A salting-out strategy for the efficient separation of an azeotropic mixture of allyl alcohol and water

Abstract

Allyl alcohol and water can form the lowest boiling point azeotrope, but it is very hard to acquire high-purity allyl alcohol by normal distillation methods. Herein, a separation and purification protocol is developed for perfectly separating an azeotropic mixture of allyl alcohol and water using the salting-out method, in which three potassium phosphate salts, K₅P₃O₁₀, K₃PO₄, and K₄P₂O₇, are systematically investigated as salting-out agents, and finally a product consisting of > 99% allyl alcohol is obtained. A thermodynamic study demonstrates that this process involves endothermy and increment entropy. There is a good correlation between the solubility of allyl alcohol and the molar concentration of salt (mol per 1 kg water), and the solubility of the allyl alcohol in the organic phase or the water phase can be forecast using the mass percentage of salts. This work provides a new methodology for the efficient separation of an azeotropic mixture of allyl alcohol and water.

Keywords

allyl alcohol azeotrope salting-out strategy separation water

Introduction

Allyl alcohol contains C=C and C–OH groups, and a series of allyl alcohol derivatives can be easily obtained by substitution, oxidation, addition, rearrangement, and polymerization reactions.^1–4 Allyl alcohol-based chemicals have numerous applications, for example, fragrances, pharmaceuticals, food formulations, lenses, coupling agents, plasticizers, crosslinking agents, and paint additives. The wide application of allyl alcohol has created a huge market demand.^5–7 In industry, the hydrolysis of allyl acetate is often used to prepare allyl alcohol, resulting in a product mixture consisting of allyl alcohol, water, minute quantities of allyl aldehyde, and acetic acid.^8,9 Allyl alcohol (b.p. 96.9 °C) and water have very similar boiling points and can form an allyl alcohol (72 wt%)–water (28 wt%) azeotrope.¹⁰ It is almost impossible to produce allyl alcohol with a purity higher than 72% using conventional distillation methods.^11,12 Special distillation methods are proposed for separating azeotropes, for example, extractive, pressure swing, azeotropic and reactive distillation,^13,14 and distillation itself requires high energy consumption.¹⁵ Consequently, it is not an ideal method to separate allyl alcohol and water.

Previous work has reported the use of vapor–liquid equilibrium (VLE) for the efficient separation of allyl alcohol and water.^11,12,16,17 The VLE data for systems consisting of allyl alcohol + water, allyl alcohol + water + calcium nitrate, allyl alcohol + water + calcium chloride, and allyl alcohol + water + magnesium nitrate were measured at a pressure of 101.3 kPa.¹² Salting-out technologies have many advantages, such as energy saving, simple operation, and no need for extractants.^18–20 Three high-solubility salting-out potassium salt-based agents (K₂CO₃, K₃PO₄, and K₄P₂O₇) showed outstanding salting-out effects in a 1,3-propanediol solution system.^21,22 K₂HPO₄ has been widely studied as a promising salting-out agent, which can be recovered (acetone + butanol + ethanol) from the pre-fractionation tower.^23–25 The salting-out effects of four high-solubility salts (K₄P₂O₇·3H₂O, K₂HPO₄·3H₂O, K₂CO₃, and K₃PO₄·3H₂O) on 2,3-butanediol from various aqueous solutions were studied at 298.15 K.^26,27

In this work, a separation and purification protocol has been developed for the highly efficient separation of allyl alcohol from its azeotrope with water using the salting-out method (Scheme 1). In addition, the salting-out effects of three high-solubility salts (K₅P₃O₁₀, K₃PO₄, and K₄P₂O₇) on the separation of an allyl alcohol from azeotrope (allyl alcohol + water) were investigated in detail. As a result, a product > 99% allyl alcohol was obtained. A thermodynamic study demonstrated that this process involved endothermy and increment entropy. There is a good correlation between the solubility of allyl alcohol and the molar concentration of salt (mol per 1 kg water), and the solubility of the allyl alcohol in the organic phase or the water phase can be forecast using the mass percentage of salts.

Scheme 1.

Process for the separation of allyl alcohol from its azeotrope with water.

Results and discussion

The liquid–liquid equilibria

Allyl alcohol and water are miscible, and no interface was observed. Thus, for their separation, the salting-out technology was applied, and three potassium phosphate salts, K₅P₃O₁₀, K₃PO₄ and K₄P₂O₇, were used to obtain high-purity allyl alcohol. Salts having a mass percentage greater than 10% induced liquid–liquid phase splits owing to the salting-out effect. The salting-out procedure is often interrelated to the cations and anions produced by the dissociation of inorganic salts. Water is largely attracted by anions and cations to form a hydration shell. Large amounts of K₅P₃O₁₀, K₃PO₄, and K₄P₂O₇ were ionized in the aqueous solution to produce cations and anions that repelled allyl alcohol, thereby enriching the organic phase. Hence, this is necessary to increase the mass percentage of salt to enrich the total ions. Although the phase separation was induced by relatively low salt contents, the separation efficiency was not significant, as shown in Figures 1 and 2. The mass percentage of allyl alcohol in the organic phase was relatively low and the mass percentage of water in the organic phase was relatively high. However, enhancing the mass percentage of salts increased the allyl alcohol content in the organic phase and reduced the water content in the organic phase. When the mass percentage of K₄P₂O₇ was increased from 10% to 60%, the mass percentage of allyl alcohol in the organic phase rose from 87.1% to > 99%. When 60% K₅P₃O₁₀, 70% K₃PO₄, and 60% K₄P₂O₇ were added to the allyl alcohol solution, the allyl alcohol solution reached its saturation point. When the mass percentage of salts continued to rise, crystals were observed at the bottom of the vials. The results indicate that K₄P₂O₇ produced the strongest hydrophilic interaction compared with the other salts. This may be because K₄P₂O₇ has a strong hydrophilic effect and hydrolyzable hydroxide ions,²⁸ and thus K₄P₂O₇ shows an excellent salting-out effect for biobutanol.²⁹ In this work, K₅P₃O₁₀, K₃PO₄, and K₄P₂O₇, as salting-out agents, were developed for perfectly separating an azeotropic mixture of allyl alcohol and water through using the salting-out method, and finally a product consisting of > 99% allyl alcohol was obtained.

Figure 1.

Mass fraction of allyl alcohol in the organic phase plotted for different salting-out agents versus the mass fraction of salts at room temperature.

Figure 2.

Mass fraction of water in the organic phase plotted for different salting-out agents versus the mass fraction of salts at room temperature.

The recovery efficiency of allyl alcohol

The ultimate goal of salting-out is to recover allyl alcohol. According to our experiment, the recovery efficiency of allyl alcohol (R) can be calculated as follows

R = \frac{m_{2} w_{22}}{m}

(1)

where m represents the mass of allyl alcohol before the salting-out experiment, m₂ represents the mass of the organic phase, w₂₂ represents the mass percentage of allyl alcohol in the organic phase, and m and m₂ are determined by a gravimetric analysis.

As shown in Figure 3, there were significant improvements in the recovery of allyl alcohol after increasing the mass percentage of K₅P₃O₁₀, K₃PO₄, and K₄P₂O_7. The order of the recovery was as follows: R(K₄P₂O₇) > R(K₃PO₄) > R(K₅P₃O₁₀) at the same salt mass percentage. After the addition of 60% K₄P₂O₇, the recovery of allyl alcohol was 99.7%, which was much higher than the results processed by 60% K₅P₃O₁₀ and 70% K₃PO₄ at room temperature. Thus, K₄P₂O₇ has a better separation effect than the other two salts.

Figure 3.

Recovery of allyl alcohol with different salting-out agents versus the different mass fraction of salts.

The thermodynamic parameters of salting-out

The salting-out temperature of the allyl alcohol system effects the distribution coefficient. The relationship between the distribution coefficient and temperature is as follows

\frac{\partial K}{\partial T} = \frac{Δ H_{0}}{R T^{2}}

(2)

where K represents the distribution coefficient, T represents the absolute scale of temperature, R represents the universal gas constant, and ΔH₀ represents the molar enthalpy change, and

K = \frac{C_{o r g}}{C_{a q}}

(3)

where C_org represents the concentration of allyl alcohol in the organic phase, and C_aq represents the concentration of allyl alcohol in the water phase.

When the type of salt was changed and the other conditions were fixed, various distribution coefficient values were calculated at different temperatures. The relationships between lnK and 1/(R × T) on a salt basis for the water + allyl alcohol system are shown in Figure 4. When the mass percentage of K₅P₃O₁₀ was 20%, then ΔH = 10.3 kJ·mol⁻¹, when the mass percentage of K₃PO₄ was 20%, ΔH = 11.6 kJ·mol⁻¹, and when the mass percentage of K₄P₂O₇ was 20%, ΔH = 15.1 kJ·mol⁻¹.

Figure 4.

Relationship between lnK and 1/(R × T) on a salt-free basis for water + allyl alcohol system with ■ 20% K₅P₃O₁₀, ● 20% K₃PO4, and▲ 20% K₄P₂O₇.

In accordance with equation (2),^30,31 if salting-out is endothermic, then the distribution coefficient K increases along with temperature and vice versa. Thus, the salting-out reactions of K₅P₃O₁₀, K₃PO₄, and K₄P₂O₇ were endothermic. ΔG and ΔS can be calculated using the following two formulae

Δ G = - R T l n K

(4)

Δ G = Δ H - T Δ S

(5)

in which ΔG represents the Gibbs energy change and ΔS represents the entropy change.

As shown in Table 1, as the temperature increased, ΔG decreased and was less than zero; consequently, the salting-out of allyl alcohol was a spontaneous and entropy-increasing process.

Table 1.

The thermodynamic parameters of salting-out at different temperatures.

	K₅P₃O₁₀ (20%, mass fraction)			K₃PO₄ (20%, mass fraction)			K₄P₂O₇ (20%, mass fraction)
Temp (K)	ΔH (kJ mol⁻¹)	ΔG (kJ mol⁻¹·K)	ΔS (J mol⁻¹)	ΔH (kJ mol⁻¹)	ΔG (kJ mol⁻¹·K)	ΔS (J mol⁻¹)	ΔH (kJ mol⁻¹)	ΔG (kJ mol⁻¹·K)	ΔS (J mol⁻¹)
293	10.3	−12.2	76.7	11.6	−12.4	81.7	15.1	−12.6	94.4
303	10.3	−12.6	75.7	11.6	−12.8	80.6	15.1	−13.6	92.9
313	10.3	−13.1	74.7	11.6	−13.3	79.5	15.1	−13.1	91.5
323	10.3	−13.6	73.8	11.6	−13.8	78.5	15.1	−14.0	90.2
333	10.3	−14.0	72.9	11.6	−14.2	77.5	15.1	−14.5	88.9

The solubility correlation

All the experimental results showed that the salting-out efficiency from the allyl alcohol aqueous solution was relevant to the type of salt and the mass percentage of salts. The process decreased the solubility of allyl alcohol in the water phase and enriched the allyl alcohol in the organic phase as the mass percentage of salts increased. The dilute aqueous solution became a supersaturated solution after a large number of water molecules had been captured by salt ions. The Setschenow equation shows that the solubility of organic compounds in salt solution has a good linear relationship with salt concentration.³² Thus, the solubility of allyl alcohol can be correlated with the mass percentage of salts. It was therefore necessary to find a method to evaluate the salting-out effects of K₄P₂O₇, K₃PO₄, and K₅P₃O₁₀ on the salting-out of allyl alcohol + water solutions.

The solubility of allyl alcohol in the water phase calculated in units of g per 100 g water is defined as follows

S_{21} = \frac{w_{21} * 100}{w_{11}}

(6)

in which w₂₁ represents the mass percentage of allyl alcohol in the water phase, and w₁₁ represents the mass percentage of water in the water phase.

The measurement of the salt concentration in the water phase is provided as follows

b = \frac{w_{31} * 1000}{M * w_{11}}

(7)

where b represents the molarity of salts in units of mol per 1 kg water, M represents the molar mass (for K₅P₃O₁₀, M = 448.4 g mol⁻¹, for K₃PO₄, M = 212.3 g mol⁻¹, for K₄P₂O₇, M = 330.3 g mol⁻¹), w₃₁ represents the mass percentage of salts in the water phase, and w₁₁ represents the mass percentage of water in the water phase.

To fully understand the salting-out process of allyl alcohol, the relationship between the natural logarithm of the solubility of allyl alcohol in the aqueous phase and the molar concentration of salt was plotted (Figure 5). The two factors showed a linear relationship. Thus, it was concluded that the solubility of allyl alcohol in the water phase was greatly influenced via the molality of the salt. The solubility of allyl alcohol in the water phase decreased as the molality of the salt increased. However, in the cases of K₄P₂O₇, K₃PO₄, and K₅P₃O₁₀, the solubility of allyl alcohol changed greatly because different anions (P₃O₁₀⁵⁻, PO₄³⁻, and P₂O₇⁴⁻, respectively) had obvious negative effects on the salting-out of alcohol.³³ The fitting empirical constants of the linear regression are displayed in Table 2.

Figure 5.

Plot of ln(S₂₁) against the molality of the salt.

Table 2.

The constants α and β and the coefficient of determination, R².

Salt	α	β	R ²
K₅P₃O₁₀	−0.3582	0.6395	0.9983
K₃PO₄	−0.3112	0.33323	0.9986
K₄P₂O₇	−0.3121	0.1516	0.9944

The linear regression of the data obtained from a particular allyl alcohol aqueous solution and a particular salting-out agent (K₄P₂O₇, K₃PO₄, or K₅P₃O₁₀) was determined using the following equation

\ln S_{21} = α b + β

(8)

The slope of α(K₄P₂O₇) < α(K₃PO₄) < α(K₅P₃O₁₀) which indicated that the solubility reduction of allyl alcohol in the water phase caused via 1 mole of K₄P₂O₇ was greater than that caused by 1 mole of K₃PO₄ or K₅P₃O₁₀. Therefore, the salting-out effects of the three salting-out agents on an allyl alcohol–water solution were as follows: K₄P₂O₇ > K₃PO₄ > K₅P₃O₁₀, which was also verified from the recovery ratios of allyl alcohol. Improving the amount of salt strengthened the hydration power of the salt in the allyl alcohol aqueous solution, with more water molecules being attracted from the organic phase to the water phase. Therefore, the solubility of allyl alcohol in the water phase was significantly reduced, and the solubility in the organic phase was significantly increased. The solubility of allyl alcohol in the organic phase is shown below

S_{22} = \frac{w_{22} * 100}{w_{12}}

(9)

The solubility of allyl alcohol in the organic phase increased and the solubility of allyl alcohol in the water phase decreased as the mass percentage of salt increased. The ln(S) value and mass percentage of salts showed a good linear relationship. The constants α and β and the correlation coefficient, R² values, are shown in Table 3.

Table 3.

The constants α and β and the coefficient of determination, R².

	Salt	α	β	R ²
Organic phase	K₅P₃O₁₀	0.0498	5.5999	0.9907
	K₃PO₄	0.0660	5.6288	0.9932
	K₄P₂O₇	0.0839	5.6866	0.9924
Aqueous phase	K₅P₃O₁₀	−0.0224	0.3734	0.9991
	K₃PO₄	−0.0221	0.0888	0.9939
	K₄P₂O₇	−0.0233	0.0109	0.9988

Thus, the solubility of allyl alcohol in the organic phase or the water phase can be projected by the mass percentage of salts, as follows

\ln S = α w + β

(10)

Once the mass percentage of a specific salt is known, the solubility of allyl alcohol in both phases can be calculated.

The K₄P₂O₇ residue in the organic phase

The mass percentage of salts in the organic phase is a key factor for assessing the energy-saving efficiency. If a salt dissolves in the organic phase, then it leads to the challenge of salt separation. As shown in Figure 6, the K₄P₂O₇ content in the organic phase decreased when more K₄P₂O₇ was added into the (allyl alcohol + water + salt) system. In addition, the mass percentage of 60% K₄P₂O₇ in the organic phase was less than 120 ppm, which may be because the high mass percentage of allyl alcohol repelled the ions. The residue of K₄P₂O₇ in the organic phase (less than 120 ppm) did not necessitate subsequent distillation. The trace of salt in the organic phase will not negatively affect the industrial applications of allyl alcohol.

Figure 6.

Mass fraction of K₄P₂O₇ in the organic phase plotted against different mass fractions of salts.

Conclusion

A separation and purification protocol has been developed for the efficient separation of the azeotropic mixture of allyl alcohol + water using the salting-out method and a greater than 99% allyl alcohol content was obtained after the salting-out process. The recovery ratio of allyl alcohol using the three tested potassium phosphate salts can be ordered as follows: K₅P₃O₁₀ < K₃PO₄ < K₄P₂O₇ at the same salt mass percentage. Hence, K₄P₂O₇ is the best salt to separate allyl alcohol compared with K₃PO₄ and K₅P₃O₁₀. When the mass percentage of K₄P₂O₇ was 60%, allyl alcohol was separated with the purity of 99%. A thermodynamic study demonstrated that the process involved endothermy and increment entropy. A new equation for predicting the solubility of the allyl alcohol in both phases using the mass percentage of salts has been constructed. This work provides a new methodology for the efficient separation of an azeotropic mixture of allyl alcohol + water.

Experimental section

Materials

The specific information for the reagents, including the supplier, purity, and analysis approach for purity, is given in Table 4. Potassium pyrophosphate, potassium tripolyphosphate, and potassium phosphate were vacuum-desiccated for at least 24 h.

Table 4.

Specification of the chemicals.

Name	Source	CAS no.	Mass fraction purity	Purification method	Final water mass fraction	Analysis approach
Allyl alcohol	Xiya Reagent Co., Ltd.	107-18-6	0.99	None	-	HPLC^a
Potassium pyrophosphate	Aladdin Reagent Co., Ltd.	7320-34-5	⩾0.98	Vacuum desiccation	0.0006	KF^b
Potassium tripolyphosphate	McLean Reagent Co., Ltd.	13845-36-8	⩾0.95	Vacuum desiccation	0.0007	KF^b
Potassium phosphate	Sigma-Aldrich Reagent Co., Ltd.	7778-53-2	⩾0.98	Vacuum desiccation	0.0006	KF^b

High-performance liquid chromatography.

Karl Fischer titration.

Salting-out experiments

The salting-out experiments were performed in gas airtight vials with sealing lids at 25 °C. Each vial was charged with 72 wt% allyl alcohol aqueous solution (10 mL) and a quantitative amount of salting-out agent, and then shaken at 25 °C at 200 rpm for 2 h, mixtures were allowed to stand for 24 h to reach complete phase separation. All the salting-out experiments were repeated two times.

Analytical approaches

The content of allyl alcohol in the salting-out experiment was ascertained through high-performance liquid chromatography (HPLC, Shimadzu, LC-20AD), and analytical methods were performed as described in the literature.³⁴ The salt ion contents were determined through inductively coupled plasma atomic emission spectrometry (Agilent 725, ICP-AES). The water content was determined through Karl Fischer titration (831 KF coulometer, Metrohm, Switzerland).

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this paper.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this paper: This work was jointly funded by the National Natural Science Foundation of China (Grant No. 52203110), the Doctoral Research Initiation Fund of Shanxi University of Chinese Medicine (Grant Nos. 2020BKS05 and 2020BK14), the Shanxi Provincial Department of Education Fund Project (Grant No. 2021L365), the Shanxi Provincial Science and Technology Department Fund Project (Grant No. 202203021212338), the Shanxi Provincial Science and Technology Innovation Talent Team (Grant No. 2022040510010), the Scientific and Technological Innovation Team of Shanxi University of Chinese Medicine (Grant No. 2022TD2010), and the Leading Team of Medical Science and Technology, Shanxi Province (Grant No. 2020TD05).

ORCID iD

Guangfu Liao

References

Wilson

Knecht

Garcia-Martinez

, et al. J Am Chem Soc 2006; 128: 4510–4511.

Nguyen

TTN

Huchede

Blanco

, et al. Appl Catal A: Gen 2018; 549: 170–178.

Long

Xia

Wang

, et al. ACS Catal 2021; 11: 13880–13890.

Syed

Hydrometallurgy 2012; 115-116: 30–51.

Sánchez

Friggieri

Adesina

, et al. Catal Sci Technol 2014; 4: 3090–3098.

Luque

Budarin

Clark

, et al. Appl Catal B: Environ 2008; 82: 157–162.

Malapit

Prater

Cabrera-Pardo

, et al. Chem Rev 2022; 122: 3180–3218.

Atkins

Walker

Manyik

RM.

Tetrahedron Lett 1970; 11: 3821–3824.

Al-Haimi

AANM

Wang

, et al. J Mol Liquids 2021; 336: 116875.

10.

Gao

Klunder

Hanson

, et al. J Am Chem Soc 1987; 109: 5765–5780.

11.

Shi

, et al. J Chem Thermodyn 2018; 118: 139–146.

12.

Shi

, et al. Fluid Phase Equilibria 2018; 457: 11–17.

13.

Malone

Doherty

MF.

Ind Eng Chem Res 2000; 39: 3953–3957.

14.

Wang

Zhuang

Liu

, et al. Sep Purif Technol 2021; 259: 118211.

15.

Liu

Luo

Lei

, et al. Sep Purif Technol 2022; 280: 119808.

16.

Jiang

Diao

, et al. J Mol Liquids 2019; 279: 524–529.

17.

Gao

, et al. J Chem Eng Data 2018; 63: 845–852.

18.

Dai

Guan

, et al. Sep Purif Technol 2017; 184: 275–279.

19.

Santos

d’Ávila

Aznar

Fluid Phase Equilibria 2001; 187-188: 265–274.

20.

Sang

S-H

Zhang

H-Z

Jiang

P-H

, et al. J Chem Eng Data 2022; 67: 500–509.

21.

Xie

Qiu

, et al. J Chem Technol Biotechnol 2016; 91: 2793–2801.

22.

Yagoub

AEA

, et al. J Mol Liquids 2021; 325: 115158.

23.

Xie

Qiu

J Chem Eng Data 2014; 59: 1507–1514.

24.

Shi

Yang

Chu

, et al. AIChE J 2022; 68: e17814.

25.

Zhang

Sang

, et al. Hydrometallurgy 2023; 216: 106017.

26.

Xie

Zhang

Zhou

, et al. J Chem Technol Biotechnol 2017; 92: 122–132.

27.

Liu

Zhang

, et al. J Chem Res 2022; 46: 17475198221117409.

28.

Jia

, et al. Energy Conversion Management: X 2021; 10: 100059.

29.

Xie

Qiu

J Chem Technol Biotechnol 2016; 91: 1860–1867.

30.

Liao

Zhong

Cheung

, et al. Mesopor Mater 2020; 307: 110469.

31.

Liao

Zhao

, et al. Appl Catal A: Gen 2018; 549: 102–111.

32.

Schlautman

Yim

Carraway

, et al. Water Res 2004; 38: 3331–3339.

33.

Villegas

Weaver

MJ.

J Phys Chem 1996; 100: 19502–19511.

34.

Jia

Zhao

Xue

, et al. J Chem Eng Data 2018; 63: 3874–3880.

A salting-out strategy for the efficient separation of an azeotropic mixture of allyl alcohol and water

Abstract

Keywords

Introduction

Results and discussion

The liquid–liquid equilibria

The recovery efficiency of allyl alcohol

The thermodynamic parameters of salting-out

The solubility correlation

The K4P2O7 residue in the organic phase

Conclusion

Experimental section

Materials

Salting-out experiments

Analytical approaches

Footnotes

Declaration of conflicting interests

Funding

ORCID iD

References

The K₄P₂O₇ residue in the organic phase