Solubility Enhancement of Hydrophobic Compounds in Aqueous Solutions Using Biobased Solvents as Hydrotropes

: Improving the aqueous solubility of poorly soluble hydrophobic compounds is a topic of great interest to the pharmaceutical, chemical, and food industries. The poor solubility of these compounds in water poses a challenge in developing sustainable processes for their extraction, separation, and formulation. Therefore, in this study, the use as hydrotropes of biobased solvents such as γ -valerolactone (GVL), Cyrene, ethyl lactate, and alkanediols (1,2-propanediol, 1,5-pentanediol, and 1,6-hexanediol) to improve the solubility of two model compounds (syringic acid and ferulic acid) in water is investigated. The effects of the concentration and structure of biobased solvents on the solubility of phenolic compounds in aqueous solutions at (303.2 ± 0.5) K were studied. The results showed that the aqueous solubility of the phenolic compounds studied typically increased with the log ( K OW ) of the hydrotrope (1,2-propanediol < Cyrene < 1,5-pentanediol < GVL < ethyl lactate < 1,6-hexanediol) and the hydrophobicity of the solute; the hydrotropic dissolution of phenolic compounds is shown to depend on both the hydrotrope and the solute. This study shows that some biobased solvents, especially GVL, are excellent hydrotropes. Their renewable nature, low price, and low toxicity make these results particularly relevant to the field of extraction and separation of bioactive compounds.


■ INTRODUCTION
In the pharmaceutical, chemical, and food industries, there is strong interest in enhancing the solubility of hydrophobic compounds due to their diverse biological activities and potential health benefits.Phenolic compounds, which possess antimicrobial, anti-inflammatory, and antioxidant properties, are prominent examples of such compounds. 1,2However, the low water solubility of hydrophobic compounds, including phenolic compounds, presents significant challenges in developing sustainable extraction, separation, and application processes.To overcome these limitations, the solubility of hydrophobic compounds in aqueous solutions can be increased by adding hydrotropes or other suitable additives.
Hydrotropes are a class of compounds that enhance the solubility of hydrophobic compounds in aqueous solutions, a phenomenon known as hydrotropy. 3,4−6 The mechanism of hydrotropy involves water-mediated aggregation of hydrotropes around hydrophobic solutes through apolar moieties, effectively minimizing water−solute contacts and, thus, enhancing solute solubility. 3,7,8As such, hydrotropes are typically amphiphilic molecules possessing polar functional groups and reasonably large apolar volumes, which are also common characteristics across many biobased solvents.From a green chemistry perspective, the use of biobased hydrotropes in the processing of hydrophobic compounds is beneficial because it minimizes the use of organic solvents or multistage, energy-intensive extraction schemes.
Recent research indicates that biobased solvents, including glycerol ethers, 7,9,10 alkanediols, 11 dihydrolevoglucosenone (Cyrene), 12−14 γ-valerolactone (GVL), 15−18 and ethyl lactate, 17 can serve as efficient hydrotropes to increase the aqueous solubility of hydrophobic compounds.Moity et al. 10 showed that monoglycerol ethers improve the solubility of hydrophobic dyes.Another study 9 compared the alkyl chain of glycerol ethers and showed that hydrophilic hydrotropes can improve the solubility of phenolic compounds.In a recent study, the alkyl chain and relative position of polar groups in alkanediols (1,2-alkanediols and 1,n-alkanediols) were studied to increase the solubility of syringic acid (by up to 60-fold). 11owever, compared to the study of glycerol ethers, 9 the solubility enhancement, in diluted regions, increases with increasing alkyl chain size of the alkanediols, 11 suggesting different trends using different hydrotrope families.While the relative position of the hydroxyl groups did not affect the solubility in dilute regions, at higher concentrations, 1,2alkanediols exhibit better performance than 1,n-alkanediols. 11e Bruyn et al. 14 demonstrated for the first time the hydrotropic ability of Cyrene, highlighting the chemical equilibrium established between Cyrene and water, yielding a geminal diol, and its ability to function as a reversible and switchable hydrotrope.Abranches et al. 12 further examined the mechanism of hydrotropy in water/Cyrene mixtures, finding the diol and ketone forms of Cyrene, which act as hydrotropes, although in different composition windows, with the diol being most prevalent at low Cyrene concentrations.Interestingly, the ability of Cyrene to increase the aqueous solubility of syringic acid (45-fold) is inferior to that of alkanediols (up to 60fold) 11 or glycerol ethers (up to 77-fold). 9erkel et al. 15 proposed GVL as a promising green solubilizer for pharmaceutical, cosmetic, and agrochemical compounds.Klossek et al. 17 suggested that the degree of selfassociation of biobased solvents has a significant impact on their hydrotropic efficiency.More specifically, the authors found that GVL, as well as ethyl lactate, form clusters in water, in contrast to other solvents such as ethanol, which may explain their high ability to increase the solubility of hydrophobic molecules like Disperse Red 13.
Not only are the biobased solvents discussed above efficient hydrotropes, but they also offer a more environmentally friendly alternative to petrochemical solvents, as they are derived from renewable sources such as wood, starch, vegetable oils, and fruits.They are also less toxic, biodegradable, and biocompatible, with GVL 19 and 1,2propanediol 20 being approved as food additives due to their toxicological safety. 21,22Moreover, they are also economically competitive and already used in certain sectors of the chemical industry.For instance, Cyrene is now being industrially produced by Circa Group with an expected price of $2.5/ kg. 23GVL, which is a stable chemical that does not oxidize or degrade at standard temperatures and pressures, is commonly used as a food additive and occurs naturally in fruits, with an industrial-level price of around $2/kg. 24Ethyl lactate, which has a high boiling point of around 481 K and a low melting point of 242 K, has a cost typically around $2/kg, 17,25 making it competitive with petroleum-derived solvents.Alkanediols such as 1,2-ethanediol, 1,2-propanediol, and 1,6-hexanediol are a subgroup of diols characterized by low toxicity and biodegradability, making them a potential alternative to conventional solvents in various applications.The prices of 2-ethanediol, 1,2-propanediol, and 1,6-hexanediol range usually from $1 to 10/kg, having potential to replace some petroleum-derived solvents.
The aim of this work is to investigate the hydrotropic effect of biobased solvents (1,2-propanediol, 1,5-pentanediol, 1,6hexanediol, Cyrene, ethyl lactate, and GVL) with a broad range of hydrophobicities (log (K OW ) varying from −1.4 to 0) in increasing the solubilization of two model compounds: syringic and ferulic acid.These solutes were selected based on their different hydrophobicities and their importance in natural extracts.Solubility results were interpreted using the Sestchenow equation 9 to better understand how the mechanism of hydrotropy operates in dilute domains, as it is often different from that in concentrated domains.For the concentrated regions, the cooperative hydrotropy model 26 was used to further investigate the mechanism of hydrotropy in the context of these solvents.The observed differences are discussed.
■ EXPERIMENTAL SECTION Chemicals.The chemical compounds used are summarized in Table 1, along with their abbreviation, CAS number, mass purity, and source.The chemical structures of all compounds studied as solutes and hydrotropes are depicted in Figure 1.The water used for all solubility measurements associated with the experiments was double distilled, passed across a reverse osmosis system, and further treated with a Milli-Q plus 185 water purification device.
Solubility Measurements.The solubility of the two phenolic compounds (syringic and ferulic acid) in aqueous solutions of biobased solvents (GVL, Cyrene, 1,2-propanediol, 1,5-pentanediol, and 1,6-hexanediol) was determined using the analytical isothermal shake-flask methodology, previously described in the literature. 9These aqueous solutions of biobased solvents or pure biobased solvents were prepared gravimetrically within ± 10 −4 g using an analytical balance Mettler Toledo Excellence XS205 Dual Range.The solutes, including syringic acid and ferulic acid, were added in excess to a fixed volume of each aqueous solution of biobased solvents or water.The samples were equilibrated in an air oven under constant agitation (1150 rpm) at (303.2 ± 0.5) K and for a minimum of 72 h, using an Eppendorf Thermomixer Comfort equipment.
After equilibrium was reached, the mixtures were placed in an oven at (303.2 ± 0.5) K, for 72 h, in order to separate the macroscopic solid phase from the liquid phase.Then, the liquid phase of each sample was carefully collected and diluted in distilled water.The quantification of ferulic acid and syringic acid was carried out by UV spectrophotometry using a SYNERGY|HT microplate reader, BioTek, at wavelengths of 316 and 266 nm, respectively, using the calibration curves previously established.Blank control samples were made in order to eliminate the interference of aqueous solutions of biobased solvents.For each concentration, three individual

Industrial & Engineering Chemistry Research
samples were prepared to determine the average and standard deviation of the results.

■ RESULTS AND DISCUSSION
Hydrotropic Effect of Biobased Solvents.The solubility of ferulic acid and syringic acid in aqueous solutions of biobased solvents (1,2-propanediol, 1,5-pentanediol and 1,6hexanediol, Cyrene, ethyl lactate, and GVL) was measured in this work.The solubility of these two phenolic compounds was determined over the entire hydrotrope concentration range, i.e., from pure water to pure biobased solvent or its saturated aqueous solutions if they are not fully water-soluble.
The solubilities measured for ferulic acid and syringic acid in pure water at (303.2 ± 0.5) K were (0.83 ± 0.05) g/L and (1.48 ± 0.03) g/L, respectively.The solubilities measured are in good agreement with the data reported in the literature for ferulic acid (0.53−0.92 g/L at 303.2 K) 28,29 and syringic acid (1.43−1.68g/L at 303.2 K). 5,9,12 The obtained hydrotropic solubility curves are depicted in Figure 2 and the detailed experimental data can be found in the Supporting Information (see Tables S1 and S2).The results obtained are shown as solubility enhancement, S/S 0 , where S and S 0 represent the solubility of phenolic compounds in each biobased aqueous solution and in pure water, respectively.
The ability of biobased solvents to enhance the aqueous solubility of ferulic acid and syringic acid is remarkable, as shown in Figure 2. Using a value of 6.0 mol/kg as a reference hydrotrope concentration, the solubility of phenolic compounds can be enhanced by biobased solvents in the following order: 1,2-propanediol < ethyl lactate < 1,5-pentanediol < Cyrene < 1,6-hexanediol < GVL.The solubility curve presents a sigmoidal shape for most of the systems in this study, typical of a hydrotropic mechanism of solvation.Moreover, many solubility curves depicted in Figure 2 pass through a maximum, which is related to the transition from a ″hydrotrope-driven″ to a ″water-driven″ regime and is accompanied by a change in the solvation mechanism. 13Significant enhancements in the solubility of the phenolic compounds were obtained for the aqueous solutions of GVL, leading to increments up to 99-and 237-fold for syringic acid and ferulic acid, respectively, when compared to pure water.Furthermore, to achieve a high enhancement in the solubility of hydrophobic compounds, even those with high hydrophobicity such as ferulic acid (log (K OW ) = 1.51), high concentrations of the hydrotrope were  required.However, our results show that even at low to moderate concentrations of biobased solvents (up to 2 mol/ kg), the solubility of syringic acid increased 14-fold and the solubility of ferulic acid increased 24-fold.Therefore, depending on the applications, low to moderate concentrations of hydrotropes may be sufficient to achieve the desired solubility level.
Since the aqueous solubility of the acidic solutes studied in this work can be affected by the pH of the medium, the starting and final pH of all solutions were assessed to detect interferences caused by changes in solute speciation (see Tables S3−S5 in the Supporting Information).It is important to highlight that no additives were introduced to the solutions in order to change their pH.The pH values obtained reflect the inherent pH of the aqueous solutions of these solvents.All water/solute/hydrotrope mixtures had a lower pH than the respective water/solute mixtures due to the higher solubility of the acid in these mixtures.Moreover, most pH values are lower than the first pK a of ferulic acid (pKa 1 = 3.58) 30 or syringic acid (pKa 1 = 3.93), 30 meaning that most of the solutes are present in their neutral form.This precludes any solubility enhancements due to a higher proportion of their more soluble ionic form.Thus, the improvement in solubility observed is due to a hydrotropic effect and not a pH effect.
When comparing the solubility enhancement of the model phenolic compounds studied achieved with biobased solvents or with other approaches such as using cholinium-based salts, De Faria et al. 31 observed a 1.6-fold increase in the solubility of syringic acid with cholinium chloride, while Yuan et al. 32 obtained a 7.8-fold increase in the solubility of ferulic acid with cholinium-based ionic liquids, particularly cholinium lysinate.In contrast, biobased solvents, including less efficient 1,2propanediol, showed significant improvements, resulting in a 30-fold increase in the solubility of syringic acid and an impressive 97-fold increase in the solubility of ferulic acid.This suggests that biobased solvents are more promising for improving the solubility of poorly soluble compounds in aqueous solutions.
Analysis of Hydrotropy in Diluted Solutions by Setschenow Constants.The solubility results discussed above reveal different trends in the dilute and concentrated regions, emphasizing the need to investigate the hydrotropy mechanisms separately in each region for a precise understanding of the molecular mechanism.This is challenging when considering the multiple interactions that may take place between the three components present in the solution (hydrotrope, water, and solute).All of these interactions may have an impact on the solubility enhancement of the solute.To simplify the analysis, we start by looking into the dilute hydrotrope region, where the interactions between solute− solute and hydrotrope−hydrotrope are expected to be negligible.To study the diluted region, the Setschenow equation was used. 9It quantifies the change in the solubility of a solute due to the presence of a hydrotrope in the dilute region according to eq 1 (1) where S and S 0 represent the solute solubility (mol/L) in the hydrotrope solution and pure water, respectively, C H represents the concentration of the hydrotrope (mol/kg), and K S is the Setschenow constant (kg/mol).This equation is only valid in a concentration region for which the variation of the natural logarithm of solute solubility remains linear.
Although the simplicity of the Setschenow equation may suggest that it is just an empirical relationship, actually it is related to Kirkwood−Buff Integrals (KBIs), through the following expression 33 (2) where G S,H and G S,W quantify, respectively, the excess of the hydrotrope (H) or water (W) around the solute (S), respectively.The KBIs measures the excess of a component present in the local vicinity of another component.Statistical thermodynamics shows that hydrotropic solubilization is mainly driven by solute−hydrotrope preferential interactions relative to solute−water, making G S,H − G S,W > 0. Thus, higher values of K S result from the preference of the solute to interact with the hydrotrope, with a consequent increase in its solubility.
The Setschenow constants for all solute−hydrotrope pairs here studied were calculated, and the results are reported in Table 2 and Figures S1 and S2 in the Supporting Information, along with the molality range used.The values of the constant give information about the efficiency of the hydrotrope in the solubilization of a specific solute.The higher the values of K S , the higher the ability of the hydrotrope to increase the solute solubility in water.
According to the values of K S obtained (Table 2), the hydrotropic power of biobased solvents in general follows the order 1,6-hexanediol > ethyl lactate > GVL > 1,5-pentanediol > Cyrene > 1,2-propanediol, which correlates well with the hydrophobicity of each hydrotrope expressed in terms of its log (K OW ) as shown in Figure 3.Note that, we used here the log (K OW ) as a measure of the solvent hydrophobicity.These results support the hydrotropy mechanism according to which the higher the hydrophobicity of the hydrotrope, the larger the solubility enhancement in the diluted region, 7,11 showing that the hydrophobic part of the hydrotrope is an important factor for hydrotropic efficacy, as demonstrated by Kunz and coworkers. 34This may be explained based on the relationship between the Setschenow constant and the Kirkwood−Buff Integrals represented in eq 2. Since G S,H and G S,W quantify the excess of the hydrotrope or water around the solute, the higher the preference of the solute to interact with the hydrotrope, G S,H , and not with water, G S,W , the higher the Setschenow constant and thus the solubility enhancement of a hydrophobic solute.The behavior in concentrated solutions is somewhat more complex, as will be discussed below.

Analysis of Hydrotropy in Concentrated Solutions by the Cooperative Model.
The cooperative model of hydrotropy 26 is now used to analyze the mechanism of hydrotropy in the region of high concentration.This model is based on statistical thermodynamics and describes the enhanced solubility of a solute (S/S 0 ) as a function of hydrotrope concentration x H , by the following equation (3)   where S represents the molar solubility (mol/L) of the solute in the hydrotropic system, S 0 is the molar solubility (mol/L) of the solute in pure water, (S/S 0 ) max is the maximum attainable relative solubility, and x H is the mole fraction of the hydrotrope in the ternary mixture.Parameters m and b afford an insight into the hydrotropy mechanism and the interactions between the hydrotrope and the solute.Parameter m represents the average number of hydrotrope molecules aggregated around the solute, while b corresponds to the facility to insert that number (m) of hydrotrope molecules in the correspondent volume around the solute.
This model was used to fit the syringic and ferulic acids solubility curves studied in this work.The results presented in Figure 4 demonstrate the ability of the model to describe the solubility curves of these solutes in aqueous solutions of biobased solvents in a wide range of concentrations.The parameters of the linearized equation are presented in the Supporting Information (Figures S3 and S4).The parameter (S/S 0 ) max of eq 3 was left as an adjustable parameter, since most of the solubility curves are not perfectly sigmoidal.Moreover, for solubility curves displaying a clear global maximum, eq 3 was fitted only to hydrotrope concentrations below it.This maximum is due to a change in the solvation mechanism of the solute, with the system moving from hydrotrope-driven solvation in water to water-driven solvation in the hydrotrope.Beyond (S/S 0 ) max , which is reached at high hydrotrope concentrations, the water is no longer the main solvent, and thus, water-mediated hydrotrope−solute interactions are no longer predominant in the system.But this region, albeit interesting, is not the focus of this work and a discussion about the effect of water upon solubility in Cyrene can be found elsewhere. 12he model parameters (m and b) obtained for all solubility curves are presented in Table S6 and correlated against log (K OW ) in Figures 5 and 6.There are clear positive trends between both parameters and log (K OW ) that are similar to those observed in Figure 3 using the Setschenow constant, K S .However, rather than exhibiting a monotonical behavior, m goes through a maximum, as previously observed for alkanediols 11 and glycerol ethers. 7s mentioned above, hydrotropy is driven by the aggregation of hydrotrope molecules around the solute through their apolar moieties.This aggregation is favorable to, and thus mediated by, water (hydrophobic effect), in the sense that minimizing apolar−water contacts enhances water− water and water−hydrotrope contacts through hydrophilic moieties. 7As such, the extent of this aggregation (quantified by parameter m) is expected to correlate positively with the apolar volume of the hydrotrope (here represented in a holistic fashion using K OW ), as the results in Figure 5 demonstrate.However, as the hydrophobicity of the hydrotrope increases, so does the extent of hydrotrope−hydrotrope self-aggregation through apolar moieties.This effect, often known as hydrotrope preaggregation (aggregation without the presence of solute), limits the amount of hydrotropes available to aggregate around the solute.Thus, m displays a maximal value when hydrotrope−hydrotrope aggregation becomes more prevalent

Industrial & Engineering Chemistry Research
than solute−hydrotrope aggregation, as demonstrated in Figure 5 and in previous works. 7,11he behavior of m also demonstrates that the differences in hydrotrope performance observed between low-and highconcentration regions are related to the balance between solute−hydrotrope and hydrotrope−hydrotrope aggregation, with the latter being mostly relevant in the high-concentration region.This effect is particularly relevant when comparing the performances of 1,6-hexanediol (largest apolar volume) and GVL (largest solubility enhancement): 1,6-hexanediol is the best-performing hydrotrope in the low-concentration region because solute−hydrotrope aggregation is maximized, while hydrotrope−hydrotrope aggregation plays a negligible role, and GVL is the best-performing hydrotrope in the highconcentration region given its excellent balance between solute−solute and solute−hydrotrope interactions.
Comparing now parameter m across different solutes (instead of different hydrotropes), ferulic acid, being the most hydrophobic compound studied, is expected to be surrounded by a high number of hydrotropes, and thus, parameter m should be the highest for this system.The results reported in Figure 5 are in agreement with this idea, and parameter m follows the same trend of the hydrophobicity of the solutes: m (ferulic acid) > m (syringic acid).This relation between m and solute log (K OW ) is in agreement with previous observations. 9,12aving discussed the relationship between m and hydrophobicity, it is interesting to note that parameter b displays a good linear correlation with the log (K OW ) of the hydrotropes studied, as depicted in Figure 6.This may be due to the physical meaning of b, which is related to the fugacity of pure hydrotrope and the fugacity of inserting m hydrotrope  & Engineering Chemistry Research molecules in the system.As long as m does not change significantly, b should depend mostly on hydrotrope−water interactions.Figure 6 supports this hypothesis given that (i) log (K OW ), being a surrogate for the extent of hydrotrope− water interactions, correlates well with b, and (ii) both correlations are identical and independent of the solute (i.e., for any given hydrotrope, the values of b exhibited by ferulic and syringic acids are approximately the same).
The interpretation of parameter b, its linear dependence on the hydrophobicity of the hydrotrope, and its independence from the nature of the solute are further supported by the values of b obtained for a series of solutes in Cyrene. 11These displayed reasonably small changes (roughly 30%) across solute log (K OW ) values ranging from 0 to 4, which were attributed to large variations across parameter m.In addition, it is important to note that solubilization in water−Cyrene mixtures is a complex process due to the partial and reversible reaction of Cyrene with water to form its corresponding geminal diol.Although this was neglected in the present work, it was accounted for in the aforementioned correlation of Cyrene-based b parameters.All in all, the relationships identified in Figures 5 and 6 are excellent guidelines for the design of novel hydrotropes targeted at enhancing the aqueous solubility of hydrophobic solutes.

■ CONCLUSIONS
The hydrotropic ability of some biobased solvents to enhance the solubility of model hydrophobic compounds was evaluated here.In general, all solvents presented a significant hydrotropic effect.According to the solubility curves here measured, the best biobased solvent was GVL, showing a 237-fold increase in solubility for ferulic acid and up to a 99-fold increase for syringic acid.
The behaviors and trends observed depended on the concentration and structure of the biobased solvents.The study of the low-concentration region based on the Setschenow equation demonstrated that the hydrotrope hydrophobicity plays the dominant role in the enhanced solubility observed in this region.The high-concentration region was analyzed using the cooperative hydrotropy model that was used to correlate the experimental data measured in this work.The parameters obtained revealed a relationship between parameter m and hydrotrope log (K OW ), which was more complex than that observed in the dilute region, resulting from hydrotrope−hydrotrope interactions that are not negligible at high concentrations.On the other hand, parameter b showed a linear dependence on the hydrophobicity of the hydrotrope, which was surprisingly independent of the type of solute.
For the solutes studied, it was shown that the hydrophobicity of GVL represents the best balance between solute− hydrotrope and hydrotrope−hydrotrope aggregation, with this hydrotrope being the third best performing in the lowconcentration region and the best performing in the highconcentration region.All in all, this work highlights useful relationships between the performance of hydrotropes and the hydrophobicity of both hydrotropes and solutes and reveals the highly attractive potential of biobased solvents as hydrotropes.
■ ASSOCIATED CONTENT * sı Supporting Information

Figure 1 .
Figure 1.Chemical structure of the solutes and hydrotropes used in this work.

Figure 3 .
Figure 3. Representation of K S as a function of the logarithm of the octanol−water partition coefficient of hydrotropes, log (K OW ), for (blue circle) syringic acid and (orange triangle) ferulic acid.

Figure 5 .
Figure 5. Parameter m of the cooperative model of hydrotropy as a function of the logarithm of the partition coefficient between octanol and water of the hydrotropes, log (K OW ), for syringic acid (blue circle) and ferulic acid (orange triangle).Coefficients of determination pertain to the secondorder polynomial regressions for each data set.

Figure 6 .
Figure 6.Parameter b of the cooperative model of hydrotropy as a function of the logarithm of the partition coefficient between octanol and water of the hydrotropes, log (K OW ), for syringic acid (blue circle) and ferulic acid (orange triangle).The coefficient of determination pertains to a single linear regression of both data sets.

Table 1 .
List of Substances Used in the Experimental Work, as well as the Logarithm of the Octanol−Water Partition Coefficients�log (K OW ), CAS Number, Purity, and Source

Table 2 .
Setschenow Constants (K S ) for Syringic and Ferulic Acids in Biobased Solvents, and the Hydrotrope Concentration Range Considered in the Calculations