at a urea concentration of 0.4 wt%, demonstrating that urea molecules primarily interact with [MAA] cations via hydrogen bonding. Taking the analysis above into account, it can be inferred that urea molecules produce the strongest hydrogen bonding interaction with both [TMGH] and [MAA] ions at the concentration between 0.1 and 0.4 wt%. This specific concentration value is further confirmed by the FTIR and DSC characterizations. As shown in Figure 1F, it is obvious that the board absorption band of solvent assigned to the N–H and O–H first shifts to a lower wavenumber and then shifts to a higher wavenumber. The lowest wavenumber is observed at the urea concentration of 0.25 wt%, indicating the highest hydrogen bond density in such a solvent system. In addition, the temperature corresponding to step signals in Figure 1G first increases and then decreases with the increase of urea concentration. The highest temperature at the urea concentration of 0.25 wt% implies that the highest hydrogen bond density significantly restricts the migration of ions in the solvent system. Similarly, the highest temperature of maximum volatilization rate (Figure S2) is seen at the urea concentration of 0.25 wt%. Thus, the possible mechanism for the interaction between urea and [TMGH][MAA] is proposed in Figure 1H. As the concentration of urea ranges from 0 to 0.25 wt%, urea molecules can interact with ion pairs and part of free ions as “hydrogen bond producers,” allowing the undisturbed migration of the active ions. This is the reason that conductivity and viscosity both decrease after adding a small amount of urea. When the urea concentration is beyond 0.25 wt%, urea molecules start to destroy the hydrogen bonds between ion pairs as “hydrogen bond breakers.” The dissociated ions further hinder the migration of the active ions, consequently leading to the increase of both conductivity and viscosity of solvent.

To investigate the impact of “hydrogen bond producers” for urea molecules on the dissolution of cellulose, the cotton linter solubility of various [TMGH][MAA]/urea solvents is summarized in Figure 2A. The solubility of cotton linters initially increases and then reduces as the urea concentration increases. It can be observed that there are nearly no cellulose fibrils and crystals in the cellulose–solvent mixture from polarized optical microscopy (POM) in Figure 2B. Compared to the POM images of original cotton linters (Figure S3A) and their partial dissolved state (Figure S3B), it suggests the excellent dissolution of cotton linters. It is worthy to notice that the solubility of cotton linters reaches as high as 17% (w/w) by adding only 0.25 wt% urea, showing a remarkable 30.8% increase relative to the original solubility. This result further strengthens the great competitiveness of [TMGH][MAA] as a powerful solvent in comparison to other solvents in terms of cellulose solubility (Table S1). To further discuss the optimization for dissolution of cellulose after adding urea, the rheological behavior of cotton linter solution was conducted. As shown in Figure 2C, the viscosity of cotton linter solution shows a decrease with the increase of shear rate, demonstrating the characteristics of non-Newtonian fluids. It suggests that cotton linters are dissolved effectively into [TMGH][MAA] solvent. After adding 0.25 wt% urea, the viscosity of cotton linter solution decreases obviously, indicating that cellulose chains diffuse better in solution after optimizing the dissolution of cellulose. Furthermore, the loss factor, the ratio of loss modulus to storage modulus (Figure S4), displays an obvious temperature dependency in Figure 2D. With the addition of 0.25 wt% urea, the higher loss factor indicates the more viscous behavior of cotton linter solution,²³ due to less entanglement of cellulose molecular chains in solution. The increased peak temperature of loss factor and substantially decreased storage modulus are also regarded as the criterion that cellulose molecular chains are further disentangled.²⁴ When adding 0.25 wt% urea, the higher peak temperature of the solution is attributed to the increased energy required for stronger intermolecular entangled network that could not be untangled in pristine cotton linter solution at the same temperature. Therefore, as the “hydrogen bond producers,” the addition of urea molecules is a highly effective approach to improve the processability of cellulose solution.

[IMAGE OMITTED. SEE PDF]

As the “hydrogen bond producers,” urea molecules exert a significant influence on the dissolution of cellulose. It is crucial to understand the evolution of hydrogen bonding interaction between solvent system and cellulose, as well as the physicochemical properties of cellulose solution for drawing the mechanism of the enhanced cellulose solubility. Herein, as the repeating unit of cellulose, cellobiose was selected for focusing on the hydroxy groups to study the evolution of hydrogen bonding interaction in cellulose solution.²⁵ The ¹H NMR spectra of cellobiose-[TMGH][MAA] mixture with various concentrations of urea are shown in Figure 3A. The signals of the backbone hydrogen atoms of cellobiose are observed between 2.9 and 4.0 ppm, and nearly no chemical shifts (Δδ < ± 0.01 ppm) are observed. Similarly, there are also almost no chemical shifts (Δδ < ± 0.01 ppm) for the peaks of cellobiose in ¹³C NMR spectra (Figure 3B). These results verify that there is no hydrogen bonding interaction between urea and cellobiose. Furthermore, the broad peak at 6.0–6.5 ppm in ¹H NMR spectra is assigned to the –NH and –OH. It initially shifts upfield and then shifts downfield, suggesting the overall hydrogen bond density decreases first and then increases in the solution system. The most significant shift observed at the urea concentration of 0.25 wt% implies the lowest hydrogen bond density, where the active ions are engaged in hydrogen bonding interaction with cellobiose and have the minimal interference from other free ions and ion pairs. The Arrhenius analysis of viscosity as a function of temperature could provide valuable insights into the molecular motion in solution. As shown in Figure 3C, the variation of viscosity of cellobiose solution as a function of temperature (Figure S5) can be further described using the logarithmic form of Arrhenius equation ²⁶: 1 $$$\begin{equation}\ln \eta = \ln {{\eta }_0} + \frac{{{{E}_\eta }}}{{RT}}\end{equation}$$$where the viscosity value is η, the absolute temperature is T, the activation energy for viscous flow is E_η, gas constant is R, and η₀ represents a constant. A good linear relationship between lnη and $$\frac{1}{T}$$ is depicted in Figure 3C, where the slope of fitted line is regarded as the E_η. As shown in Figure 3D, the E_η initially decreases to the minimum value at the urea concentration of 0.25 wt% and then increases moderately. This suggests that the weakest hydrogen bonding interaction facilitates the mobility of ions and molecules at the urea concentration of 0.25 wt%.²⁷ In addition, the logarithmic form of Arrhenius equation of the temperature-dependent conductivity (σ) in Figure S6 can also be expressed as follows ²⁸: 2 $$$\begin{equation}\ln \sigma = \ln {{\sigma }_0} - \frac{{{{E}_\sigma }}}{{RT}}\end{equation}$$$where E_σ is the activation energy to evaluate the mobility of free ions. The pre-exponential factor (σ₀) is positively related to the concentration of free ions.²⁹ We can discern from Figure 3E that there is a good linear fit between lnσ and $$\frac{1}{T}$$. The slope and intercept of the fitted line can be used to calculate the E_σ and σ₀, respectively. Similar with the trend of E_η, the E_σ reaches the minimum value at the urea concentration of 0.25 wt% (Figure 3F), exhibiting the fastest mobility of free ions. More importantly, the σ₀ also falls to the lowest point at the urea concentration of 0.25 wt% in Figure 3G, implying the least number of free ions. This arises from the maximum hydrogen bonding interaction between urea and free ions. To further directly verify the changes in the number of free ions with various urea concentration, Walden plot of the solution system is carried out, which describes the molar conductivity–viscosity relationship in Figure S7.³⁰ The ideal reference of ionicity is the extrapolation line of the 0.01 M aqueous KCl solution at 25°C, representing the full dissociation of ions. It is found in Figure 3H that the fitted lines with the increase of urea concentration first deviate from the reference line and then approach it at the fixed temperature. At the urea concentration of 0.25 wt%, the fitted line moves the furthest away from the reference line, indicating the lowest ionicity. In a word, the role of urea is gradually transformed from “hydrogen bond producers” to “hydrogen bond breakers” with the increase of the urea concentrations in the solution. When the concentration is below 0.25 wt%, urea molecules produce hydrogen bonds with free ions and ion pairs, leading to fewer hydrogen bonding interaction of free ions or ion pairs with the active ions engaged in hydrogen bond formation with cellobiose. As the urea concentration is beyond 0.25 wt%, urea molecules start to disrupt the original hydrogen bonds between ion pairs, leading to more dissociated ions to compromise the dissolution of cellobiose. Therefore, the urea concentration of 0.25 wt% in our case serves as the transition point for the change in the roles of urea in solution, which is conducive to producing the hydrogen bonds with free ions and ion pairs with maximum efficiency.

[IMAGE OMITTED. SEE PDF]

To further confirm the mechanism of urea-enhanced dissolution of cellulose, molecular dynamics (MDs) simulations were performed to describe the dissolving process of cellulose in such a solvent system. Three concentrations of urea molecules are selected in Figure S8: 0 wt% (pure [TMGH][MAA]), 0.25, and 1.25 wt%. After the 800 ns simulation (Videos S1–S3), the ending conformations of cellulose bunches in solvent system are depicted in Figure S9. The cellulose bunches are gradually swelling, indicating that the crystalline structure of cellulose is broken down. Figure 4A shows the change in the number of hydrogen bonds between urea molecules and cellulose as a function of simulation time. Notably, there are almost no hydrogen bonds between cellulose and urea molecules at the urea concentration of 0.25 wt%. However, a few hydrogen bonds start to be observed when urea concentration increases to 1.25 wt%. To further shed light on the roles of urea in solvent system on the dissolution of cellulose, the hydrogen bond numbers in urea-[TMGH][MAA] and [TMGH]-[MAA] are calculated, as shown in Figure 4B,C. The urea molecules can interact with [TMGH][MAA] by hydrogen bonding, and the density of hydrogen bonds is dependent on urea concentrations. It can be found that the number of hydrogen bonds between [TMGH] and [MAA] at the urea concentration of 0.25 wt% is almost same as that in pure [TMGH][MAA]. This result indicates that urea molecules of 0.25 wt% only form hydrogen bonds with [TMGH][MAA] as “hydrogen bond producers” without breaking their original hydrogen bonds. However, the excessive urea molecules tend to disrupt the hydrogen bonds within [TMGH][MAA] as “hydrogen bond breakers.” Therefore, it is clear from Figure 4D,F that more [TMGH] and [MAA] interact with cellulose by hydrogen bonding at the urea concentration of 0.25 wt% because of fewer interference from part of free ions and ion pairs. Meanwhile, the hydrogen bond number at excessive concentration of urea is comparable with that in pure [TMGH][MAA]. This demonstrates that more dissociated ions interact with the active ions engaged in hydrogen bonding with cellulose, and even excessive urea molecules directly occupy the hydrogen bonding site on cellulose. As a result, fewer hydrogen bonds within cellulose bunches are seen at the urea concentration of 0.25 wt% in Figure 4E,F, exhibiting the enhancement on the dissolution of cellulose. And almost no reduction in the number of hydrogen bonds between cellulose chains is observed with the addition of excessive urea, showing no improvement on the dissolving process of cellulose. The results of MD simulations are highly consistent with the cellulose solubility and the physicochemical properties of [TMGH][MAA] as a function of urea concentration.

[IMAGE OMITTED. SEE PDF]

According to the results and comprehensive analysis in the preceding sections, a schematic diagram is proposed in Figure 5 to illustrate the mechanism of urea-enhanced dissolution of cellulose in [TMGH][MAA] solvent. As for pure [TMGH][MAA] solvent, the protons in [TMGH] cations can attack on the hydroxyl oxygen atoms of C6 in cellulose by hydrogen bonding, whereas the carboxyl oxygen atoms in [MAA] anions can occupy the hydrogen bonding sites of the hydroxyl hydrogen atoms of C2 and C3 in cellulose. This synergistic interaction from [TMGH] and [MAA] breaks the hydrogen bond network among cellulose molecular chains, leading to the dissolution of cellulose. When a low concentration of urea is added, the urea molecules only interact with the free ions and a few ion pairs via hydrogen bonding to stabilize the [TMGH][MAA] solution system. To be specific, the amino nitrogen atoms of urea form hydrogen bonds with the protons of [TMGH] cations, and the amino hydrogen atoms of urea interact with the carbonyl oxygen atoms of [MAA] anions through hydrogen bonding. In this case, urea molecules act as “hydrogen bond producers” in the solvent, reducing the interference on the active ions bonded to cellulose from free ions and ion pairs. Thus, the enhanced dissolution of cellulose is achieved. However, when the excessive urea is added, the urea molecules are required to disrupt the hydrogen bonds between ion pairs, and even between the active ions and cellulose. Consequently, urea molecules, as “hydrogen bond breakers,” release more free ions to break the equilibrium among [TMGH] cations, [MAA] anions, and cellulose, inhibiting the dissolution of cellulose. In a word, it is simple and feasible to unlock the role of “hydrogen bond producers” by controlling the concentration of urea, thereby achieving the optimization of dissolution of cellulose in [TMGH][MAA] solvent.

[IMAGE OMITTED. SEE PDF]

In addition to the enhanced dissolution of cellulose in [TMGH][MAA] solvent, the addition of urea also exerts a positive influence on the regeneration of cellulose from the solution system. To describe the chemical structure of RCF and RCF-U, ATR–FTIR spectra are performed in Figure 6A. The regenerated cellulose films display the absorption bands at ∼3340, ∼2902, ∼1157, and ∼1023 cm⁻¹, corresponding to OH, CH₂, C–O–C, and C–O, respectively. The absorption band of RCF-U assigned to the O–H vibration moves from 3344 cm⁻¹ to a lower wavenumber at 3336 cm⁻¹ compared to RCF, implying the denser hydrogen bond network formed in regenerated cellulose with the addition of urea. TGA curves in Figure 6B illustrate the thermal stability of RCF and RCF-U. Owing to residual water and solvent in the regenerated cellulose, the TGA curves display the drops in weight at low temperature. The region of rapid weight loss of RCF is observed from 295 to 368°C, whereas that of RCF-U appears from 310 to 379°C. A noticeable increased onset decomposition temperature and the temperature of maximum decomposition rate with addition of urea suggest that the addition of urea contributes to the improvement on the thermostability of regenerated cellulose. Wide angle X-ray diffraction (WAXD) analysis in Figure 6C is used to figure out the crystalline structure of regenerated cellulose. The uniform diffraction rings are observed in the inserted 2D-WAXD patterns for the regenerated cellulose films, implying a random distribution of cellulose crystals. It can be found in the intensity profile that both regenerated cellulose films exhibit the characteristic diffraction peaks at 8.4, 13.8, and 15.2 nm⁻¹, assigned to the plane (1 1− 0), (1 1 0), and (0 2 0) of cellulose II, respectively. The peak of RCF-U at 15.2 nm⁻¹ is more pronounced compared to that of RCF, confirming its more complete crystalline structure of plane (0 2 0). Additionally, the crystallinity of RCF-U from the deconvolution of WAXD patterns is higher than that of RCF (Figure S10), being in-line with the TGA. The optimization for hydrogen bond network and crystallinity of cellulose possibly improves the properties of regenerated cellulose films. As shown in Figure 6D, RCF and RCF-U both exhibit high light transmittance above 90% and low transmission haze below 10% in the visible light range of 390–780 nm. The two values greatly demonstrate that the transparency of regenerated cellulose films remains excellent with the addition of urea. The mechanical properties, involving the tensile strength, elongation at break, and Young's modulus, are crucial for determining the service life of cellulose films in practical applications. It can be observed in Figure 6E that the regenerated cellulose films both reveal a typical brittle fracture behavior with an elongation at break from 4% to 6%. As shown in Figure 6F, the tensile strength and Young's modulus of RCF-U are both higher than those of RCF, reaching 124.5 MPa and 6.4 GPa, respectively. This result shows an expected enhancement in the mechanical properties of regenerated cellulose films due to structural optimization during the regeneration process of cellulose. Overall, urea molecules promote the reconstruction of intra-cellulose hydrogen bond network and the crystallization during the regeneration process of cellulose, thereby improving the thermostability and mechanical properties of the regenerated cellulose films. This simple and green method for enhancing cellulose solubility of [TMGH][MAA] solvent is highly promising for advancing the conversion of cellulose into high-performance products.

[IMAGE OMITTED. SEE PDF]

CONCLUSION

In summary, with the purpose of significantly enhancing the dissolution of cellulose in protic ILs, the concept of “hydrogen bond producers” was proposed to manipulate the hydrogen bonding structure within protic ILs for improving the interaction between the active ions and cellulose. In this work, urea molecules were selected as “hydrogen bond producers” to markedly enhance the dissolution effectiveness of cellulose in [TMGH][MAA] solvent system. After an optimization of the physicochemical properties of [TMGH][MAA] as a function of urea concentrations, a significant enhancement in the cellulose solubility was achieved with an increase from 13% (w/w) to 17% (w/w) at 80°C at the urea concentration of only 0.25 wt%, consolidating the great competitiveness of [TMGH][MAA] as a powerful solvent for the dissolution of cellulose. The mechanism of urea-enhanced dissolution of cellulose was that the urea molecules exclusively interacted with ion pairs and part of free ions via hydrogen bonding as “hydrogen bond producers” at a low concentration, reducing the interference with the active ions and thus promoting the interaction between active ions and cellulose. Moreover, the thermostability and mechanical properties of highly transparent regenerated cellulose films were improved with the addition of urea, exhibiting the great implication of “hydrogen bond producers” to the large-scale processing of cellulose products. This work is expected to provide a valuable clue to enhance the efficient dissolution of natural cellulose, which will be beneficial for the extensive utilization of industrial cellulose resources.

EXPERIMENTAL SECTION

Materials and chemicals

Cotton linters (DP = 741) were obtained from Hubei Jinhuan Co., Ltd. Before use, they were completely dried at 60°C under vacuum for 24 h. 1,1,3,3-Tetramethylguanidine, urea, urea-¹⁵N₂, and d-(+)-cellobiose were provided by Aladdin Biochemical Technology Co., Ltd. Methoxyacetic acid and dimethylsulfoxide-d₆ were purchased from Macklin Biochemical Co., Ltd. Chemicals and reagents were used without additional purification.

Dissolution of cellulose in [TMGH][MAA] solvent with addition of urea

The [TMGH][MAA] was obtained by combining TMG and MAA in the molar ratio of 7:3 and stirring for 5 min. Subsequently, urea was added in the [TMGH][MAA] through continued stirring. The dissolution of cotton linters was carried out by adding the cotton linters to 50 g of [TMGH][MAA]/urea solvent and mechanically stirring at 300 rpm at 80°C. The cellulose solubility was evaluated by the continuous addition of cotton linters until the cellulose fibrils and crystals in POM (BX51, Olympus Co.) were visible. The cellulose solubility was defined as the mass ratio (w/w) of the dissolved cotton linters to the [TMGH][MAA]/urea.

Preparation of regenerated cellulose films

The 7% (w/w) cotton linter solutions were chosen to study the impact of urea on the regeneration of cellulose. After centrifugation for 5 min at 8000 rpm, these solutions were poured onto glass substrates with a controlled wet film thickness of 0.80 mm by the mayer-rod. Then wet films were coagulated by soaking in a 20 vol% ethanol/water mixture for 24 h, followed by deionized water for 48 h. Finally, the regenerated cellulose films were obtained by drying at 30°C for 24 h, designated as RCF and RCF-U, respectively, for the pure [TMGH][MAA] and [TMGH][MAA]/0.25 wt% urea.

Nuclear magnetic resonance

The ¹H NMR, ¹³C NMR, and ¹⁵N NMR spectra were performed by NMR (Avance II-400 MHz, Bruker) at 30°C. The samples were fabricated by dissolving cellobiose, [TMGH][MAA] solvent with 0, 0.1, 0.25, 0.5, and 1.0 wt% urea-¹⁵N₂ and [TMGH][MAA]/cellobiose (77 wt%) with 0, 0.1, 0.25, 0.5, and 1.0 wt% urea-¹⁵N₂ into DMSO-d₆, and subsequently transferred to the NMR sample tubes.

Conductivity measurement

The conductivity of solvent system was conducted by conductometer (DDS-11A, Shanghai Leici Chuangyi Instrument Factory) at different temperature and urea concentration. The average values were calculated from at least five samples.

Rheological properties

The rheological properties were measured using a rheometer (MCR320, Anton Paar). Viscosity measurements of solvents were performed from 30 to 100°C at the fixed shear rate of 5 s⁻¹, and the viscosity of cellulose solutions was collected from 0.1 to 20 s⁻¹ at 30°C. The dynamic temperature sweep measurements were conducted at the heating rate of 5°C min⁻¹ with the fixed angular frequency of 1 rad s⁻¹.

Differential scanning calorimetry

DSC measurements were performed by the differential scanning calorimeter (Q2000, TA Instruments). The samples were tested from −35 to 10°C at the heating rate of 5°C min⁻¹ in a nitrogen environment.

Molecular dynamics simulation

The structural model of cellulose was created using experimental crystallographic data³¹ by Cellulose-Builder,³² where the cellulose model consists of 8 cellulose Iβ chains and each chain is composed of 10 glucose units. The [TMGH] cations, [MAA] anions, and urea molecules were treated using the parameters from the generalized AMBER force field, and the GLYCAM_06j-1 force field was used for cellulose.³³ The ion structure of [TMGH][MAA] was optimized, and the atom charges were obtained by Multiwfn alligned with ORCA.³⁴ The charges were determined using the restrained electrostatic potential method.³⁵ The cellulose model was solvated in a cubic box containing 500 ion pairs and certain amounts of urea molecules. All MD simulations were carried out in GROMACS.³⁶ The bulk system along all directions was mimicked on the basis of the periodic boundary conditions. To avoid potential coordinate collisions, the steepest descent method was initially employed until the maximum force between atoms was below 100 kJ mol⁻¹ nm⁻¹. Subsequently, the system was equilibrated through energy minimization using conjugated gradient method.³⁷ Next, the simulation of 10 ns was used under the isothermal-isobaric (NPT) ensemble to equilibrate the solvents. Then, the production run of 800 ns was conducted in the NPT ensemble with restraints removed, utilizing a time step of 2 fs. The particle-mesh Ewald summation was employed to calculate long-range electrostatic interactions with the cutoff radius of 1.2 nm for electrostatics and Van der Waals (VDW) interactions.³⁸ The temperature was held constant at 353 K via velocity rescaling,³⁹ and the pressure coupling was kept at 1 bar by the Parrinello–Rahman algorithm.⁴⁰

Fourier transform infrared spectroscopy characterization

FTIR spectroscopy (Nicolet 6700, ThermoFisher Scientific) was carried out for [TMGH][MAA]/urea solvents and regenerated cellulose films in the reflection-transmission mode and the ATR mode, respectively. The spectra were acquired from 4000 to 700 cm⁻¹, averaging 32 scans at the resolution of 2 cm⁻¹, with background subtraction applied.

Two-dimensional wide angle X-ray diffraction

2D-WAXD patterns were recorded with the X-ray CCD detector (Model Mar165, 2048 × 2048 pixels of 80 µm × 80 µm, Rayonix Co. Ltd.) at the beamline 16B (λ = 1.24 Å), Shanghai Synchrotron Radiation Facility (SSRF). The distance from the sample to the detector was 262.7 mm. The one-dimensional WAXD intensity was derived by integrating from 0 to 360° based on the 2D pattern via the xPolar software. After fitting the peaks, the crystallinity (χ_c) can be calculated by the following equation: 3 $$$\begin{equation}{{\chi }_c} = \frac{{\sum {{A}_{cryst}}}}{{\sum {{A}_{cryst}} + \sum {{A}_{amorp}}}} \times 100\% \end{equation}$$$where the fitted areas of the crystal and amorphous phases are ∑A_cryst and ∑A_amorp, respectively.

Thermal characterization

TGA measurements were performed on a thermogravimetric analyzer (TG209F1, Netzsch Scientific) from 30–800°C at the heating rate of 10°C min⁻¹ under the nitrogen environment.

Optical measurement

Light transmittance was carried out using the UV–vis spectrometer (Lambda 750S, Perkin Elmer) from 200 to 800 nm.

Tensile test

The tensile tests were performed on the universal testing machine (5699, Instron) equipped with the load cell of 1000 N. The rectangular films with the width of 10 mm were measured with the span length of 25 mm at the strain rate of 5 mm min⁻¹, and the mechanical properties were averaged from at least five samples.

ACKNOWLEDGMENTS

This work is financially supported by the National Natural Science Foundation of China (Grant Nos. 51973141, 52033005, and U21A2090), the Science and Technology Department of Sichuan Province (Grant number 2022YFH0094). We are also grateful for the kind help with WAXD measurements in the beamline BL16B of Shanghai Synchrotron Radiation Facility (SSRF, Shanghai, China).

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

References

Sherwood J. The significance of biomass in a circular economy. Bioresource Technol. 2020;300: [eLocator: 122755].

D'Amato D, Droste N, Allen B, et al. Green, circular, bio economy: a comparative analysis of sustainability avenues. J Cleaner Prod. 2017;168:716‐734.

Kuehnel MF, Reisner E. Solar hydrogen generation from lignocellulose. Angew Chem Int Edit. 2018;57(13):3290‐3296.

Wang Y, Wang H, Chen L, et al. Robust ionic liquid/ethanolamine‐superbase solvents enable rapid, efficient and mild dissolution of lignocellulosic biomass. Green Chem. 2023;25(12):4685‐4695.

Han ZM, Sun WB, Yang KP, et al. An all‐natural wood‐inspired aerogel. Angew Chem Int Edit. 2023;62(6): [eLocator: e202211099].

Quintana AA, Sztapka AM, Santos Ebinuma VC, Agatemor C. Enabling sustainable chemistry with ionic liquids and deep eutectic solvents: a fad or the future? Angew Chem Int Edit. 2022;61(37): [eLocator: e202205609].

Akiba T, Tsurumaki A, Ohno H. Induction of lignin solubility for a series of polar ionic liquids by the addition of a small amount of water. Green Chem. 2017;19(9):2260‐2265.

Swatloski RP, Spear SK, Holbrey JD, Rogers RD. Dissolution of cellose with ionic liquids. J Am Chem Soc. 2002;124(18):4974‐4975.

Zhang H, Wu J, Zhang J, He J. 1‐allyl‐3‐methylimidazolium chloride room temperature ionic liquid: a new and powerful nonderivatizing solvent for cellulose. Macromolecules. 2005;38(20):8272‐8277.

Zhang J, Wu J, Yu J, Zhang X, He J, Zhang J. Application of ionic liquids for dissolving cellulose and fabricating cellulose‐based materials: state of the art and future trends. Mater Chem Front. 2017;1(7):1273‐1290.

King AWT, Asikkala J, Mutikainen I, Järvi P, Kilpeläinen I. Distillable acid‐base conjugate ionic liquids for cellulose dissolution and processing. Angew Chem Int Edit. 2011;50(28):6301‐6305.

Michud A, Tanttu M, Asaadi S, et al. Ioncell‐F: ionic liquid‐based cellulosic textile fibers as an alternative to viscose and lyocell. Text Res J. 2015;86(5):543‐552.

Greaves TL, Drummond CJ. Protic ionic liquids: evolving structure‐property relationships and expanding applications. Chem Rev. 2015;115(20):11379‐11448.

Kelley SP, Narita A, Holbrey JD, Green KD, Reichert WM, Rogers RD. Understanding the effects of ionicity in salts, solvates, co‐crystals, ionic co‐crystals, and ionic liquids, rather than nomenclature, is critical to understanding their behavior. Cryst Growth Des. 2013;13(3):965‐975.

Chen S‐P, Zhu J‐L, Chen X‐R, et al. Guanidine‐based protic ionic liquids as highly efficient intermolecular scissors for dissolving natural cellulose. Green Chem. 2023;25(22):9322‐9334.

Gahtori P, Gunwant V, Pandey R. Role of hydrophobic side chain in urea induced protein denaturation at interface. Chem Phys Impact. 2023;7: [eLocator: 100314].

Sagle LB, Zhang Y, Litosh VA, Chen X, Cho Y, Cremer PS. Investigating the hydrogen‐bonding model of urea denaturation. J Am Chem Soc. 2009;131(26):9304‐9310.

Harris KDM. Meldola lecture: understanding the properties of urea and thiourea inclusion compounds. Chem Soc Rev. 1997;26(4):279‐289.

Gund P. Guanidine, trimethylenemethane, and “Y‐delocalization”. Can acyclic compounds have “aromatic” stability. J Chem Educ. 1972;49(2):100‐103.

Grabowski SJ. N–H···O intramolecular hydrogen bonds – the influence of external agents on π‐electron delocalization. J Mol Struct‐Theochem. 2007;811(1–3):61‐67.

Wood PA, Pidcock E, Allen FH. Interaction geometries and energies of hydrogen bonds to C=O and C=S acceptors: a comparative study. Acta Crystallogr B. 2008;64(4):491‐496.

Jena S, Routray C, Dutta J, Biswal HS. Hydrogen bonding directed reversal of ¹³C NMR chemical shielding. Angew Chem Int Edit. 2022;61(41): [eLocator: e202207521].

Li FM, Zheng YY, Wang B. Rheological behaviors of graphene oxide/polyacrylonitrile spinning solutions. Mater Sci Forum. 2017;898:2187‐2196.

Sun J, Song Y, Zheng Q, Tan H, Yu J, Li H. Nonlinear rheological behavior of silica filled solution‐polymerized styrene butadiene rubber. J Polym Sci Pol Phys. 2007;45(18):2594‐2602.

Zhang J, Zhang H, Wu J, Zhang J, He J, Xiang J. NMR spectroscopic studies of cellobiose solvation in EmimAc aimed to understand the dissolution mechanism of cellulose in ionic liquids. Phys Chem Chem Phys. 2010;12(8):1941‐1947.

Wang J, Li Y, Liu H, Tong J. Surface tension, viscosity and electrical conductivity characteristics of new ether‐functionalized ionic liquids. J Mol Liq. 2022;351: [eLocator: 118621].

Ebrahimi M, Moosavi F. The effects of temperature, alkyl chain length, and anion type on thermophysical properties of the imidazolium based amino acid ionic liquids. J Mol Liq. 2018;250:121‐130.

Castro VIB, Mano F, Reis RL, Paiva A, Duarte ARC. Synthesis and physical and thermodynamic properties of lactic acid and malic acid‐based natural deep eutectic solvents. J Chem Eng Data. 2018;63(7):2548‐2556.

Arkhipova EA, Ivanov AS, Levin MM, et al. New asymmetrical morpholinium‐ and 1,1‐dioxidothiomorpholinium‐based dicationic ionic liquid: structure, thermophysical and electrochemical properties of propylene carbonate solutions. J Mol Liq. 2022;346: [eLocator: 117095].

Greaves TL, Drummond CJ. Protic ionic liquids: properties and applications. Chem Rev. 2008;108(1):206‐237.

Nishiyama Y, Langan P, Chanzy H. Crystal structure and hydrogen‐bonding system in cellulose I from synchrotron X‐ray and neutron fiber diffraction. J Am Chem Soc. 2002;124(31):4974‐4975.

Gomes TCF, Skaf MS. Cellulose‐Builder: a toolkit for building crystalline structures of cellulose. J Comput Chem. 2012;33(14):1338‐1346.

Li Y, Liu X, Zhang Y, Jiang K, Wang J, Zhang S. Why only ionic liquids with unsaturated heterocyclic cations can dissolve cellulose: a simulation study. ACS Sustain Chem Eng. 2017;5(4):3417‐3428.

Lu T, Chen F. Multiwfn: a multifunctional wavefunction analyzer. J Comput Chem. 2011;33(5):580‐592.

Bayly CI, Cieplak P, Cornell WD, Kollman PA. A well‐behaved electrostatic potential based method using charge restraints for deriving atomic charges: the RESP model. J Phys Chem. 1993;97(40):10269‐10280.

Hess B, Kutzner C, Spoel D, Lindahl E. GROMACS 4: algorithms for highly efficient, load‐balanced, and scalable molecular simulation. J Chem Theory Comput. 2008;4(3):435‐447.

Payne MC, Teter MP, Allan DC, Arias TA, Joannopoulos JD. Iterative minimization techniques forab initiototal‐energy calculations: molecular dynamics and conjugate gradients. Rev Mod Phys. 1992;64(4):1045‐1097.

Darden T, York D, Pedersen L. Particle mesh Ewald: an N·log(N) method for Ewald sums in large systems. J Chem Phys. 1993;98(12):10089‐10092.

Bussi G, Donadio D, Parrinello M. Canonical sampling through velocity rescaling. J Chem Phys. 2007;126(1): [eLocator: 014101].

Parrinello M, Rahman A. Polymorphic transitions in single crystals: a new molecular dynamics method. J Appl Phys. 1981;52(12):7182‐7190.