Phenylethyl alcohol recovery from aqueous solutions

Phenylethyl alcohol (PEA), a valuable compound widely used in various fields, is the main odorant component of rose. In this study, PEA adsorption from aqueous solutions was carried out using Purolite MN202 adsorbent. To evaluate the performance of the adsorbent in PEA adsorption, both batch and dynamic adsorption experiments were conducted. In the batch adsorption experiments, the effects of initial PEA concentration (250–6000 mgL-1), adsorbent amount (0.025–1 g), and temperature (25–40°C) on the adsorption capacity were investigated. To determine the adsorption mechanism and its degree, the batch experimental data obtained at 30°C were analyzed using Langmuir, Freundlich, Temkin, and Dubinin–Radushkevich (D-R) isotherm models. The Langmuir model yielded a maximum adsorption capacity (qm) of 1250 mgL-1. The regression coefficients (R²) of the isotherm models for Langmuir, Freundlich, Temkin and D-R were found in the range of 0.93±0.010, 0.93±0.014, 0.95±0.012 and 0.96±0.011 respectively. Both the Langmuir and Freundlich models, along with adsorption–desorption cycles, confirmed that the adsorption mechanism was physical. Additionally, the Temkin model indicated that the adsorption process was favorable. Moreover, the mean free energy calculated from the D-R model was 8.2249±0.2033 kJmol-1, indicating that the mechanism is related to reversible ion exchange. The kinetics of the adsorption process were studied using pseudo-first-order (PFO), pseudo-second-order (PSO), and intraparticle diffusion kinetic models. The regression coefficients (R²) for PFO, PSO, and intraparticle diffusion models were 0.86±0.020, 0.97±0.0002, and 0.97±0.0010, respectively. The equilibrium adsorption capacity (qe) was calculated as 938.68±99.11 mgg-1 and 833.33 mgg-1 for PFO and PSO models, respectively. The rate constants for PFO (k1), PSO (k2) and intraparticle diffusion models (ki) were obtained as 0.06 min-1, 7.16×10-5±3.56×10-7 gmg-1min-1 and 56.67±0.15 mgg-1min-0.5, respectively. When the adsorption process was examined from a thermodynamic perspective; it was seen that the Gibbs free energy was negative (ΔG°<0). This finding showed that the process was spontaneous. The negative heat of adsorption indicated that the adsorption was exothermic, and the negative standard entropy change indicated that less disorder occurred at the solid-liquid interface. In the dynamic adsorption experiments, breakthrough curves were obtained by varying the initial PEA concentration (250–500 mgL-1), adsorbent mass (1–2 g), and flow rate (11–18.8 mLmin-1). The time to reach 50 % breakthrough decreased with increasing initial concentration and flow rate, while it increased with a higher amount of adsorbent. Additional dynamic adsorption experiments were performed using distillation water obtained from a rose oil factory's veronica tank. It was observed that increasing the flow rate shortened the time to reach 50 % breakthrough in these tests as well. The maximum column adsorption capacities (qₜₖ) were found in the range between 765.42 and 6458.51 mgg-1, the equilibrium column adsorption capacity (qₑₖ) in the range between 764.42 and 3225.87 mgg-1 and the removal efficiency in the range between 22.68 % and 61.80 %. Dynamic adsorption data were modelled using Yoon–Nelson (YNM), Thomas (TM), and Clark (CM) models. The regression coefficients for YNM, TM, and CM ranged from 0.86 to 0.97, 0.81 to 0.97 and 0.83 to 0.97, respectively. The time required for 50 % breakthrough (τ), the Yoon–Nelson rate constant (kYN) and the YNM column adsorption capacity (qYN) were calculated in the range between 146.97 and 2293.67 min, 0.0012 and 0.0036 min-1, and 439.80 and 3150.48 mgg-1, respectively. The TM constant as kTM and maximum adsorption capacity qTM were determined 0.006–0.24 mLmg-1min-1 and 460.83–3323.26 mgg-1, respectively. For the CM constants A and the mass transfer rate r were found in the range between 2.68–75.061 and 0.0014–0.0055 min-1, respectively. Among these, the YNM demonstrated the best performance in fitting the dynamic adsorption data. It was concluded that the adsorbent is suitable for recovering PEA from aqueous solutions.