1. Introduction

Supercritical technology is studied in several international research centers, where some studies are mentioned [1-15]. This technology utilizes renewable solvents, such as CO₂, and is a green technology because supercritical processes avoid or minimize environmental damage. The main characteristic of this technology is the use of solvents above or near their critical conditions for temperature and pressure. A solvent in a supercritical state easily penetrates inside a solid matrix and solubilizes the solute even when it is strongly attached to the cellular wall.

The availability of substantial amounts of raw materials in regions near the place of manufacture contributes to the development of supercritical fluid processes and improves their economic feasibility for industrial applications. This fundamental point is satisfied in several regions of Brazil. This country has one of the highest levels of biodiversity worldwide, which allows the production of low-cost raw materials.

In the latest 29 years, LASEFI (Laboratory of Supercritical Technology: Extraction, Fractionation and Identification of Vegetal Extracts) has been developing processes using supercritical technology. Several patents have been registered, and many scientific research studies have been published in Journals and Conference Proceedings (44 full length and 65 abstracts). Therefore, the objective of this review is to present the knowledge acquired from the recent results obtained in the laboratory and to provide a brief discussion of the advances that occurred from 2009 to 2013.

2. Research Studies Conducted at LASEFI from 2009 to 2013

Some of the current scientific investigations focused on supercritical technology being carried out by LASEFI’s research group are as follows: the extraction of bioactive compounds from botanic matrices and the evaluation of the process parameters, the production of micro- and nano-encapsulated compounds using Supercritical Antisolvent (SAS) and Supercritical Fluid Extraction from Emulsions (SFEE) techniques and the hydrolysis of agroindustrial co-products using sub/supercritical water + CO₂ for the production of sugars that can be eventually utilized for second-generation ethanol production. The laboratory has seven systems to perform these types of experiments. Table 1 and Figure 1 present the characteristics of each one of the systems belonging to LASEFI.

Table 1. Characteristics of LASEFI’s systems for SFE, PLE, micronization and hydrolysis

Item 1 (SFE-2×1L, Figure 1-1) is a home-made system containing two extractors of 1 L each with different geometries. It is a multipurpose unit that can be used for extracting bioactive compounds in continuous mode from several raw materials. Item 2 (SFE-I, Figure 1-2) is also a system that was assembled at LASEFI. It is used when the addition of co-solvents is required during extraction. Both Items 2 and 5 (PLE-I, Figure 1-5) can be used for PLE.

Item 3 (SFE-Spe-ed, Figure 1-3) is a commercial unit for extraction (Applied Separations, 7071, Allentown, USA) and can be used with extractors with volumes from 0.005 L to 0.29 L. Item 4 (SFE-2×5L, Figure 1-4) is the largest unit owned by LASEFI. The respective extraction system possesses two extractors of 5 L each (Thar Technologies, Pittsburgh, USA) arranged in parallel and three 1-liter separators operating in series.

Item 6 (ARADIME, Figure 1-6) is a system that enables the development of several types of processes using pressurized fluids such as sub/supercritical fluid extraction with or without co-solvent and the production of particles via RESS or SAS. The integration of unit operations is one of the recent research topics undertaken by the scientists at LASEFI. As a result of this research, the HYDRO system (Item 7, Figure 1-7) was integrated with upstream extraction units to achieve biomass hydrolysis.

Figure 1. Supercritical technology unitsbelonging to LASEFI for extraction: (1) SFE-2×1L; (2) SFE-I; (3) SFE-Spe-ed; (4) SFE-2×5L; for micronization: (5) ARADIME; and for hydrolysis: (6) HYDRO

2.1. Extracting Bioactive Compounds

Several botanic matrices that contain compounds with potential applications in the food, chemical and pharmaceutical industries have been studied. One of these compounds is β-ecdysone, a saponin with therapeutic properties found in Brazilian ginseng (Pfaffia glomerata). Leal et al [16] obtained Brazilian ginseng extracts using supercritical CO₂ in the SFE-I system. According to their results, the extract’s antioxidant activity was a function of the extraction conditions, where the largest antioxidant activity was obtained at 30 MPa and 303 K.

A patent by Meireles et al [17], PI0900551-0A2, was deposited with the Brazilian National Institute of Industrial Property (INPI). The patent presents a supercritical fluid extraction process for active compounds from Brazilian ginseng roots. The described process can be performed in continuous or semi-continuous mode. The choice of one extraction mode over the other depends on the number of extractors and on the control devices that make up the process operation plant.

Takeuchi et al [18] extracted phenolic compounds from macela (Achyroclinesatureioides) using mixtures of CO₂ plus ethanol using the SFE-I system (Item 2, Figure 1). The addition of a co-solvent provided another means of manipulating the solvent selectivity, improved the extraction yield and intensified the functional properties of the extracts. High and stable antioxidant activities were obtained.

Phenolic compounds were extracted from pomegranate (Punicagranatum). The compounds present in this fruit show anti-inflammatory and antimicrobial properties. The supercritical fluid extraction was efficient because the extract presented a high TPC (389 mg.g^-1). The appropriate operating conditions were 30 Pa and 323 K [19].

The extraction of antioxidant compounds from jabuticaba (Myrciaria cauliflora) byproducts was investigated using SFE with a co-solvent. The SFE-I system has a co-solvent line (as seen in Scheme 1) and was used for performing the experimental assays. A greater extraction yield was achieved at 30 MPa and 333 K, resulting in 25 g extract/100 g raw material. However, the highest antioxidant activity was obtained at 20 MPa and 323 K. An increase in temperature resulted in decreased recovery of the antioxidant compounds, indicating the extraction of undesirable compounds. This is likely because most of the antioxidant compounds are unstable and highly susceptible to thermal degradation [20].

Scheme 1. Flow diagram for the SFE-I experimental apparatus for extraction, including cosolvent feed system (adapted from Veggi [21])

Another patent (PI0903275-4A2) developed at LASEFI was deposited by Meireles and Rosa [22] with the INPI. This patent mentions the extraction and purification process of artemisinin from the solid mass of Artemisia annua using supercritical technology. The extraction process is divided into three stages. The first stage entails the contact and dissolution of the solid particulate mass in supercritical CO₂ inside the extractor. The second stage aims to obtain the purified extract and requires contact between the CO₂ and the vegetal extract dissolved in a fixed polar phase in a fractionation column. The third stage, elution, involves contact between the CO₂ and co-solvent mixture with a fixed polar phase in a second fractionation column.

Scheme 2 shows the flow diagram of the SFE-Spe-ed system (Applied Separations, model 7071, Allentown, USA). This unit is an efficient and simple apparatus used to obtain various target compounds. SFE of chamomile (ChamomillarecutitaL.) extract was studied in the SFE-Spe-ed system using a 0.29 L extractor. A yield of approximately 3.5 g extract/100 g raw material was obtained. Compounds such as en-in-dicycloether and cis-β-farnesene were identified in the extracts. Mathematical modeling and optimization were carried out to determine the optimum conditions that maximize the extract amount. The effect of particle diameter on extraction yield was also investigated. The results show that, for chamomile, the extraction yield is a weak function of particle diameter. Nevertheless, the other process parameters, temperature and pressure, strongly influenced the extraction yield. Increasing either of these parameters led to enhanced extraction of bioactive compounds from chamomile [23].

Scheme 2. Flow diagram for the SFE-Spe-ed experimental apparatus for extraction (adapted from Prado [25])

SFE of annatto (Bixaorellana L.) was performed in the SFE-Spe-ed apparatus to obtain an extract rich in tocotrienols and defatted bixin-rich seeds. The highest extraction yield (2.2 g extract/100 g raw material) was reached at 40 MPa and 333 K. In this study, cycles of CO₂ pressurization/depressurization were tested, but their influences on the bixin yield were negligible. The effects of pressure release on modifying the cell membrane due to rapid gas expansion did not provide an improvement in the process efficiency [24].

Another piece of equipment belonging to LASEFI is the SFE-2×1L system (Scheme 3), which includes two extractors of 1 L each operating in parallel. The system was assembled to test and validate an extraction process for bioactive compounds in continuous mode. This configuration allows vegetal extracts to be obtained uninterruptedly, and thus, the productivity can be improved and the costs of production can be reduced. However, operating in this configuration requires the kinetic curves of both extractors to be identical, despite their different geometries. Currently, experimental assays are being carried out to establish the operation criteria and process parameters that will enable comparison of the extraction curves in these extractors. The purpose is to obtain equals mass transfer rates for both extractors. Thus, one of the criterion can be the maintenance of the S/F (mass of solvent/mass of raw material) ratio and the extraction time constants. The review by Zabotet al [26] provides detailed information about the influence of extractor geometries on the kinetic profiles of the extraction.

Scheme 3. Flow diagram for the SFE-2×1L experimental apparatus for extraction (adapted from Zabotet al [28])

To assess the feasibility of SFE in industrial applications, knowledge obtained from laboratory scale experiments should be used to conduct a pilot run. Therefore, scale-up criteria that are able to reproduce the overall extraction curves (OEC) need to be defined.

Prado et al [27] utilized the available equipment at LASEFI and studied processes involving SFE scale up. The investigations aimed to obtain extracts from clove (Eugenia caryophyllus) and sugarcane residue. The SFE-Spe-ed system (extractor of 0.29 L) and the SFE-2×5L system (Scheme 4) were used in this research. The proposed scale-up criteria consisted of maintaining the S/F ratio and the extraction time constants. Considering a 15-fold increase in the raw material mass for the SFE-2×5L, the solvent mass was also increased 15 times. Operationally, the solvent mass flow rate was 15 times higher and the extraction time was maintained. Adopting these simple criteria of maintaining the S/F ratio and the extraction time constants resulted in the appropriate replication of the kinetic profiles of the OEC in the scale-up experiment. Similar yields were reached in both the SFE-Spe-ed (I) and the SFE-2×5L (II) system. The yields of clove extracts were 15 g/100 g (I) and 14.5 g/100 g (II). The yields of sugarcane residue were 2.5 g/100 g (I) and 2.8 g/100 g (II).

Scheme 4. Flow diagram for the SFE-2×5L experimental apparatus for extraction, including the co-solvent feed system (adapted from Prado [25])

Prado et al [29] investigated the same scale-up criteria mentioned above on SFE of grape (Vitisvinifera) seed extract. The results were satisfactory because the yields were 11.9 g/100 g (I) and 11.2 g/100 g (II) using an S/F ratio of 8.4. The authors further tested the separation step on the SFE-2×5L system using three separators of 1 L operating in series (Scheme). The temperature was kept at 313 K, and pressures of 10 MPa, 6 MPa and 3 MPa were used for the separators 1, 2 and 3, respectively. Most of extract was recovered in separator 1 (86%), and the rest of it was recovered in separator 2. There was no extract in separator 3, which indicates that, using the operating conditions selected for separators 1 and 2, it is possible to precipitate all the extract. Similar results were obtained for extraction from clove [27].

Figure 6 shows a flow diagram for the PLE-I system, which uses pressurized liquid, generally water or ethanol, for extracting mainly polar compounds that possess chemical affinity for these solvents. To summarize, the research studies conducted at LASEFI emphasize the extraction of bioactive compounds at the highest extraction yield by optimizing process parameters through experimental work coupled with mathematical modeling. In addition, these studies investigate the chemical composition of the extracts and the functional properties, such as antioxidant activity, of the target compounds.

Scheme 5. Flow diagram for the PLE-I experimental apparatusfor extraction (adapted from Rodrigues et al [38])

Scheme 6. Flow diagram for the ARADIME experimental apparatus for micronization (adapted from Santos [30])

2.2. Production of Micro- and Nano-particles

Currently, novel techniques to produce micro- and nano-particles using supercritical technology are being developed to overcome the drawbacks found in conventional processes. In our research group, studies aiming to produce micrometer-scale particles have been carried out since 2009. Santos [30] designed, assembled and tested the ARADIME apparatus, the flow diagram of which is shown in Scheme. This home-made multipurpose system is used to conduct experiments with pressurized fluids that allow the production of particles of functional pigments using RESS (Rapid Expansion of Supercritical Solutions) and SAS (Supercritical fluid Anti-Solvent).

Santos et al [31] demonstrated that the SAS process could be successfully utilized to co-precipitate microparticles of polyethylene glycol (PEG) loaded with bixin-rich extract. Moreover, the RESS process using ethanol as co-solvent can be effectively employed to encapsulate rutin and anthocyanin-rich extract in the PEG matrix.

Anthocyanins extracted from jabuticaba (Myrciaria cauliflora) skins were encapsulated using supercritical CO₂ and ethanol as the co-solvent. Particles encapsulated by RESS at different pressures and temperatures retained the extracts’ biological activity and preserved the extract’s stability against light and heat degradation. The best operational conditions were 313 K and 20 MPa [32].

Carotenoid particles of sub-micron size were produced using SFE with an oil-in-water emulsion. Suspensions containing stabilized carotenoids with final particle size of 344–366 nm, an encapsulation efficiency of 34–89% and a degree of isomerization from trans- to cis-carotenoid forms in the range of 0.02–15% were obtained [33]. A novel process, known as OEPO (organic solvent extraction and online particle formation), was developed by Santos et al [34]. The process consists of the combination of Pressurized Liquid Extraction (PLE)-Supercritical Anti Solvent (SAS) precipitation, PLE-SAS co-precipitation and PLE-Supercritical Fluid Extraction of Emulsions (SFEE). This is a suitable and promising process to obtain, in only one step, different products as precipitated extract, co-precipitated extract or encapsulated extract in suspension.

All of the recent works completed at LASEFI related to this emerging technology [31-37] belong to a broader project whose overall goal is the development of alternative sustainable systems for the formation of particles of sensitive bioactive substances from many vegetable matrices.

2.3. Hydrolysis of Agroindustrial Co-products Using Supercritical Technology

Another research line, pursued at LASEFI since 2004, relates to the hydrolysis of agroindustrial waste using sub/supercritical water + CO₂ for the production of sugars that can be eventually utilized for second-generation ethanol production. The HYDRO system (Scheme 7), containing a 0.05 L reactor, was designed, assembled and tested by Prado et al [39]. The system was designed to operate at pressures up to 40 MPa and temperatures up to 673 K. Water is pumped from reservoir 2, and if addition of CO₂ is enabled, the static mixer (instrument 9) serves to homogenate the solvents before their percolation inside the reactor containing the biomass. Generally, the biomass consists of lignocellulosic material, a non-edible source of fermentable sugars widely available in the nature. To validate the HYDRO system, hydrolyses of cellulose and sugarcane bagasse were performed. The glucose equivalent concentrations in the hydrolysate fractions were approximately 6 g/100 g of cellulose for a reaction time of 60 min and 5.9 g/100 g sugarcane bagasse for a reaction time of 40 min.

Scheme 7. Flow diagram for the HYDRO experimental apparatus for hydrolysis (adapted from Prado et al [39])

The reducing sugars obtained in the hydrolysate fractions can be used in fermentative processes for the production of renewable energy sources. This method of obtaining energy can become the basis for the development of modern industrial economies in the future. Ethanol is emerging as an efficient and economically feasible fuel. With the goal of obtaining sugar precursors for ethanol production, Follegatti- Romero et al [40] performed experimental assays with the HYDRO system using cellulose with water under subcritical conditions of 457 K, 470 K and 482 K with a reaction time of 68 min and at a pressure of 20 MPa. The authors concluded that the results obtained were satisfactory and that the total reducing sugar content recovered at 482 K was the highest. At this temperature, the recovery of sugars was also the fastest.

3. Analytical Procedures

Chemical characterization of the products is important to assess their qualities. At LASEFI, researchers usually perform thin layer chromatography (TLC) for qualitative analysis of the extracts. An important method of quantitative analysis is gas chromatography (GC) coupled to a flame ionization detector (FID). HPLC (high performance liquid chromatography) analyses are also performed to identify and quantify the target compounds obtained. Additionally, supercritical fluid chromatography (SFC) has been recently implemented in the laboratory. The system will be used to develop methodologies for analyzing a wide range of components (phenolic acids, catechins, flavonoids, alkaloids, carotenoids, sterols, etc.) present in different natural sources.

4. Economic Evaluation of Processes Employing Supercritical Technology

Processes that make use of supercritical technology are usually technically feasible. Nonetheless, detailed investigations about their economic feasibility need to be performed if these processes are to be scaled up for industrial applications. Therefore, simulations of the cost of manufacturing (COM) extracts from several botanic matrices are being developed by researchers at LASEFI. Cavalcantiet al [20] simulated the COM of jabuticaba extract obtained by SFE in extractors of various sizes using a commercial simulator. According to their results, the cost of extract production was lower than US$ 10.00/kg in a 300 L extractor. Veggiet al [41] compared the COM of jabuticaba extract obtained by PLE and by conventional techniques in a 300 L extractor. The values of COM of the extracts were US$ 15.53/kg in PLE, US$ 410.21/kg in ultrasound-assisted extraction and US$ 778.42/kg in Soxhlet extraction. The values of COM of extracts from Amazonian plants as buriti, pupunha and pressed palm fiber [42], annatto [24], pomegranate [19], grape [29] and sugarcane residue [27] were also simulated.

5. Summary of the Current Scientific Investigations Conducted at LASEFI

Table 2 shows a summary of the recent projects developed at LASEFI coveringthe application of supercritical fluids to several experimental processes. The large biodiversity of bioresources in Brazil has favored the use of different raw materials, as annatto (Bixaorellana L.), which contains tocotrienols and bixin [24] to be used in the treatment of cardiovascular diseases [43] and to be applied as colorant in food processing [44], respectively.

Table 2. LASEFI’s research studies from 2009 to 2013

6. Concluding Remarks

Some progress has been made compared to the recent research results. For instance, one ongoing projectaims to develop an integrated analysis system by gathering the sample preparing with the chromatographic analyses. The sample preparing step is done by pressurized liquid extraction (PLE) assisted by ultrasound or supercritical fluid extraction (SFE) online coupled to a purification step using solid phase extraction (SPE). This integrated system will be applied to develop methodologies for determining a wide range of bioactive compounds.

Likewise, producing high-added vegetal extracts using green processes is a continuous effort done by the research group. Emerging applications of supercritical fluids, such as micronization and the encapsulation of nanoparticles, are being studied to produce several phytochemicals useful in the food and pharmaceutical industries. The processes include the encapsulation of nanoparticles containing target components from annatto, jabuticaba waste and Brazilian ginseng. Moreover, the use of environmentally friendly pressurized fluid technologies to obtain products from biomass has shown encouraging results. Another ongoing project is focused on the attainment of multiple products through a combination of different processes and sequenced with the use of multiple fluids. The goal is to use the knowledge and available tools on energy integration and life cycle analysis seeking an integral use of all parts of the Brazilian ginseng (root and aerial parts), expecting a Brazilian ginseng biorefinery based on supercritical technology.

Process intensification, process integration, extraction in continuous mode and hydrolysis of agroindustrial residues are also ongoing projects performed at LASEFI. In summary, the present objectives of the research group revolve around the scale up of the laboratory processes for the production of various extracts to effectively start up an industrial plant employing supercritical technology in Latin America.

ACKNOWLEDGMENTS

Giovani L. Zabot thanks FAPESP (2011/23665-2) and Moyses N. Moraes thanks CAPES for granting their Ph.D. assistantships. M. A. A. Meireles acknowledges a productivity grant (301301/2010-7) from CNPq. The authors acknowledge the financial support from CAPES, CNPq and FAPESP.