Computer-aided simulation and exergy analysis of TiO₂ nanoparticles production via green chemistry

Exergy analysis of chemical processes

Exergy analysis is an important tool which can be used as an instrument to evaluate existing and emerging processing pathways from energy and thermodynamic viewpoints (Singh et al., 2019). Hamdy et al. (2019) applied an exergy-based method to quantify the effect of cold storage on thermodynamic performance of six liquefaction processes and identified the most cost-efficient process. In this study, exergy analysis was used as a decision-making tool for the selection of the best alternative for the liquefaction process. Zoder et al. (2018) developed a simulation and exergy analysis of energy conversion processes applied to gas and steam turbine cycles. The authors introduced a novel approach which combines process simulation, exergy analysis and object-oriented programming of a combined cycle, gas turbine system. In this case, exergy sensitivity analysis allowed identification of thermodynamic inefficiencies. Peters, Petrakopoulou & Dufour (2015) performed an exergy analysis to assess synthetic biofuel production via fast pyrolysis and hydrodeoxygenation upgrading. In such work, process simulation provided operating information to conduct exergy balances for each plant section and exergy inefficiencies were determined. These performance parameters revealed the potential for improvement of the process. Koç, Yağlı& Koç (2019) developed an exergy analysis of a subcritical/supercritical organic Rankine cycle for a gas heat recovery system. Exergy analysis was performed based on mass, energy and general exergy balance according to the net power production and energy/exergy inlet flows. Peralta-Ruiz, González-Delgado & Kafarov (2013) applied exergy analysis for screening process alternatives in microalgae oil extraction. The authors considered exergy analysis as a selection criteria for novel/emerging processing routes under sustainability targets. In the study, processing topologies were simulated using Aspen Plus^® software, and the exergy analyses were developed based on the extended mass and energy balances obtained from process simulation. Meramo-Hurtado, Ojeda-Delgado & Sánchez-Tuirán (2018) analyzed a bioethanol production plant design under exergy parameters considering rice residues as the main feedstock. Results showed that acid pretreatment was the stage with the lowest exergy efficiency. Bahadori & Nalbland Oshnuie (2019) developed an exergy analysis of indirect dimethyl production process to find the stages with low exergetic efficiencies. All of these contributions to development of the exergy concept show the relevance of the application of exergy analysis as a useful process evaluation tool.

Process simulation

The plant size was fixed according to the availability of raw materials, mainly, lemongrass. After selecting the production capacity of the plant, process simulation was performed based on the following steps:

I. Select chemical components of the process and operations.

II. Choose a proper thermodynamic model and state equation.

III. Set processing capacity.

IV. Set up input parameters as mass/energy flow rates, temperature, pressure, and stoichiometry of the reactions (Xuan, Lim & Yeo, 2014).

The commercial simulation software Aspen Plus^® was used to model and simulate the large-scale production process of TiO₂ nanoparticles. Process boundaries were defined by the highlighted sections in red and green shown in Fig. 2. Several contributions use this tool to model existing and emerging technologies for chemical processing, which derived from a wide variety of feedstocks (e.g., biomass), processing routes and products (Meramo-Hurtado, Ojeda-Delgado & Sanchez-Tuiran, 2018; Ojeda et al., 2011b; Luo, Van der Voet & Huppes, 2010). The software Aspen Plus^® is characterized by an extensive, flexible and trusted property database containing a vast property collection of many compounds. Hence, most compounds involved in TiO₂ nanoparticle synthesis were available in the database. Components unavailable in the software were created using the Molecule Editor in Aspen Plus^® and the physicochemical properties reported in the literature (Wooley & Putsche, 1996). This software also allows one to introduce many known constant properties and temperature-related parameters such as normal boiling point, Gibbs free energy of formation, critical properties, and acentric factors. In this work, it was necessary to create a user-defined component such as titanium isopropoxide. The Non-Random Two Liquids (NRTL) solution model was chosen as the thermodynamic-based model for process simulation because of its accuracy to predict properties of polar/non-polar mixtures.

Process description

The large-scale process was designed for a capacity of 5,724.16 t/y of TiO₂ nanoparticles because of the availability of raw materials. In Colombia, lemongrass production is the limiting factor on TiO₂ nanoparticles synthesis with a national production above 30,000 t/y over the two last decades (Espinal et al., 2000). The extract of lemongrass (lemongrass oil) was used as a surfactant to ensure nanosized particles. Titanium isopropoxide (TTIP) was employed as a precursor of titanium in the photocatalytic product. The green synthesis process was divided into two main sections as shown in Fig. 2:

1. Lemongrass oil extraction.

2. Synthesis of TiO₂ nanoparticles.

In the first section, lemongrass is pretreated to remove cellulosic material that may reduce the oil extraction yield. Then, dried lemongrass is cooled to room temperature (28 °C) and sent into a crushing unit for particle size reduction. An infusion-evaporation unit was used as a solid–liquid extraction method to collect oil extract with low moisture content. The oil extract is mainly composed of myrcene, neral, geraniol, citral and nerol, with a total oil composition of 1.10 wt% (Meramo-Hurtado et al., 2018). This process stream contains 7,284.31 t/y of phytochemicals (myrcene, neral, citral, among others) and an oil composition of 4.2 wt%. The second section involves a hydrolysis reaction to form nanoparticles. For simulation purposes, kinetics are significant parameters for reactor modeling (Sun et al., 2019). The general reaction mechanisms of the sol–gel synthesis of TiO₂ from titanium alkoxide are shown as follows: (1)${\displaystyle \mathrm{Ti}{\left(\mathrm{OR}\right)}_4+4{\mathrm{H}}_2\mathrm{O}\to \mathrm{Ti}{\left(\mathrm{OH}\right)}_4+4\mathrm{ROH}}$ (2)${\displaystyle \mathrm{Ti}{\left(\mathrm{OH}\right)}_4\to \mathrm{Ti}{\mathrm{O}}_2+2{\mathrm{H}}_2\mathrm{O}}$

In the presence of water, titanium alkoxide is hydrolyzed (hydrolysis reaction) and then, condensation takes place to form an oxide network (Mahshid, Askari & Ghamsari, 2007; Gupta & Tripathi, 2012). The Arrhenius rate constants for TiO₂ nanoparticle synthesis were reported by Oskam et al. (2003) with values ranged 10⁻²-10⁻³ nm³s⁻¹ for 1/temperature between 0.002–0.0024 K⁻¹. The overall reaction using TTIP as precursor is given by Nour, Sulaiman & Nour (2014): (3)${\displaystyle \mathrm{Ti}{\left(\mathrm{O}{\mathrm{C}}_3{\mathrm{H}}_7\right)}_4+2{\mathrm{H}}_2\mathrm{O}\to \mathrm{Ti}{\mathrm{O}}_2+4{\mathrm{C}}_3{\mathrm{H}}_7\mathrm{OH}.}$

After mixing the titanium precursor with water, lemongrass oil was added into the hydrolysis reactor. At the lab-scale, previous authors reported a conversion yield of 0.93 mol TTIP/mol TiO₂ (Nour, Sulaiman & Nour, 2014). The resulting nanoparticles were sent to repetitive cycles of centrifugation and washing to remove the residual water and ethanol. Finally, moisture content of high purity nanoparticles was reduced in the calcination stage.

Mathematical formulation for exergy analysis

The exergy analysis provides key indicators to improve process performance such as exergy loss or irreversibilities, percentage of exergy loss and exergy efficiency, which can be calculated per stage or around the entire system. Assumptions reported by Moreno-Sader, Meramo-Hurtado & González-Delgado (2019) were considered in this work:

• The whole process was assumed to be steady state.

• Kinetic exergy and potential exergy were neglected.

• The temperature of reference was taken as 298 K.

The performance indicators from exergy analysis are functions of variables such as exergy of work, exergy due to heat transfer or exergy of the process stream. The calculations for exergy analysis are based on the research conducted by Yan et al. (2019), Yang et al. (2019b) and Yang et al. (2019a). For a better understanding of this section, a systematic procedure for exergy-based indicators is presented in Fig. 3.

As a first approach, the composition of process streams was needed to identify the main properties of components. The following equation was used to calculate physical exergy of process streams, where H and S, are the stream enthalpy and entropy, respectively. (4)${\displaystyle {\mathrm{Ex}}_{\mathrm{p}}=\mathrm{H}-{\mathrm{H}}_{\mathrm{o}}-{\mathrm{T}}_0\left(\mathrm{S}-{\mathrm{S}}_0\right)}$

The second step is to calculate the specific chemical exergies of compounds, which are usually reported in the literature. This parameter is defined as the work that can be obtained when a substance reaches thermodynamic equilibrium through chemical reactions and is given by Eq. (5). (5)${\displaystyle {\mathrm{Ex}}_{\mathrm{che},\mathrm{i}}=\Delta {\mathrm{G}}_{\mathrm{f}}^0+\sum _{\mathrm{j}}{\mathrm{v}}_{\mathrm{j}}{\mathrm{Ex}}_{\mathrm{Ch}-\mathrm{j}}^0}$

where $\Delta {\mathrm{G}}_{\mathrm{f}}^0$ is the standard free energy of formation, ${\mathrm{Ex}}_{\mathrm{Ch}-\mathrm{j}}^0$ is the standard chemical exergy for element j, v_j is the number of atoms of j. The chemical exergy of process stream, i.e., a mixture (Ex_che, depends on the chemical exergy of its components and is given by Eq. (6). (6)${\displaystyle {\mathrm{Ex}}_{\mathrm{che}}=\sum {\mathrm{y}}_{\mathrm{i}}\ast {\mathrm{Ex}}_{\mathrm{ch}-\mathrm{i}}+\mathrm{R}{\mathrm{T}}_0\sum {\mathrm{y}}_{\mathrm{i}}\ast ln{\mathrm{y}}_{\mathrm{i}}}$

where R is the universal gas constant and y_i is the mole fraction of component i.

The following step is to estimate the exergy of mass flow by Eq. (7), as the sum of chemical exergy, physical exergy, potential exergy and kinetic exergy. (7)${\displaystyle \mathrm{E}{\mathrm{x}}_{\text{mass}}={\mathrm{Ex}}_{\mathrm{che}}+{\mathrm{Ex}}_{\mathrm{p}}+{\mathrm{Ex}}_{\text{potential}}+{\mathrm{Ex}}_{\mathrm{kin}}}$

Kinetic and potential exergies are neglected in most cases because of their low values compared to physical and chemical exergies; hence, Eq. (7) can be simplified into Eq. (8). (8)${\displaystyle {\mathrm{Ex}}_{\text{mass}}={\mathrm{Ex}}_{\mathrm{p}}+{\mathrm{Ex}}_{\mathrm{che}}}$

The exergy of utilities involves both exergy of work (Ex_work) and exergy of heat stream (Ex_Q) that are calculated through Eq. (9) and 10, respectively. Exergy for work is equal to work (W), as long as there are no changes in volume. Exergy by heat transfer depends on the temperature of the stream and the environmental temperature (taken as 298 K and 1 atm). (9)${\displaystyle {\mathrm{Ex}}_{\text{work}}=\dot{W}}$ (10)${\displaystyle {\mathrm{Ex}}_{\mathrm{Q}}=\left(1-\frac{T_o}{T}\right)\dot{Q}}$

where Q is the heat transfer rate, T is the system temperature, and T_o is the environmental temperature. Then the total exergy (Ex_total−in entering into the system is estimated from the results of exergy of streams and industrial services (Ojeda et al., 2011a). This parameter is given by Eq. (11). (11)${\displaystyle {\mathrm{Ex}}_{\text{total}-\mathrm{in}}=\sum {\mathrm{Ex}}_{\text{mass}-\mathrm{in}}+\sum {\mathrm{Ex}}_{\text{utilities}-\mathrm{in}}}$

Where ∑Ex_mass−in is the exergy of all streams entering the system and ∑Ex_{utilities−in} is the total exergy of the utilities.

Exergy can leave a system by products’ and residues’ streams. This parameter is estimated by Eq. (12), where Ex_total−out is the total output exergy flow, ∑Ex_{products−out} is the total exergy of product streams, and ∑Ex_{residues−out} is the total exergy flow by the process wastes. (12)${\displaystyle {\mathrm{Ex}}_{\text{total}-\mathrm{out}}=\sum {\mathrm{Ex}}_{\text{products}-\mathrm{out}}+\sum {\mathrm{Ex}}_{\text{residues}-\mathrm{out}}}$

Process irreversibilities Ex_loss) is the exergy loss, i.e., the work not used through the process, and it is given by Eq. (13). (13)${\displaystyle {\mathrm{Ex}}_{loss}=\sum {\mathrm{Ex}}_{\text{total}-\mathrm{in}}+\sum {\mathrm{Ex}}_{\text{products}-\mathrm{out}}}$

To identify the sources of highest losses within the process, the percentage of exergy loss (% Ex_loss) in process stage k is calculated by Eq. (14). (14)${\displaystyle \text{\% Ex}{}_{\text{loss}}=\left(\frac{\mathrm{Ex}{}_{\text{loss},\mathrm{k}}}{\mathrm{Ex}{}_{\text{total loss}}}\right)\ast 100}$

The global exergy efficiency (η_exergy) indicates the profitability and effectiveness of the process at the industrial level. This performance indicator is given by Eq. (15). (15)${\displaystyle {\eta }_{\text{exergy}}=\left(1-\left(\frac{\mathrm{Ex}{}_{\text{loss}}}{{\mathrm{Ex}}_{\text{total}-\mathrm{in}}}\right)\right)\ast 100}$

Process simulation

The process simulation of the large-scale synthesis of TiO₂ nanoparticles via green chemistry was performed under defined assumptions as follows:

• Process simulation was carried out at steady state with fixed conditions such as processing capacity, pressure of process stages set in 1.01 bar, room temperature at 301.15 K. The detailed operating conditions are summarized in Table 1.

• The hydrolysis reactor was simulated using a RStoich model for a defined conversion yield of 0.93 mol TTIP/mol TiO₂.

• The centrifuge was modeled as a membrane to retain solid nanoparticles and the permeate was mainly a mixture of residual water and ethanol.

• The pump was assumed as part of the mixing stage to facilitate calculations

• Manipulator blocks were employed for washing, infusion and calcination stages.

The stream properties were estimated by a non-random two liquid (NRTL) thermodynamic model due to its well-known accuracy for this type of mixture. However, more detailed validation of the selected model is summarized in Table 2 by comparing the chemical properties of TiO₂ nanoparticles provided by Aspen Plus^® with those reported in literature (Shi et al., 2013).

Figure 4 shows the process flowsheet for large-scale production of TiO₂ nanoparticles via green chemistry. In first section, the STREAM 1 is the lemongrass stream used as raw material for oil extraction. Such stream is sent into a cleaning (CLEAN), washing (LAV1) and drying (DRYER 1) stages. After feedstock pretreatment, resulting stream (STREAM 11) passed through a crushing (CRUSH) stage to reduce size of particles and increase surface area. Then, liquid–solid extraction (INFUS) was carried out, followed by an evaporation (EVP) stage. The oil extract is collected, cooled until room temperature and stored with 96% wt. moisture content (STORA). The second section was simulated using models for reactors, mixer and centrifuge in Aspen Plus^® library. First, TTIP precursor is mixed with water (MIX 1) and the resulting mixture is fed into a hydrolysis reactor (HYDRO) along with the oil extract. The product stream is mainly nanoparticles (STREAM 26) requiring further purification. To this end, three centrifuges (CENTR1, CENTR2 and CENTR3) and one dryer (CALC) were employed to achieve high purity. Finally, TiO₂ nanoparticles were cooled to room temperature (STREAM 39). Table 3 lists mass flowrates for main process streams of the simulated route. For a processing capacity of 32,675 t/y lemongrass and 5,724 t/y TTIP, simulation reported a production rate of 1,496 t/y TiO₂ nanoparticles. Hence, the overall production yield was calculated in 0.26 kg TiO₂/kg TTIP for greener synthesis of titanium-derived nanoparticles.

Exergy analysis results

According to the well-defined methodology for exergy analysis, components of process streams were identified in order to search their specific chemical exergies. Open literature layout an extensive database of chemical exergy, e.g., Bilgen, Keles & Kaygusuz (2012) reported values of chemical exergy for many solvents and bio-oils. In this work, the standard chemical exergies reported by Rivero & Garfias (2006) and Szargut (2007) were employed to perform exergy analysis. Table 4 shows the specific chemical exergy for myrcene, uncedyne, nerol, cellulose, water, oxygen, TTIP, TiO₂, among others, which are the main components of the oil extract, precursor and nanoparticles.

The thermodynamic properties of specific chemical exergy, enthalpy and entropy were key data to continue with the next step in exergy analysis. These properties’ values were entered into Eqs. (2) and (4) to calculate the physical and chemical exergy of process streams, respectively. Table 5 summarizes the results obtained for the main process streams including raw materials and product.

Figure 5 depicts the contribution of process stages to exergy loss (or irreversibilities) to identify stages needing improvement from an energy viewpoint.

Figure 6 shows the exergy analysis results for each process stage reporting irreversibilities, exergy loss or destroyed exergy (%), exergy of residues and exergy efficiency.

The overall energy performance of large-scale TiO₂ production via green chemistry is shown in Fig. 7.