A. Cimpeanu, Sulzer Chemtech, Houston, Texas (U.S.)
The refining and petrochemical industries are constantly evolving, driven by the dual imperatives of maximizing operational efficiency and minimizing environmental impact, including the critical goal of reducing carbon emissions.
Aromatic hydrocarbons benzene, toluene and xylenes (collectively known as BTX) are vital intermediates in the petrochemicals industry. These molecules are extracted from hydrocarbon streams such as reformate, pyrolysis gasoline (pygas) and coke oven light oil, which contain a mix of aromatics and non-aromatics (paraffins, olefins and naphthenes). Due to the close boiling points of these compounds, conventional distillation is often insufficient, necessitating the use of solvent-based extraction techniques. Two primary methods are employed for aromatic extraction:
Liquid-liquid extraction (LLE): Separating close-boiling components based on liquid-liquid separation. It operates at ambient temperatures, making it energy efficient.
Extractive distillation (ED): A selective solvent alters the relative volatility of components, enabling separation by distillation.
Sulfolane—also known as tetrahydrothiophene 1,1-dioxide—is a highly polar, aprotic solvent widely used in the petrochemicals industry, particularly for aromatic extraction in both ED and LLE technologies. The advantages are high selectivity and solvency for aromatics, low vapor pressure (reduces solvent loss), and thermal and chemical stability. It is also environmentally safer than older solvents like phenol. However, some limitations are still present:
Limited flexibility for optimizing/improving plant performance
May not address specific needs of the unit
Not suitable for low-energy/low-carbon intensity operations
Inflexible for new feed sources
Limitations in optimizing the value of aromatics (recovery and purity) in older units
Corrosion issues due to stability and decomposition.
These limitations in traditional sulfolane-based BTX extraction systems in LLE and ED units arise because they operate with inherent selectivity and solvency limitations (primary solvents properties), particularly when processing high-aromatic content feedstocks such as pygas and reformates. These constraints require elevated solvent-to-feed (S:F) ratios, resulting in increased utility consumption and associated carbon emissions. Next-generation custom solvent formulations like the author’s company’s proprietary solvent technologya should demonstrate superior aromatic/non-aromatic discrimination and enhanced dissolution characteristics, enabling solvent circulation reductions of 15%–25% while maintaining or improving aromatics recovery rates.
Recent commercial implementations have achieved utility consumption reductions exceeding 10%, directly correlating with reduced carbon dioxide (CO2) emissions. The advanced solvents function as drop-in replacements for existing sulfolane systems, enabling upgrades without unit shutdown or major equipment modifications. In these cases, adding co-solvents to the main existing solvent, typically sulfolane, improved separation in aromatic extraction units. The benefits of customizing solvent approaches have demonstrated improved unit performance in commercial applications. This approach focuses primarily on improving sulfolane-based aromatic extraction processes by blending new formulated solvents (named co-solvents) into the main sulfolane solvent.1,2 The main goal in aromatic extraction units is to reduce energy consumption (mainly steam) and, consequently, CO2 emissions.
Screening and selection. The initial step in co-solvent development involves pre-selection based on key molecular properties such as polarity (dipole moment) and dielectric constant. A thorough literature review also supports the identification of promising candidates. Once a co-solvent is shortlisted, its performance, in combination with the main solvent (typically sulfolane), is evaluated through laboratory screening. For LLE applications, a simple shaking test at constant temperature followed by gas chromatography (GC) analysis is sufficient. For ED processes, vapor-liquid equilibrium (VLE) behavior is assessed using an equilibrium cell under total reflux conditions, simulating a theoretical distillation tray.
Enhancing sulfolane’s performance through co-solvent addition targets improvements in its primary properties: selectivity, solvency (critical for LLE) and relative volatility (essential for ED). Co-solvents are tailored to meet specific process requirements and are typically blended in proportions < 50% with the main solvent. While selectivity is vital for both LLE and ED, solvency plays a more dominant role in LLE operations.
In both LLE and ED, the addition of a co-solvent can significantly increase selectivity and relative volatility between close-boiling components, making the separation process more technically and economically viable. Beyond primary performance metrics, secondary selection criteria are equally important. These include:
Freezing and boiling point
Toxic free
Corrosion free
Thermally stable
Non-hazardous
Commercially available (cost)
Sulfolane compatible.
Following successful initial screening, the next critical step involves pilot plant testing of the solvent/co-solvent blend to fully validate the selected co-solvent. Prior to industrial implementation, a detailed commercial evaluation based on economic analysis and business case development is conducted. Subsequently, a robust simulation model incorporating accurate thermodynamic parameters must be developed to predict process performance.1,2,3
Selectivity is the primary criterion during the co-solvent screening process, as improvements in this parameter can significantly enhance the efficiency of both ED and LLE. In addition to selectivity, solvency, which is particularly critical in LLE applications, was also evaluated. These two properties represent the core performance indicators for co-solvent selection in aromatic extraction. Beyond these, secondary attributes such as boiling point, toxicity, thermal stability, commercial availability and compatibility with sulfolane were also considered to determine the suitability of a co-solvent candidate.
In LLE operations, selectivity is the first and most crucial parameter to assess, as it reflects the solvents’ ability to differentiate between aromatic and non-aromatic hydrocarbons. When a solvent is introduced into a hydrocarbon mixture at a defined ratio, it enhances the separation by making non-aromatics less affine with solvent and an aromatics-rich phase, thereby facilitating phase separation. A higher selectivity under constant S:F ratios improves the overall separation efficiency. Co-solvents are typically blended with the main solvent in varying proportions to enhance these properties, with selectivity being the primary target in LLE optimization.
The selectivity (β) is calculated for each non-aromatic/aromatic hydrocarbon pair using Eq. 1:
β = (xNA1 / xNA2 ) / (xA1 / xA2) (1)
where:
xNA1 and xNA2 are the molar fractions of non-aromatic hydrocarbons in the raffinate (lighter portion of a two liquid phase) and extract phases (bottom layer of double liquid phase), respectively.
xA1 and xA2 are the molar fractions of aromatic hydrocarbons in the raffinate and extract phases, respectively. This ratio reflects the solvent’s ability to preferentially separate aromatics from non-aromatics.3
In highly non-ideal systems, selectivity can also be expressed in terms of activity coefficients (Eq. 2):
β = γNA / γA (2)
γNA and γA are the activity coefficients of non-aromatics and aromatics in the solvent-rich phase. These coefficients are typically greater than unity, especially for non-aromatics, as the interaction between aromatics and solvents is stronger than the interaction between non-aromatic and solvent.
Solvency or solvent power (SP), another critical parameter, is defined in Eq. 3:
SP = xA2 / xA1 (3)
This ratio indicates the solvent’s capacity to dissolve aromatic hydrocarbons, also on a solvent-free basis.
In LLE, a fundamental trade-off exists between selectivity and solvency. Typically, higher selectivity corresponds to lower solvency, and vice versa. Therefore, when developing a co-solvent, the objective is to enhance selectivity while maintaining solvency at a comparable level. Solvency directly influences the S:F ratio in practical aromatic extraction processes, while increased selectivity contributes to improved product purity and supports S:F ratio optimization. When both selectivity and solvency are favorably balanced, the process can achieve lower utility consumption and improved overall efficiency.1
In ED, selectivity plays a crucial role in determining separation efficiency, which is typically quantified by relative volatility (a), which measures how effectively a solvent facilitates the separation of close-boiling aromatic and non-aromatic hydrocarbons. Similar to LLE, introducing a co-solvent into the hydrocarbon mixture at a defined ratio enhances , making the separation between non-aromatics (raffinate) and aromatics (extract) more feasible under VLE conditions. As a increases under a constant S:F ratio, the separation becomes more efficient, improving the overall performance of the aromatic extraction process.
Relative volatility () for each non-aromatic/aromatic hydrocarbon pair is calculated using Eq. 4:
α = (yNA / xNA ) / (yA / xA ) (4)
yNA and xNA are the molar fractions of non-aromatics in the vapor and liquid phases, respectively.
yA and xA are the molar fractions of aromatics in the vapor and liquid phases, respectively.4
By applying equilibrium relationships in non-ideal solutions, the relative volatility () can be expressed in terms of activity coefficients and vapor pressures (Eq. 5):
α = (γNA / γA) × (pNA / pA) (5)
γNA and γA are the activity coefficients of non-aromatic and aromatic hydrocarbons in the solvent-rich phase.
pNA and pA are the vapor pressures of the pure non-aromatic and aromatic components at system temperature.
This relationship is applicable to both single-phase (VLE) and dual-phase [vapor LLE (VLLE)] systems, with the specific note that in VLLE conditions, all parameters in Eq. 5 refer to the solvent-rich phase. In ideal mixtures, such as pure hydrocarbons, the activity coefficients approach unity, simplifying the equation to a vapor pressure ratio. However, when a selective solvent is introduced into an aromatic/non-aromatic mixture, the activity coefficients diverge significantly, with γNA typically much greater than γA. This disparity increases α above unity, enabling effective separation of close-boiling components.4
By combining Eqs. 2 and 5, relative volatility (α) can be alternatively expressed as Eq. 6:
α = β’ × (pNA / pA) (6)
In a VLLE system (where two liquid phases are formed), β’ corresponds directly to the LLE selectivity defined in Eq. 1. In ED operations, which typically involve a single liquid phase (VLE), β’ does not represent true LLE selectivity but still serves as a useful indicator of solvent performance. In both cases, an increase in selectivity leads to higher relative volatility, which enhances separation efficiency and enables a reduction in the S:F ratio, ultimately lowering energy consumption.
Eq. 6 highlights a key insight: when a co-solvent improves selectivity in LLE, it also tends to increase a in ED, making it beneficial across both separation technologies.
Considering all the above relationships for calculation of β, α and SP, rigorous lab screenings for various aromatics/non-aromatics raw material were performed.
Based on the relationships used to calculate β, α and SP, a series of rigorous laboratory screening tests were conducted using various aromatic and non-aromatic hydrocarbon mixtures. To simulate industrial conditions, multi-component feeds were prepared with key constituents, representative of typical raw streams such as pygas and reformate. Three standard feed compositions were used, varying the aromatic content between 25% and 75% by weight.1
As previously discussed, β quantifies the effectiveness of separating aromatics from non-aromatics between two liquid phases. In LLE operations, when a solvent is added to a hydrocarbon mixture, its β determines how easily the two groups can be separated typically by gravity-driven phase separation. Selectivity increases with higher S:F ratios, but the goal of co-solvent development is to enhance selectivity while enabling a reduction in the S:F ratio. This results in improved separation efficiency and reduced utility consumption, thereby optimizing the overall performance of the aromatic extraction process.
FIG. 1 illustrates the selectivity performance for several aromatic/non-aromatic hydrocarbon pairs at a S:F ratio of 3:1. The base solvent, sulfolane, was blended with 30% of a proposed co-solvent, and the resulting selectivity values were compared against those of sulfolane alone.
Three representative binary pairs were selected to demonstrate selectivity trends:
Total non-aromatics vs. total aromatics
n-heptane (nC7) vs. toluene
n-octane (nC8) vs. paraxylene (PX).
The addition of co-solvent consistently enhanced selectivity across all aromatic content levels. Notably, the improvement was more pronounced at lower aromatic concentrations in the feed, a behavior typical of LLE operations. All screening experiments were conducted under controlled temperature and pressure conditions.
Similar to LLE, α in ED is a key indicator of separation efficiency between aromatic and non-aromatic hydrocarbons. α increases with higher S:F ratios, but the strategic use of co-solvents, such as blending with sulfolane, can improve α while allowing for a reduction in the S:F ratio.
This improvement not only enhances separation but also reduces the number of trays and reboiler duty in industrial distillation columns, leading to lower energy consumption. The effectiveness of α depends on several factors, including feed composition, co-solvent type and concentration, and the specific hydrocarbon pairs selected for evaluation.
FIG. 2 illustrates α trends for four representative binary pairs’ total non-aromatics/total aromatics, cyclohexane/benzene, nC8/toluene and ethyl-cyclohexane/PX across S:F ratios of 3:1, 5:1 and 10:1. The data, based on a feed composition of 75% aromatics and 25% non-aromatics with a 20% co-solvent added to sulfolane, clearly show that α increases with the S:F ratio. Notably, the sulfolane/co-solvent blend consistently outperforms sulfolane alone.
FIG. 1 shows that the β for total aromatics vs. total non-aromatics using sulfolane alone is 6.975, while the blend with 30% co-solvent achieves a of 7.725., an approximate 10% improvement. Similarly, FIG. 2 demonstrates that at a 5:1 S:F ratio, the a increases from 4.29 with sulfolane alone to 4.58 with the co-solvent blend. This enhancement becomes even more pronounced at higher S:F ratios, which reflect actual concentration profiles in ED columns.
Using Eq. 6, and assuming vapor pressure ratios of 0.916 for sulfolane and 0.896 for the blend, the calculated pseudo-selectivity (β′) values are 4.68 and 5.11, respectively, again showing a ~10% increase. This consistency between LLE and ED confirms that a co-solvent that improves selectivity in LLE also enhances α in ED.
While lab screening provides valuable insights, it is insufficient for industrial implementation. The next step involves developing a robust process simulation model using commercial simulators and thermodynamic packages, which account for system non-ideality. These models help determine optimal co-solvent concentration and predict impacts on energy consumption and product quality. Final validation is achieved through pilot plant testing, which confirms readiness for full-scale deployment.
Industrial applications. Sulfolane and sulfolane-based solvents are primarily used in two aromatic extraction technologies: LLE and ED. In the LLE process (FIG. 3), the solvent flows countercurrently to the hydrocarbon feed, selectively extracting aromatic compounds while leaving non-aromatics in the raffinate product. The extracted aromatics exit the extractor at its bottom as extract and undergo further purification in an extractive stripper (ES) to remove any entrained non-aromatics. The solvent is then recovered in a solvent recovery (SR) column and recycled back to the extractor.4
FIG. 4 illustrates the ED process, which centers around a single column [ED column (EDC)], where the solvent functions as a selective reflux. Solvent is introduced countercurrently to the vaporized feed and column bottoms, enabling the extraction of aromatics at the bottom, while non-aromatics are removed as raffinate from the top. Similar to the LLE process, the aromatic-rich solvent stream is routed to an SR column (SRC), where the solvent is separated and recycled back to the EDC. In both technologies, the purified BTX product exits from the top of the SRC.
In ED, selectivity is a key driver of energy efficiency and product quality. However, excessive selectivity can lead to operational challenges, such as the formation of dual liquid phases on trays, particularly in the upper section of the ED column. In contrast, LLE benefits from both selectivity and solvency: high selectivity improves product purity, while high solvency can reduce the S:F ratio and enhance aromatics recovery.4
To address these challenges, co-solvents are introduced to enhance sulfolane’s performance. These are carefully selected to balance selectivity and solvency without compromising plant operability. Typically, co-solvents are blended in concentrations ranging from 10%–30% within the sulfolane matrix, offering a practical and scalable solution for upgrading existing units.
Sulfolane-based aromatic extraction technologies are commercially available as licensed processes, including the author’s company’s aromatics recovery technologyb, which is based on ED. While both LLE and ED are widely used, many older units, particularly LLE systems, have been gradually replaced by ED due to its superior efficiency and lower operating costs (OPEX).5,6
However, to emphasize again, when sulfolane is used as the sole solvent, both technologies face limitations in process optimization and adaptability:
Limited ability to tailor performance to specific unit needs
Unsuitability for low-energy or low-carbon-intensity operations
Inflexibility in handling new or variable feedstocks
Constraints in maximizing aromatics recovery and purity, especially in aging units
Corrosion risks due to solvent degradation.
A novel solvent platform: A next-generation solution for aromatics extraction. The author’s company’s proprietary solvent technologya represents a significant advancement in aromatic extraction technology. These co-solvent blends are engineered to deliver superior selectivity and solvency, offering substantial performance improvements over traditional solvents such as sulfolane.
The solvent technologya can be deployed as a drop-in replacement in existing sulfolane-based units, allowing operators to upgrade solvent systems without requiring shutdowns, equipment modifications or revamps. This seamless integration enables immediate operational benefits, including enhanced throughput, improved aromatics recovery and reduced energy consumption. In most cases, full performance gains are realized within one month of complete solvent addition, while preserving the existing sulfolane inventory.
In addition to retrofit applications, the solvent technologya is also the proprietary solvent used in the author’s company’s aromatics recovery technologyb, positioning the process as a leading solution in the industry. With > 50 operating unitsb worldwide utilizing the author’s company’s solvents, the company has accumulated extensive operational experience, reinforcing its leadership in solvent innovation.1,6
Research and development efforts, supported by its U.S., Singapore and Swiss facilities with a combined 65 yrs of expertise in solvent testing and pilot plant operations, have focused on replacing conventional solvents with high-performance alternatives. The testing of the company’s novel solvent technologya in both LLE and ED units has demonstrated measurable improvements aligned with customer goals:
Increased aromatics recovery
Enhanced energy efficiency
Higher unit throughput.
At the core of the solvent technologya is a breakthrough in solvent design. Enhanced selectivity enables more precise separation of aromatics from non-aromatics, while improved solvency ensures efficient dissolution and recovery of valuable components. These properties allow units to operate at significantly lower S:F ratios, resulting in reduced utility consumption (steam, electricity, cooling water) and a smaller environmental footprint.
The solvent technologya contributes directly to sustainability by lowering carbon emissions through reduced energy usage. This supports compliance with environmental regulations and enables customers to benefit from carbon reduction incentives such as tax credits and subsidies. The drop-in nature of the upgrade accelerates the path to sustainability without added capital costs.
Traditional sulfolane systems often struggle with feeds containing high concentrations of aromatics and heavy aromatics, such as pygas and reformates. The solvent technologya addresses this limitation by improving recovery efficiency at lower S:F ratios, translating into higher profitability through increased yield of high-value BTX products.
Unlike one-size-fits-all solvents, the solvent technology’sa formulations are tailored to specific unit objectives, considering feed composition, operating conditions and desired product specifications. This customization maximizes unit performance and profitability.
The technology’sa solvents are also designed with safety and operability in mind. They are non-toxic, non-reactive and thermally stable, reducing corrosion risks and simplifying handling procedures. Their compatibility with sulfolane ensures smooth integration into existing systems without requiring specialized infrastructure.
The co-solvents used in the technologya are chemically similar to sulfolane, ensuring linear blending behavior and full compatibility. They enhance not only the solvent’s core properties but also process conditions by adjusting boiling points, density, melting point and thermal stability.1,3
Technologya co-solvent examples and broader applications. While the author’s company’s primary focus has been on upgrading sulfolane-based LLE and ED units, the company has also extended its solvent development efforts to other aromatic extraction technologies that utilize solvents such as n-methyl-2-pyroolidone (NMP), n-formylmorpholilne (NFM) and glycols. In these cases, customized co-solvent blends have been formulated to optimize performance and energy efficiency.
TABLE 1 shows a comparison between sulfolane and the solvent technologya, which also is compared against other commercially available aromatic extraction solvents listed in FIG. 5, highlighting the solvent technology’sa advantages. Notably, the solvent technologya also offers a compelling retrofit solution for non-sulfolane LLE and ED units, enabling performance upgrades through the author’s company’s reformulated solvent blends.
Case Study 1: LLE unit optimization in the U.S. A major U.S. producer requested the author’s company to evaluate one of its LLE units with the goal of increasing mixed xylene recovery from the raffinate stream. An overview is provided in TABLE 2. The results of the study showed that enhancing aromatics recovery not only improved the BTX product value and profitability, but also elevated raffinate purity, simplifying downstream processing and contributing to a lower carbon footprint.
A secondary objective was to implement the solvent upgrade without any plant modifications or operational downtime. The author’s company successfully applied the solvent technologya platform, specifically utilizing a co-solvent, which delivered the desired improvements in performance and efficiency, validating the technologya as a robust solution for existing LLE units.
Case Study 2: ED unit optimization in Asia. A leading producer in Asia asked the author’s company to evaluate one of its ED units, with the goal of reducing utility consumption by at least 10%, thereby lowering the unit’s carbon footprint. A key requirement was to implement the solvent upgrade without any plant modifications or operational downtime.
The author’s company successfully formulated a customized solvent technologya blend featuring a new co-solvent, which was tailored to the unit’s specific operating conditions. Within 5 mos, the new solvent was deployed, achieving the targeted improvements in energy efficiency and overall unit performance (TABLE 3), demonstrating the technology’sa effectiveness as a drop-in solution for ED applications.
Takeaways. Solvent-based aromatic extraction remains a fundamental process in petrochemical operations, directly impacting product purity, energy efficiency and operating costs. While sulfolane has long been the industry standard, its limitations—particularly in energy intensity, feedstock flexibility and corrosion susceptibility—underscore the need for innovation.
The introduction of customized co-solvents through a novel technologya platform marks a significant advancement in this field. These formulations enhance critical solvent properties such as β, SP and α, enabling more efficient separation in both LLE and ED units, significantly lowering the energy intensity of the process.
The solvent technologya offers a drop-in upgrade for existing sulfolane systems, requiring no major equipment changes or downtime. Its implementation delivers tangible benefits, including:
Lower S:F ratios
Reduced utility consumption
Improved aromatics recovery
Enhanced sustainability and carbon footprint reduction.
The platform’s adaptability to specific unit configurations ensures optimized performance across diverse operational scenarios. Validated through laboratory testing, pilot-scale trials and successful commercial deployments, the author’s company’s solvent technologya stands out as a next-generation solution for BTX extraction.
By enabling producers to meet evolving environmental standards, reduce emissions and improve profitability, the solvent technologya is not just a solvent innovation, it is a strategic asset for the future of petrochemical processing.
By decreasing energy consumption and emissions, the technologya aligns with global sustainability targets and regulatory pressures. Its adaptability to diverse unit configurations ensures optimized performance across a wide range of operational scenarios.
Validated through lab-scale testing, pilot trials and commercial implementation, the solvent technologya represents a next-generation solution for BTX extraction, one that empowers producers to reduce greenhouse gas emissions, enhance energy efficiency and advance toward net-zero objectives. More than a solvent innovation, the technologya is a strategic enabler for sustainable petrochemical operations. HP
NOTATIONS
α = Relative volatility
ED = Extractive distillation
LLE = Liquid-liquid extraction or liquid-liquid equilibrium
VLE = Vapor liquid equilibrium
SP = Solvent power or solvency
β = Selectivity
β’ = Pseudo selectivity
P = Liquid true vapor pressure of a component or a mixture
xNA1, xNA2 = Molar fraction of non-aromatics in raffinate and extract phase
xA1, xA2 = Molar fraction of aromatics in raffinate and extract phase
γNA , xNA = Molar fraction of non-aromatics in vapor phase respectively in liquid phase
γA , xA = Molar fraction of aromatics in vapor phase respectively in liquid phase
NOTES
Sulzer’s Techtiv™ solvent
Sulzer’s GT-BTX™
LITERATURE CITED
Cimpeanu, A. and G. Prakasham, “Maximize value from aromatics extraction units,” Sulzer internal presentation, 2025.
Fu-Ming, L., “Extractive distillation: Separating close boiling point components,” Chemical Engineering, November 1998.
Lo, T. C., M. H. I. Baird and C. Hanson, Handbook of Solvent Extraction, Krieger Publishing Company, 1991.
V. Winkle, M., Distillation, McGraw-Hill Chemical Engineering Series, 1967.
Gerbaud, V., D. Rodriguez, L. Hégely, P. Lang, F. Denes and X. You, “Review of extractive distillation: Process design, operation, optimization and control,” Chemical Engineering Research and Design, 2019.
Meyers, R. A., Handbook of petrochemicals production processes, 2nd Ed., McGraw-Hill Education, 2019.
Andrei Cimpeanu is a Technology Manager for aromatics for Sulzer Chemtech. He has a demonstrated history of working in the refinery, petrochemical and chemicals industries. He is skilled in process modelling and simulation, process development, engineering and design activities. Cimpeanu earned an MS degree in chemical engineering and is a licensed Professional Engineer.