New Reactive Extraction Based Reclaiming Technique for Amines

Apr 7, 2016 - A new reclaiming technique based on reactive extraction for removal of heat stable salts (HSS) from monoethanolamine (MEA) used in carbo...
0 downloads 0 Views 7MB Size
Article pubs.acs.org/IECR

New Reactive Extraction Based Reclaiming Technique for Amines Used in Carbon Dioxide Capture Process from Industrial Flue Gases Phattara Akkarachalanont,‡ Chintana Saiwan,‡ Teeradet Supap,§ Raphael Idem,*,§ and Paitoon Tontiwachwuthikul§ ‡

Petroleum and Petrochemical College (PPC), Chulalongkorn University, Bangkok, Thailand 10330 Clean Energy Technologies Research Institute (CETRI), University of Regina, Regina, Saskatchewan S4S 0A2, Canada

§

S Supporting Information *

ABSTRACT: A new reclaiming technique based on reactive extraction for removal of heat stable salts (HSS) from monoethanolamine (MEA) used in carbon dioxide (CO2) capture process has been developed. The extraction process was based on the use of tri-n-octylamine (TOA), Aliquat 336, OH− modified Aliquat 336, a two-step extraction (modified Aliquat followed by TOA), and mixed extractant (modified Aliquat and TOA) in 1-octanol diluent. The best parameters were 69% OH− modified Aliquat, two-step extraction, and mixed extractant (modified Aliquat and TOA) and under optimum extraction conditions were able to improve the extraction efficiency of the original Aliquat and TOA to over 90%. The two-step extraction and mixed extractant were also capable of managing Cl− contamination in MEA solution. Extraction was found to be independent of temperature whereas efficiency reduced with increase of CO2 loading. Therefore, it is recommended to apply this extraction technique to the lean MEA stream after the rich/lean heat exchanger either with or without cooling. Regeneration of used extractant (OH Aliquat) was implemented and optimized with the use of 4 kmol/m3 NaOH. In addition to HSS removal, the new extraction technique was also able to remove major nonionic degradation products also by up to 90%.

1. INTRODUCTION Amine-based absorption for carbon dioxide (CO2) capture is an effective technique for removing CO2 from low pressure flue gases generated from various industrial processes. Effective amines used include monoethanolamine (MEA), diethanolamine (DEA), and methyldiethanolamine (MDEA). Although these amines are effective, they easily degrade during CO2 removal by irreversible reactions with flue gas impurities (e.g., O2, SOx, and NOx). Negative impacts caused by degradation include loss of active amine which reduces the CO2 removal efficiency and capacity of the amine plant. In addition, unwanted degradation products are introduced into the absorption system. Undesirable degradation products including heat stable salts (HSS) are a major concern as they are known to be corrosive. The salts once tied up with the amine are also difficult to release, at least under the normal regeneration conditions used in the CO2 treating units. To ensure stability of the absorption cycle, a reclaiming process is needed to periodically recover the amine from the HSS. The conventional reclaiming process often uses distillation to separate the amine from the HSS contaminants. Since distillation is achieved based on boiling point difference, a large energy input is supplied to the degraded solvent to separate the amine and water from the HSS. This makes distillation a very energy-demanding process, which gives rise to high operating costs. In addition, coevaporation could also © 2016 American Chemical Society

occur since various HSS and degradation products may have boiling points close to that of the amine. As a result, the amine may still be contaminated after distillation. To improve effectiveness of the reclaiming process for amine-based CO2 capture, several techniques have been developed to separate HSS from the amine. In our view, solvent extraction stands out since it can clean up the HSS in the amine solution potentially down to less than 10 ppm levels. In addition, the energy input is not as demanding as that of the distillation based reclamation process because extraction can be done at ambient temperatures. Although solvent extraction has not yet been applied to the CO2 capture process, various works have been done to demonstrate the use of extraction for the removal of carboxylic acids from aqueous phase systems, implying that solvent extraction can be used as a reclaiming technique in amine-based CO2 capture process. Solvent extraction has found many uses in industrial separation, purification, and product recovery, i.e., organic acids from fermentation broth, food, pharmaceutical, refining of uranium, plutonium, and radioisotopes, and, most recently, has gained increasing importance in biotechnology, a combination Received: Revised: Accepted: Published: 5006

January 4, 2016 March 21, 2016 April 7, 2016 April 7, 2016 DOI: 10.1021/acs.iecr.6b00035 Ind. Eng. Chem. Res. 2016, 55, 5006−5018

Article

Industrial & Engineering Chemistry Research of the bioreactor with liquid−liquid extraction. 1−4 In fermentation processes, carboxylic acids (i.e., formic acid, acetic acid, propionic acid, hydroxycarboxylic acids, oxalic acid, glycolic acid, succinic acid, citric acid, lactic acid, etc.) which frequently occur as diluted aqueous solutions require development of an effective method for their isolation and concentration. Various extractants including mixed extractants such as alkyl phosphine oxide, alkyl phosphates, alkyl amines (e.g., trialkyl amine or TAA, quaternary alkyl ammonium halide or Alamine and Aliquat) were used in hydrocarbon diluents, e.g., toluene, acetone, hexane, heptane, decane, and kerosene, alone or with modifiers, e.g., alcohols, chloroform, tributyl phosphate or TBP, methyl isobutyl ketone or MIBK, to increase extractability of the extractant and, when the inert diluent is used, to prevent the formation of a second organic phase. One research work used alamine 336 and Aliquat 336 to extract carboxylic acids, specifically acetic acid, propionic acid, and butyric acid, from diluted aqueous phase for fermentation and wastewater application.5 The result showed that the extraction performance using both extractants was pH dependent. Aliquat 336 was found to extract both undissociated and dissociated acids and, thus, could be used under both acidic and basic conditions. The working concentration of Aliquat 336 was also found to be limited to 25% to avoid the formation of an emulsion between the aqueous and organic extractant phases. A diluent was also recommended as it helped to improve the physical properties of Aliquat 336 and to speed up the separation of the two phases after extraction. Alamine 336, on the other hand, was found to only be able to extract undissociated acid, thus working best at pH lower than 4. Aliquat 336 in both chloride (Cl−) and hydroxide (OH−) forms was also investigated for the extraction of acetic acid in calcium magnesium acetate production using quaternary amine.6 For Aliquat 336 in 1-octanol diluent, its use as in the form of being partially converted to OH− was found to have a higher extraction capacity than the unconverted Aliquat. The reason was the existence of a more favorable exchange equilibrium for replacing the OH− anion by acetate than for replacing Cl− by acetate could be obtained. The extraction capacity of both aliquats was also found not to be affected by temperature. A tertiary amine, tri-n-octylamine (TOA), and Aliquat 336 in butyl acetate diluent were compared for the extraction of 7aminocephalosporanic acid.7 Tri-n-octylamine was found to extract the acid better at a lower pH, and its extraction efficiency decreased with an increase in the pH of the acid solution. An opposite result was observed for Aliquat 336 in which it provided a higher extraction capacity than that of trioctylamine in the higher pH range. The extraction using a mixture of TOA and Aliquat 336 was also found to be promising.8 The mixed TOA/Aliquat 336 system was used to extract lactic acid for fermentation broth application. The mixture was found to work more effectively than the individual extractants. More studies on the extraction by tertiary amines and quaternary amines can be found in the literature. For example, the extraction of p-aminobenzenesulfonic acid using alamine 336 and Aliquat 336 has been reported.9 Six diluents (i.e., kerosene, n-octanol, chloroform, butyl acetate, tetrachloromethane, and benzene) were evaluated for their abilities to increase the extracting power of the amine extractants over a wide pH range. Alamine 336 was found to extract most acids from the aqueous phase by forming a complex with the neutral

form of the acid molecule whereas Aliquat 336 was extracted by the ion exchange process. Aliquat 336 was also found to work more effectively in the ion exchange process if used with protic diluents such as n-octanol and chloroform than nonpolar diluents such as benzene and tetrachloromethane at high pH conditions. TOA in 1-octanol and kerosene was used in the extraction of glycolic acid from an aqueous glycolonitrile hydrolysate.10 The influences of TOA concentration, extraction temperature, and phase volume ratio of organic phase to aqueous phase were determined. The results showed that a lower temperature was beneficial to the extraction of glycolic acid from the hydrolysate. On the other hand, a higher temperature and higher phase ratio were favorable to the regeneration of the acid loaded extractant. TOA mixed in 2ethyl-hexanol was also used to recover acetic acid from woodbased pyrolysis oil derived aqueous phase. This diluent was deemed more effective than MIBK and ethyl acetate due to its higher solvating power, high boiling point, and low affinity for water.11 A separate study using TOA in 2-ethyl-1-hexanol diluent to remove acetic acid from a wood pyrolysis oil obtained similar results.12 About 40 wt % TOA provided the best acetic acid extraction performance. In a preliminary study, we used a low molecular weight organic phase extractant, tributyl phosphate (TBP), to extract HSS from aqueous MEA solution. High molecular weight extractants trioctyl phosphine oxide (TOPO) and TOA were also used. Heptane, chloroform, MEK, and 1-octanol were used as diluents for TBP, TOPO, and TOA. These organic solvents were also used alone for HSS extraction. The use of diluents alone and their mixtures with TBP and TOPO were found to be problematic due to uncontrollable evaporation of most diluents and precipitation of TBP and TOPO during extraction. The formation of a third phase emulsion at high concentrations of TBP and TOPO also occurred resulting in a long equilibration time and difficulty in separating the organic phase from the aqueous phase after extraction. The only extractant/diluent combination that was found promising from the screening process was TOA/1-octanol which could be used without precipitation and evaporation problems. Also, TOA/ octanol was the only reactive extractant system in this screening step which could chemically extract HSSs more effectively than TBP and TOPO. The last two extractants removed HSSs solely based on solubility only via a less effective physical extraction mechanism. We further developed a method to separate HSS from aqueous amine solution using reactive extraction as a new amine reclaiming technique. Not only is extraction capable of removing ppm-level HSS from amine, but also, it can be a potential low-energy demanding process. The reactive extractant selected based on initial screening was TOA while 1octanol was chosen as an active diluent. Aliquat 336 was also chosen to extract HSS from the amine based on the ionexchange process. A chemical modification using OH counterion in Aliquat 336, a two-step extraction, and mixed extractants consisting of Aliquat 336 and TOA were also investigated to improve extraction efficiency. Concentration, extraction temperature, and phase volume ratio of the extractants and aqueous amine phase for extraction were optimized. The regeneration of used extractant using aqueous NaOH solution was also evaluated to determine the optimum conditions (i.e., NaOH concentration, mixing time, temperature, and extractant to NaOH volume phase ratio). 5007

DOI: 10.1021/acs.iecr.6b00035 Ind. Eng. Chem. Res. 2016, 55, 5006−5018

Article

Industrial & Engineering Chemistry Research

2. EXPERIMENTS 2.1. Equipment and Chemicals. The organic phase chemicals used consisted of trioctylamine (TOA; 98% purity) and aliquat336 which were both obtained from Sigma-Aldrich, Canada. 1-Octanol (reagent grade with 99% purity) also from Sigma-Aldrich was used as a diluent to prepare the desired TOA and Aliquat concentrations. Sodium hydroxide (NaOH, reagent grade of 97% purity) and potassium hydroxide (KOH, reagent grade of 87% purity) were both purchased from Fischer Scientific, Canada. Predetermined weights of these bases were mixed with deionized water and used for Aliquat conversion to the hydroxide form (OH−). Mohr’s method was adopted as a procedure to determine % OH− conversion as the equivalent concentration of Cl− replaced by OH−. This determination was carried out by titration using 0.1 kmol/m3 silver nitrate (AgNO3, Sigma-Aldrich) with 0.25 kmol/m3 potassium chromate (K2CrO4, Fischer Scientific, Canada) as an indicator. A hydrophobic filter and a rotary evaporator (model RII, BÜ CHI Labortechnik AG, Switzerland) with vacuum pump (model 2025, Wisconsin, U.S.A.) were used to remove water residue from the converted Aliquat. A similar grade of NaOH used for conversion experiments was also used in the regeneration experiments except that the concentration used for the latter tests ranged from 1 to 4 kmol/m3. In the aqueous phase, concentrated MEA (>99% purity; Fisher Scientific, Canada) was used to prepare 5 kmol/m3 aqueous MEA solution by diluting a predetermined weight of MEA with deionized water. In the case of CO2 loaded experiments, 100% CO2 cylinder (research grade) supplied by Praxair, Regina, Saskatchewan, Canada, was used to obtain the desired CO2 loadings in MEA solution. Details of CO2 preloading procedure can be found in our previous work.13 The exact MEA concentration and CO2 loading calculated based on mol CO2/mol MEA were confirmed by titration with standard 1 N hydrochloric acid (HCl) to methyl orange end point and CO2 displacement solution made of sodium chloride (NaCl), sodium bicarbonate (NaHCO3), and methyl orange adjusted to acidic pH with sulfuric acid (H2SO4). All chemicals for titration were obtained from Fisher Scientific, Canada, except NaCl and NaHCO3 supplied by Sigma-Aldrich, Canada. Formic acid, acetic acid, propionic acid, butyric acid, glycolic acid, oxalic acid, succinic acid, sodium sulfate, sodium sulfite, and sodium thiosulfate (Sigma-Aldrich, Canada) were all reagent grade with more than 99% purity. The acids were used to prepare degraded MEA solution containing HSS used for extraction tests by adding predetermined concentrations into the solution. The HSS stock solution of similar composition/concentrations but without MEA was also prepared and used in the tests for comparison. In addition to HSS, 1000 ppm of N-(2-hydroxyethyl) acetamide, 1-(2hydroxyethyl)-2-imidazolidinone (75% in water), imidazole, N-(2-hydroxyethyl) succinimide (95%), and 2-oxazolidone (98%) all obtained from Sigma-Aldrich, Canada, were also spiked in the MEA solution and used for extraction tests of nonionic degradation products. Capillary electrophoresis (CE) equipped with diode array detector (DAD) (CE, model HP 3D CE, Hewlett-Packard Canada Ltd., Montreal, Quebec, Canada) was used for aqueous phase analysis of HSS and chloride concentrations. The first CE method used in this study was adopted from the literature.14 For all CE analysis, a bare-fused silica capillary column of extended light path (150 μm) with dimensions of 50 μm i.d. ×

645 mm length (560 mm effective length, Agilent Technologies Canada, Mississauga, Ontario, Canada) was used. For the first CE method, background electrolyte (BGE) was trimellitatebased solution prepared by mixing 0.8406 g of trimellitic (1,2,4benzenetricarboxylic) acid (≥99% purity), 0.4000 g of poly(vinyl alcohol) (average molecular weight 30 000− 70 000), and 9.6880 g of trizma base (tris(hydroxymethyl) aminomethane (ultrapure grade) with 400 g of nanopure water. All chemicals were purchased from Sigma-Aldrich, Canada. The electrolyte was degassed in an ultrasonic bath (model 75D, VWR International, Pennsylvania, United States) and filtered through 0.2 μm nylon filter before use. 10% (w/w) hexadimethrine bromide (≥95% purity, Sigma-Aldrich) was also used for capillary coating. The second CE method used organic acid buffer for CE (pH 5.6, Agilent Technologies Canada, Mississauga, Ontario, Canada) to additionally determine acetate and glycolate when Aliquat 336 was used in the extraction. CE water (ultra pure) and 0.1 and 1 M NaOH solution purchased from Agilent Technology, Canada, were also used for capillary flushing. Sodium molybdate (NaMoO4, ≥98% purity, Sigma-Aldrich, Canada) was selected as an internal standard for HSS quantitative analysis. The pH meter used was pH/CON 510 standard model (Oakton, New York, U.S.A.) with a precision of ±0.01 pH unit. All CE samples were prepared using nanopure water. A gas chromatograph−mass spectrometer (GC-MS, model 6890-5073, Hewlett-Packard Canada, Ltd., Montreal, Quebec, Canada) was used to analyze nonionic degradation products (N-(2-hydroxyethyl)acetamide, 1-(2-hydroxyethyl)-2-imidazolidinone, imidazole, N-(2-hydroxyethyl)succinimide, and 2oxazolidone). The chromatographic capillary column was HPInnowax with a cross-linked poly(ethylene glycol) stationary phase. The column had the dimensions of 0.25 μm thickness × 0.25 mm i.d. × 30 m length and was obtained from Agilent Technologies, Canada. The introduction of the sample was done by an autosampler/autoinjector (model 7683, HewlettPackard Canada, Ltd., Montreal, Quebec, Canada). 2.2. Extraction Procedures. A typical run was carried out in a 40 mL extracting bottle. A total of 10 mL of 1000 ppm of HSS spiked in 5 kmol/m3 aqueous MEA or water was loaded into the bottle. In CO2 loaded experiments, 5 kmol/m3 MEA solution was preloaded with the desired CO2 concentration before adding HSS. For 1:1 phase volume ratio (volume ratio of organic extractant phase to aqueous phase of MEA), equal volumes of 10 mL of the two phases was loaded in the extracting bottle. For phase volume 1:2 experiments, the extractant phase used was 10 mL while the aqueous phase was double the volume of the extractant and vice versa for a 2:1 ratio. Details of extraction conditions used in this study are given in Table S1. A magnetic stirrer regulated at 1200 rpm was used to mix the two phases thoroughly at a predetermined time of 10 min or as noted. The mixing condition sufficiently allowed the HSS to transfer from the aqueous phase into the extractant phase at its maximum. In the case of extraction at room temperature, two phases were mixed as they were prepared. For higher temperature experiments, extractant and aqueous phases were separately heated to the desired temperature in a temperature controlled water bath prior to mixing. The two phases were then mixed and stirred at that temperature in the bath throughout the test. After the extraction, the mixture was set to ensure phase equilibration. To completely separate the two phases as well as speed up the separation process, the mixture was subsequently centrifuged at 5008

DOI: 10.1021/acs.iecr.6b00035 Ind. Eng. Chem. Res. 2016, 55, 5006−5018

Article

Industrial & Engineering Chemistry Research

volume ratio of Aliquat and NaOH, the degree of conversion obtained was 87% for the third conversion. The final conversion of 88% was also obtained which was carried out using similar NaOH concentration, temperature, and mixing time to the third conversion. It was only contact time and volume ratio that changed to 21 and 1:1, respectively. A vacuum-rotary evaporator set at 150 mmHg and 353 K was used to remove water residue from the third and the fourth converted Aliquat. The conversion conditions are summarized in Table S2. 2.4. Regeneration Procedures. Procedures for regeneration were similar to extraction experiments. Preloaded HSS Aliquat in 1-octanol kept after extraction was loaded into a 40 mL bottle. NaOH solution of desired concentration (i.e., 1 and 4 kmol/m3) was then added into the bottle. The volume of the organic phase and NaOH solution following the phase ratio of 1:2, 1:1, to 2:1 was used. The two phases were mixed using a magnetic stirrer with the mixing time varied from 5 to 10 min at room temperature, 313 K, and 323 K. Subsequently, a centrifuge set at 4000 rpm and 8 min was used to quickly separate the two phases, and the organic layer of Aliquat was later removed from the bottle. The amount of HSS removed from Aliquat was directly analyzed from the bottom phase of aqueous NaOH by the CE technique. The regeneration efficiency was determined using eq 2 as follows:

4000 rpm for 8 min. The upper layer extractant phase was carefully removed and kept at 277 K to be used later in regeneration studies. The aqueous phase was measured for pH and analyzed by CE techniques for the HSS concentration that remained after extraction. Extraction efficiency was calculated using eq 1: % Extraction efficiency ⎛ [HSS]before − [HSS]after ⎞ =⎜ ⎟ × 100 [HSS]before ⎝ ⎠

(1)

where [HSS]before and [HSS]after denote ppm concentration of HSS before and after extraction in aqueous MEA or water. Similar procedures were applied for the extraction of nonionic degradation products (i.e., N-(2-hydroxyethyl)acetamide, 1-(2-hydroxyethyl)-2-imidazolidinone, imidazole, N-(2-hydroxyethyl)succinimide, and 2-oxazolidone) in 5 kmol/m3 MEA solution. The only difference was that the GC-MS technique was used to quantitatively determine these compounds before and after extraction. 2.3. Chemical Modification of Aliquat 336. Modification of Aliquat 336 to OH form was initially carried out following the procedures given in the literature.6 The procedures were needed to replace the chloride ion (Cl−) of Aliquat 336 with a more effective OH− ion. Also, Aliquat modification helped reduce the introduction of chloride into the amine solution after extraction. Precisely, 2 kmol/m3 KOH was mixed with Aliquat in a 250 mL flask using 105 mL equivalent volume. The flask was shaken vigorously for 5 min at room temperature to maximize the ion exchange process between Cl− and OH− before setting for phase separation. Aliquat was removed from the upper layer and contacted again with freshly prepared KOH using the same steps just described. The procedures were repeated for a total of 10 contact times before residual water was removed from the final Aliquat by filtration through a hydrophobic filter. The degree of conversion (% OH in Aliquat) was measured by analyzing the Cl− concentration that remained in the converted Aliquat using Mohr’s method. The difference of Cl− concentration in the original Aliquat and that of after conversion equivalent to OH− concentration in the modified Aliquat structure was determined as 69%. Attempts were also made to increase the % OH in Aliquat as it could possibly help extraction efficiency. Although KOH was initially used effectively for chemical conversion of Aliquat 336 to OH form, it was decided to continue the rest of the conversion tests of Aliquat with NaOH that exhibited the same basicity and effectiveness required for ion-exchange process of Cl− and OH−. This is because not only did NaOH provide the same level of conversion efficiency, but also, it was much cheaper than KOH and, thus, more economical. Thus, instead of KOH, NaOH was used as it was cheaper but yet had similar basicity to KOH. Instead of 2 kmol/m3, 4 kmol/m3 NaOH was used for conversion. The mixing temperature was raised from room temperature as done with KOH to 313 K. Mixing time was also increased from 5 to 10 min. The Aliquat was repeatedly contacted with fresh NaOH for 15 contact times. The only parameter that was kept similar to the first conversion was Aliquat to NaOH volume ratio of 1:1. The rest of the procedure was followed as described previously. Using Mohr’s analysis, the second conversion with NaOH successfully increased the degree of conversion to 79%; thus, stronger conditions were further assessed. With 5 kmol/m3 NaOH, 333 K temperature, 10 min mixing time, 15 contact times, and 1:2

% Regeneration efficiency ⎛ [HSS]inNaOHafterregeneration ⎞ ⎟⎟ × 100 = ⎜⎜ ⎝ [HSS]inAliquat beforeregeneration ⎠

(2)

2.5. Analysis of HSS Using Capillary Electrophoresis Technique (CE). Two CE methods were used to analyze HSS concentration in all aqueous phases in this study (i.e., HSS in MEA and water with TOA and OH Aliquat extraction, KOH, and NaOH). The first method used was adopted from the literature14 and used to directly determine formate, propionate, butyrate, oxalate, succinate, sulfite, sulfate, and thiosulfate. This method was also used to directly analyze partially overlapped acetate and glycolate in TOA and OH Aliquat extractions. However, samples extracted by original Aliquat 336 (without conversion to Cl form) showed almost complete overlap of these HSS when analyzed by the first CE method. Therefore, the second CE method given later in the section was needed to help analysis of acetate and glycolate. For the CE first method, prior to analysis, the capillary was preconditioned by initially flushing with 1 kmol/m3 NaOH for 20 min followed by 10% hexadimethrine bromide solution for 20 min. The 0.1 kmol/m3 NaOH was subsequently flushed for 10 min to remove excess bromide left by the previous step. CE water was used to flush for an additional 10 min. The capillary was then flushed with trimellitate-based BGE for 20 min. A voltage of −30 kV was finally applied for 10 min to complete the preconditioning process. For an actual analysis, a sample was injected using the hydrodynamic mode in which 50 mbar was applied for 8 s. Negative voltage of 30 kV was used throughout the run. HSS was detected using an indirect UV detection mode set at 350 nm with a bandwidth of 80 nm, with a reference of 240 nm with a 10 nm bandwidth. The capillary was kept at 298 K, also throughout the analysis. In between runs the capillary was flushed with BGE for 5 min. Since the aqueous phase was only analyzed for HSS concentration, a material balance was used to determine HSS concentration in the extractant phase. 5009

DOI: 10.1021/acs.iecr.6b00035 Ind. Eng. Chem. Res. 2016, 55, 5006−5018

Article

Industrial & Engineering Chemistry Research

complete elution of all degradation products. A constant flow rate mode was used with helium carrier gas regulated at a flow rate of 1 mL/min. The GC-MS interface, MS quad, MS source, and EM voltage were kept at 523 K, 423 K, 503 K, and 1858, respectively. The MS scan mode used had a mass range from 10 to 300 Da. Prior to analysis, standard curves of the degradation products were made. Standard mixture containing 100 to 1000 ppm of N-(2-hydroxyethyl) acetamide, 1-(2-hydroxyethyl)-2imidazolidinone, imidazole, N-(2-hydroxyethyl) succinimide, and 2-oxazolidone were repeatedly analyzed three times, after which their areas were averaged. A plot of areas and the corresponding concentrations was generated for all compounds. The exact concentrations of the degradation products in samples were identified using appropriate calibration curves. The accuracy of GC-MS technique reported in terms of standard error was less than 5%.

For the second method, the same capillary was initially conditioned by flushing by a ready-made organic acid buffer of pH 5.6 BGE for 15 min. Hydrodynamic mode of sample introduction was still used by applying pressure of 50 mbar for 2 and 4 s to sample and BGE vials, respectively. Run voltage and temperature were −25 kV and 293 K, respectively. DAD signal set at 350 nm with a bandwidth of 20 nm with reference of 200 nm with a bandwidth of 10 nm was used for HSS detection. The capillary was flushed for 4 min with BGE in between analysis. The analysis time was also 10 min. 2.5.1. Quantitative Analysis of HSS. Quantitative analysis of HSS was obtained using internal standard calibration curves. Prior to sample analysis, stock solution of standard 100 ppm formate, acetate, propionate, butyrate, glycolate, succinate, oxalate, sulfite, sulfate, and thiosulfate was prepared in 0.5 kmol/m3 MEA solution. The stock solution was carefully diluted to 10, 20, 30, 40, 50, and 75 ppm using a predetermined volume of 0.5 kmol/m3 MEA. Each standard including 100 ppm stock was then spiked with 200 ppm molybdate internal standard and run using the first CE condition described earlier. All standards were analyzed three times to check for repeatability. Standard curves were all generated by plotting averaged corrected peak area ratios of standard HSS and molybdate against corresponding HSS concentrations. For samples extracted by TOA and OH Aliquat, all HSS in the aqueous phase were analyzed by these area-concentration curves, except acetate and glycolate whose analysis were done using additional curves made by height ratio. Extraction samples were prepared by diluting with nanopure water using a dilution ratio of 1 in 10 and later spiked with 200 ppm molybdate. This ratio was selected so that the MEA concentration in the diluted sample remained the same as that of the standards. Injections were done twice, and the area ratios (height ratio for acetate and glycolate) of HSS and molybdate were averaged. The exact concentrations of HSS in the samples were obtained by comparison of their ratios with the corresponding standard curves. The error involved was 5− 10%. Since acetate and glycolate extracted by original aliquate 336 (extractant with Cl form) could not be completely separated by the first CE method, the second CE method was used to help determine their concentrations. Initially, the sample was run with the second CE method to determine the acetate concentration by calibrating its area with its standard curve also generated by the second CE procedure. To determine glycolate concentration, the acetate concentration obtained earlier was used to find its corresponding area from its curve generated by the first CE method. The first CE method was then used to run the same sample from which its overlapped area of acetate and glycolate was obtained. Subtraction of the combined area with acetate area found from the previous step gave the glycolate area. This area was equivalent to the peak area of glycolate that would have responded to the analysis, if present alone in the sample. The glycolate area was finally found by calibrating for the exact concentration with its calibration curve produced from the first CE method. 2.6. Analysis of Nonionic Degradation Products Using Gas Chromatography−Mass Spectrometry Technique (GC-MS). For a typical GC-MS analysis, 1 μL of sample was injected at the GC inlet set at 523 K using a split injection mode with a split ratio of 30 to 1. The GC oven was initially set at 373 K and ramped to 513 K at the rate of 280 K/min. The temperature was kept at 513 for an additional 10 min to ensure

3. RESULTS AND DISCUSSION 3.1. Extraction Using TOA and Aliquat 336. 3.1.1. Extraction of HSS in Water. Initially, aqueous HSS solution without MEA was used to study the effect of concentrations of TOA and Aliquat 336 and also to determine the optimum concentration of the extractants. The results were used to establish the effect of MEA. TOA and Aliquat 336 were investigated using concentrations in the range of 0.2 to 1 kmol/ m3 in 1-octanol diluent. Higher concentrations were not used since they were found unsuitable, especially for Aliquat 336. Concentrations higher than 1 kmol/m3 Aliquat 336 generated emulsion during extraction. The formation of emulsion made separation between extractant and aqueous phase extremely difficult. The HSS solution used for extraction contained 1000 ppm of formate, acetate, propionate, butyrate, oxalate, succinate, and glycolate. Sulfate, sulfite, and thiosulfate of equivalent concentration were also added to represent SO2derived HSS. The extraction temperature, extractant to aqueous phase volume ratio, and mixing time were kept at room temperature, 1:1, and 10 min, respectively, unless otherwise noted. Figure S1 shows the concentration extraction efficiency of HSS with TOA. The extraction efficiency for HSS was found to increase as the concentration of TOA increased from 0.2 to 0.6 kmol/m3. The maximum efficiency was reached at 0.6 kmol/m3 and found to be in the range of 58 to 96%. The extraction efficiency slightly decreased at higher concentrations (0.7 to 1.0 kmol/m3) possibly due to increased TOA viscosity limiting mass transfer of HSS from the aqueous phase to the extractant phase.10 The effect of Aliquat concentration is given in Figure S2. The extraction efficiency of most HSS was found to increase steadily from 0.2 kmol/m3 until it reached the maximum at 1 kmol/m3. The efficiency obtained at this concentration was in the range of 36 to 91% for various HSS. Data shown in Figures S1 and S2 for TOA and Aliquat, respectively, were also correlated using the Langmuir based expression. This was done to check for a possible interaction of each HSS and the extractants for a future study of the extraction mechanism. The equation used for the correlation is shown in eq 3: Extraction efficiency = 5010

a[Extractant] 1 + b[Extractant]

(3)

DOI: 10.1021/acs.iecr.6b00035 Ind. Eng. Chem. Res. 2016, 55, 5006−5018

Article

Industrial & Engineering Chemistry Research where [Extractant] is concentration of TOA or Aliquat (kmol/ m3), and the coefficients a and b represent the interaction of HSS and the extractant derived from the HSS chemical nature. For TOA, the data in Figure S1 did not quite follow the Langmuir relationship, evidently seen from a low curve fitting R2 value for all HSSs, which averaged to only 0.67 as shown in Table S3. Unlike the Aliquat system, its extraction data in Figure S2 could be adequately described by the Langmuir equation with an average R2 value above 0.9. The coefficients a and b in the Aliquat system were also calculated and reported in Table S3 for all HSS which varied by HSS type. The variations of a and b indicate a difference in interaction behavior of Aliquat and HSS, possibly due to the chemical nature of each salt (e.g., hydrophobicity and electronegativity). Although TOA did not quite obey the Langmuir relationship, this extractant’s coefficients a and b for each HSS were also given for comparison. More work needs to be done in order to determine specifically the type of interaction HSS will have with the extractants. 3.1.2. Extraction of HSS in Aqueous MEA Solution. The extraction technique was also carried out for the removal of HSS in 5 kmol/m3 aqueous MEA solution. Similar concentrations of HSS as used in Section 3.1.1 were also used in the MEA solution. The concentrations of TOA and Aliquat 336 in 1-octanol were also varied from 0 to 1 kmol/m3. The extraction efficiencies of HSS using TOA and Aliquat 336 are shown in Figures S3 and S4, respectively. In Figure S3, all HSSs show a similar trend in which the efficiency increased as the concentration of TOA increased from 0 kmol/m3 until the maximum extraction was reached at 0.6 kmol/m3 concentration. At the optimum concentration, the extraction efficiency was found in the range of 44 to 86%. Unlike the non-MEA system shown previously in Figure S1, the efficiency dropped dramatically when extraction was performed using higher concentrations of TOA (e.g., 1 kmol/m3). Similar explanation given for the non-MEA system in which high viscosity limited the mass transfer could be used to explain the decrease in extraction efficiency with increase of concentration of TOA beyond 0.6 kmol/m3. In addition, the presence of MEA also affected the extraction of all HSS explained in more details in Section 3.1.3. The effect of Aliquat 336 concentration is shown in Figure S4. The increase in extraction efficiency of all HSS was clearly seen when the Aliquat concentration was increased from 0 to 0.4 kmol/m3. The effect became smaller after 0.4 kmol/m3 concentration before reaching a maximum at the concentration of 1 kmol/m3 Aliquat 336. The extraction efficiency obtained was in the range of 40 to 88%. 3.1.3. Effect of MEA. 3.1.3.1. Extraction with TOA. MEA played a significant role in the extraction of HSS using TOA and Aliquat 336 shown previously in Sections 3.1.1 and 3.1.2. Figure 1 shows a comparison of the HSS extraction in water and aqueous MEA solution using TOA. It is clear that the efficiency of all HSS decreased when MEA was present in the system. The extraction efficiency was reduced to as much as 31% for sulfate, thiosulfate, and oxalate while butyrate was affected the least with a 9% decrease. The remaining HSS showed a decrease between 11 and 25%. To understand the decrease of TOA extraction performance when MEA was present, an extraction mechanism of HSS by TOA needs to be established. TOA is an aliphatic amine which extracts acids from an aqueous phase by forming an acid−base complex with the

Figure 1. Effect of MEA on extraction of HSS using 0.6 kmol/m3 of TOA in 1-octanol (room temperature, 10 min mixing time, and 1:1 phase ratio).

undissociated acids5,15 The extraction mechanism corresponding to the acid−base complex formation by hydrogen bonding for HSS extraction can be given as in eq 4. (4)

TOA + HA ↔ TOA−HA

The extraction by ion-pair formation of TOA and acid is also suggested as a possible mechanism.8 The reaction corresponding to HSS extraction in the current study is given in eq 5. TOA + H+ + A− ↔ TOA−H+A−

(5) −

where species with an overbar, HA and A , represent species in organic extractant phase, undissociated HSS, and dissociated HSS, respectively. Dissociation of HSS and pH of the aqueous phase play a significant role in determining the extraction mechanism and, more importantly, the extraction efficiency. HSS can exist in undissociated or dissociated forms (i.e., HA or A−) following the reaction given in eq 6: HA ↔ H+ + A−

(6)

Equation 7 can be used to relate the pH of the aqueous phase to the concentration ratio of A− and HA: pH = pK a + log

[A−] [HA]

(7)



where Ka, [A ], and [HA] are the acid dissociation constant of HSS and concentrations of dissociated and undissociated HSS, respectively. For the extraction without MEA, the pH of the aqueous phase was measured at 2.7. Based on eq 7 alone, the organic HSS used in this study except oxalate existed at 10 to 100 times more in the undissociated form than the dissociated. The stronger acidic oxalate and inorganic HSS (i.e., sulfate and thiosulfate) showed the opposite trend where the more dissociated form was present. This could suggest a difference in the dominant extraction mechanism. When HSS was added in 5 kmol/m3 MEA solution, the aqueous phase pH changed to 11.5. At this pH, the organic HSS was present mostly in the dissociated form. It was reported that TOA preferentially extracted undissociated 5011

DOI: 10.1021/acs.iecr.6b00035 Ind. Eng. Chem. Res. 2016, 55, 5006−5018

Article

Industrial & Engineering Chemistry Research acid;15 thus, this could account for the reduction of the extraction efficiency of the organic HSS as shown in Figure 1. Oxalate and inorganic HSS were also affected at this pH. The dramatic increase of [A−] as compared to [HA] could also be blamed for the decrease in the extraction efficiency. This trend is confirmed by the literature in which the efficiency of both mechanisms was found to deteriorate with an increase of aqueous phase pH.8 The negative effect of MEA could be twofold. In addition to the decrease of extraction efficiency induced by pH change, the strong bonding of MEA and HSS could also contribute to the reduced TOA extraction performance. This effect is more prominent for stronger acidic species (i.e., oxalate and inorganic sulfate and thiosulfate) since they chemically bond more strongly to MEA. Therefore, the larger reduction in extraction efficiency was observed (over 30% for oxalate, sulfate, and thiosulfate). The remaining HSS (i.e., formate, acetate, propionate, butyrate, glycolate, and succinate) experienced less bonding interaction with MEA due to a much lesser acidity. Thus, smaller decrease in extraction efficiency (less than 20%) was seen when MEA was present with these HSS. 3.1.3.2. Extraction with Aliquat 336. Figure 2 shows the MEA effect on the extractability of HSS at room temperature

acetate, propionate, succinate, and glycolate were observed because their existence in the undissociated form was unfavorable to Aliquat extraction. Butyrate was the exception possibly due to its long-chained molecule and higher hydrophobicity, which allowed a better solvation by Aliquat 336 and 1-octanol, thereby giving a better extraction.5 It is clear from Figure 2 that MEA also affected the performance of Aliquat 336. A strong interaction of MEA and oxalate, thiosulfate, and sulfate similar to what was explained in the case with TOA could have played a major role in the decrease of extraction efficiency respectively by 19%, 28%, and 30%. On the other hand, the extractability of organic HSS increased to as high as 12% with MEA, especially for formate and glycolate and to a lesser extent acetate and succinate. The increase of pH to 11.5 in MEA solution could be responsible because it totally shifted eq 6 to the right resulting in these organic HSS existing mostly in the dissociated form which was preferred for Aliquat 336 extraction. Although MEA interacted with these organic HSS, its effect would have been less pronounced than that of the previous HSS group (i.e., oxalate, thiosulfate, and sulfate) due to a much weaker interaction with MEA. The effect of pH, therefore, could have overcome the MEA interaction effect, thus allowing formate, acetate, succinate, and glycolate to be extracted more into the Aliquat. MEA did not have a significant effect on propionate and butyrate though it helped promote formation of their dissociated form. In our view, the extraction capacity limit of both HSS must have been reached and no further increase in extraction efficiency could be obtained. 3.2. Selection of Extractant for CO2 Absorption Process. On the basis of the results obtained so far, TOA worked much more effectively than Aliquat 336 at a low pH (essentially the system without MEA) as plotted in Figure S5 for a clear comparison. Figure S6 also shows a side by side comparison of HSS extraction efficiency in MEA solution using TOA and Aliquat 336. When MEA was present, a superior extraction efficiency of Aliquat 336 to TOA was observed for sulfate and oxalate. Formate, acetate, propionate, butyrate, glycolate, and succinate showed somewhat similar affinity to both extractrants. Thiosulfate was the only HSS extracted for which TOA was 12% better than Aliquat 336 in the MEA system. Although the extraction efficiencies of TOA and Aliquat 336 were generally close in aqueous MEA solution, Aliquat 336 was selected for further investigation for two reasons. First, the extraction efficiency of Aliquat 336 could be improved if the Aliquat 336 was chemically modified by replacing its Cl− for a more effective OH− counterion. Aliquat 336 is a quaternary ammonium based anion exchange extractant which its Cl− exchanges with HSS in aqueous amine phase during extraction. Based on the degree of affinity, hydroxide (OH−) is found more suitable in terms of ease of ion exchange ability. It has been reported also that the ion-exchange selectivity of OH− by quaternary amine resins was lower and the interaction between the resins and OH− was also weaker than that of Cl−; thus, ion exchange reaction could occur more easily.16 Therefore, replacing with OH− would help increase the HSS extractability from the aqueous amine phase. Second, the use of Aliquat 336 could introduce Cl− into the amine solvent via anion-exchange reaction which could potentially induce corrosion in the CO2 absorption plant. Thus, chemical modification of Aliquat 336 described earlier was also required to prevent chloride contamination in the amine solution. Further extraction with

Figure 2. Effect of MEA on extraction of HSS using 1.0 kmol/m3 of Aliquat 336 in 1-octanol (room temperature, 10 min mixing time, and 1:1 phase ratio).

by 1 kmol/m3 Aliquat 336 in 1-octanol. As a quaternary ammonium salt, Aliquat 336 can extract HSS based on the ion exchange reaction as follows: R 4N+Cl− + A− ↔ R 4N+A− + Cl− +



(8)



where R4N Cl and A represent Aliquat 336 and dissociated HSS. In the non-MEA system, oxalate, sulfate, and thiosulfate still yielded a high percentage of extraction efficiency which was higher than those of the remaining HSS except butyrate, respectively measured at 89%, 84%, and 82%. On the basis of previous discussion, inorganic HSS and oxalate, though at low pH, existed mostly in the dissociated form, thus favoring the ion-exchange reaction of Aliquat 336 as given in eq 8, while the rest shows efficiency only between 36 to 75% except butyrate for which 91% was extracted. The low extractability of formate, 5012

DOI: 10.1021/acs.iecr.6b00035 Ind. Eng. Chem. Res. 2016, 55, 5006−5018

Article

Industrial & Engineering Chemistry Research

except glycolate; its efficiency was increased from 69% to 81%. Also, in Figure S10, a mixture of the two extractants using their optimum concentrations also performed well showing an extraction efficiency close to that of 69% converted Aliquat. Above 90% extraction efficiency was obtained for most HSS with glycolate and acetate being only above 80% efficiency range. Though, the extraction abilities using two-step and mixed extractants were equal and could be an alternative technique to using 69% OH− modified Aliquat. It must be noted that these two techniques generally would require more steps in doing extraction. One benefit gained from using twostep extraction and mixed extractants, however, is their ability to manage chloride contamination explained in detail in the next section. 3.3.3. Management of Chloride Contamination in Amine Solution. As mentioned in Section 3.2, the chemical modification of Aliquat 336 also served to reduce Cl− contamination in MEA solution during extraction. Table 1

TOA was also carried out. However, it was only done by the use of TOA and modified Aliquat in two-step or mixed extraction to investigate a possible synergistic effect and further reduction of Cl− contamination. 3.3. Improvement of HSS Extraction Efficiency. 3.3.1. OH− Modified Aliquat. Attempts were made to increase the extraction efficiency of HSS in MEA solution using chemical modification done by replacing Cl− with OH− in the structure of Aliquat 336. The degrees of conversion obtained were 69%, 79%, 87%, and 88%. To ensure that the best working concentration for the modified Aliquat still remained at 1 kmol/m3, 69% OH− Aliquat was initially used to confirm this condition. Figure S7 shows the extraction efficiency of HSS in 5 kmol/m3 MEA solution using various concentrations of 69% OH− Aliquat in 1-octanol. A similar trend obtained previously for the original Aliquat was also observed. The efficiency dramatically increased from 0 to 0.4 kmol/m3 after which it began to slow down and finally reached the maximum extractability at 1 kmol/m3. 88% OH− Aliquat gave the same trend which 1 kmol/m3 concentration was also found to be the optimum as shown in Figure S8. Therefore, this concentration was still the best and used for the remaining modified aliquats to determine the extraction efficiency. Figure S9 compares the effect of different % OH− in Aliquat on HSS extraction from 5 kmol/m3 MEA solution at room temperature. The original Aliquat 336 was also included for comparison. At the optimum concentration of 1 kmol/m3, 69% conversion to OH− clearly improved the extractability of the original Aliquat. The extraction efficiency was increased by 45, 44, 42, 39, 38, 36, 29, 22, 20, and 9%, respectively for sulfite, acetate, succinate, sulfate, thiosulfate, formate, oxalate, propionate, glycolate, and butyrate which most HSS reached 90% extraction efficiency with 69% OH− Aliquat. This confirmed the effectiveness of OH− Aliquat in the extraction process with HSS. It is also clear that the difficult to remove HSS including inorganic sulfate and thiosulfate benefited the most from the OH− form of extractant, though they bound strongly with MEA in the aqueous solution as analyzed previously in Figure 2. Highly hydrophilic HSS such as formate and acetate were also extracted into the organic phase much better with the OH− modified Aliquat. OH− Aliquat of 79%, 87%, and 88% were additionally tested, also using 1 kmol/m3 in 1-octanol. Also shown in Figure S9, an increase of % OH− in the Aliquat structure from 69 to 79, 87, and 88% did not give any significant increase in the extractability of HSS. In fact, the efficiency of most HSS still remained approximately 90% similar to those obtained from 69% conversion. The anion-exchange equilibrium between OH− and HSS must have reached its maximum capacity at about 60 to 70% OH−. Further increase of extraction with higher % OH− was therefore not possible under the prevalent extraction condition. 3.3.2. Two-Step Extraction and Mixed OH− Aliquat and TOA. Two-step extraction was carried out by initially applying 69% OH− Aliquat of 1 kmol/m3 to extract HSS from aqueous MEA solution. 0.6 kmol/m3 TOA was then used in the second step to extract additional HSS from the aqueous MEA solution. The efficiency of the two-step extraction is shown in Figure S10. The data for TOA and 69% OH− Aliquat alone are also given in the same figure for comparison. The two-step extraction clearly extracted HSS much better than TOA alone giving the removal efficiency in the range of 81−98%. However, it showed similar extractability to 69% OH− Aliquat for all HSS

Table 1. Cl− Concentration in MEA Solution after Extraction with Various OH Modified Aliquats OH modified Aliquat (%)

chloride concentration after extraction (ppm, ±10)

69 79 87 88

218 188 130 120

shows the concentration of Cl− released into MEA solution after the extraction using various OH− modified aliquats. Less concentration of Cl− was found if a higher % conversion to OH− of Aliquat was used. The Cl− contamination was a result of competitive ion exchange reaction of the remaining Cl− and OH− in the modified Aliquat structure for HSS in the amine solution. This again emphasizes the need for modification of Aliquat in terms of reducing Cl− contamination in MEA solution. The mixed extractants and two-step extraction techniques were also tested for Cl− removal. To clearly see the Cl− removal efficiency, competitive reaction of HSS-Aliquat was eliminated. Therefore, tests were carried out with only Cl− in MEA solution. The 5 kmol/m3 MEA solution initially containing 60 ppm of Cl− was used for extraction. The 87% OH− Aliquat and TOA at 1 and 0.6 kmol/m3, respectively, was used in both extraction techniques. Table 2 shows concentration of Cl− in MEA solution before and after mixed extractant and two-step extractions. The results showed that the use of the mixed extractants reduced Cl− concentration from 60 to 36 ppm, which accounted for 40% removal. The two step extraction was found to be superior to the mixed extractants because it Table 2. Removal of Cl− from 5 kmol/m3 MEA Solution Using Two-Step Extraction and Mixed Extractant of 0.6 kmol/m3 of TOA and 1 kmol/m3 of 87% OH Aliquat at Room Temperature, 10 min Mixing Time, and 1:1 Phase Ratio chloride concentration (ppm) after extraction

5013

before extraction

mixed extractant

two-step extraction

61

36

16 DOI: 10.1021/acs.iecr.6b00035 Ind. Eng. Chem. Res. 2016, 55, 5006−5018

Article

Industrial & Engineering Chemistry Research reduced the concentration of Cl− to 16 ppm equivalent to 73% removal efficiency. This shows that Cl− contamination is manageable with both the two-step extraction and mixed extractants, but more so by the two-step extraction. 3.4. Optimization of Extraction Parameters. 3.4.1. E€ect of Phase Ratio of Aliquat and Aqueous MEA Solution. The 69% Aliquat at 1 kmol/m3 was used to determine the most effective Aliquat/aqueous amine phase ratio. Figure 3 shows the

of temperature on extraction of HSS from MEA solution is illustrated in Figure 4. Extraction efficiency of HSS remained

Figure 4. Effect of temperature on HSS extraction in 5 kmol/m3 MEA using 1 kmol/m3 69% OH Aliquat (10 min mixing time and 1:1 phase ratio).

unchanged throughout the tested temperature range. The exception only applied to glycolate, which was extracted better at a higher temperature (i.e., 338 K). However, temperature did not significantly affect the overall extractability of the Aliquat. Although a reaction of acid-amine complex in organic extractant phase is exothermic. The stability of the complex was less affected by temperature due to the stronger basicity of the amine in removing the acids at any given temperatures.6,17 In fact, temperature independent extraction of HSS found from the test could be favorable to the CO2 capture process, since temperature in the study range has no effect. The capture plant was given more choices in choosing locations within the capture process at which to apply the extraction technique. Possible locations where amine slipstream could be withdrawn and treated by the Aliquat based extraction process included lean MEA stream after the rich/lean heat exchange either with or without further cooling. 3.5. Effect of CO2 Loading. It is crucial to determine effect of CO2 loading on efficiency of HSS extraction with the modified Aliquat. The effect of CO2 loading was demonstrated using the 88% conversion extractant at the optimum concentration of 1 kmol/m3. Room temperature and 1:1 phase ratio were selected for this study. Figure 5 shows that the increase of CO2 loading from 0 to 0.1, 0.2, and 0.3 mol/mol MEA respectively decreased the extractability of the extractant and, thus, decreased % extraction of HSS. When MEA solution was loaded with CO2 at 0, 0.1, 0.2, and 0.3 mol/mol, its pH was reduced. Thus, the working ability of the Aliquat was also reduced as explained earlier. The decrease of pH with CO2 loading could be used to account for the reduced HSS extraction efficiency because the ion-exchange mechanism was made less favorable. Also, having CO2 in MEA solution is known to generate various anions such as carbonate (CO32−), bicarbonate (HCO 3 − ), and carbamate (OHCH 2 CH 2 NHCOO−). These CO2 induced anions potentially competed for Aliquat, thus reducing HSS extraction efficiency of the system. The study of the CO2 the effect provided a crucial information to help select the proper location of the extraction unit within the capture process. To maximize the extractability

Figure 3. Effect of phase ratio on HSS extraction in 5 kmol/m3 MEA using 1 kmol/m3 69% OH Aliquat (room temperature, 10 min mixing time, and 1:1 phase ratio).

effect of phase ratio defined as volume ratio of Aliquat in 1octanol to aqueous MEA solution. Two phase ratios of 1:2 and 2:1 were tested and compared with previous data of 1:1. The phase ratio of 1:2 was found to result in a decrease in the extraction efficiency of all HSS when compared with the base run using 1:1 phase ratio. It was also the least effective among the ratios tested giving the least efficiency for all HSS. For the 1:2 ratio, most HSS showed extraction efficiencies were well below 90% (e.g., 41% for acetate and 48% for glycolate were extracted from aqueous MEA solution). Oxalate and succinate were the only HSS for which more than 90% were extracted into the Aliquat. On the other hand, the extraction efficiency of all the HSS improved after switching from phase ratio 1:2 to 2:1. Doubling the volume of Aliquat was able to extract over 94% of most of the HSS from the amine phase. However, the 2:1 phase ratio only showed either a little increase or unimportant change in the extraction efficiency over a 1:1 ratio with the exception of glycolate. A marked improvement was seen from glycolate in which its extraction efficiency increased from 69% to 91% by switching the phase ratio from 1:1 to 2:1. This increase in extraction efficiency was derived from a higher Aliquat content, which increased the extraction capacity of the system. In our view, although the phase ratio of 2:1 was able to reach maximum extractability, the ratio of 1:1 still performed fairly equivalent and was a preferred technique due to its cost saving benefit. In addition, the use of less extractant also helped to minimize the generation and disposal of waste from the extraction process. 3.4.2. Extraction Temperature. Extraction at temperatures of 313, 323, and 338 K were carried out and compared with the previous run at room temperature. The 69% OH Aliquat with 1:1 phase ratio was still used throughout this section. The effect 5014

DOI: 10.1021/acs.iecr.6b00035 Ind. Eng. Chem. Res. 2016, 55, 5006−5018

Article

Industrial & Engineering Chemistry Research

Figure 5. Effect of CO2 loading on extraction of HSS in 5 kmol/m3 MEA using 1 kmol/m3 88% OH Aliquat 336 (room temperature, 10 min mixing time, and 1:1 phase ratio).

Figure 6. Effect of NaOH concentration on regeneration efficiency using HSS preloaded 79% OH Aliquat of 1 kmol/m3 (316 K, 10 min mixing time, and 1:2 phase ratio).

of the Aliquat and minimize the undesirable competitive reactions, the MEA stream to be extracted should be as lean as possible such as that before and after the lean/rich heat exchanger to the absorber column. 3.6. Regeneration of Used Aliquat. In order to recycle the Aliquat in the extraction process, regeneration of used extractant is desired. After HSSs have been removed from MEA solution by the Aliquat 336, cleaner MEA with only trace amounts of HSSs could be returned for the capture of CO2, thereby ensuring that the amine could be used effectively in the absorption process again. The Aliquat which is now loaded with HSSs is sent for regeneration using aqueous solution spiked NaOH and conditions optimized in this section, to recover the extractant for reuse. HSSs in aqueous NaOH which could exist in the form of sodium salt are then sent for disposal in the same manner as other wastes generated from the capture process. This section aims at determining the best regeneration conditions for replenishment of the used Aliquat. Based on initial screening, NaOH was selected as a regenerant due to its strong basicity. HSS preloaded Aliquat previously obtained from the extraction experiments was used throughout this section. Preloaded HSS concentrations in OH Aliquat used in the regeneration study are given in Table S4. Mixing time, temperature, and volume phase ratio of used Aliquat and NaOH are also optimized. 3.6.1. Effect of NaOH Concentration. The effect of NaOH concentration was evaluated using 1 and 4 kmol/m3. Mixing time, temperature, and phase ratio were respectively set at 10 min, 316 K, and 1:2. The 79% OH Aliquat preloaded with known concentration of HSS shown in Table S4 was used for regeneration. Regeneration efficiency was calculated based on HSS concentration released into NaOH phase after contacting with used Aliquat. Figure 6 shows clearly that 4 kmol/m3 NaOH was over 50% more effective than its 1 kmol/m3 concentration. At 4 kmol/m3, regeneration efficiency ranging between 59% to 89% was obtained except for sulfite, propionate, and butyrate which respectively showed 48, 15, and 5% efficiency. NaOH of 1 kmol/m3 produced only less than 39% efficiency for most HSS. 3.6.2. Effect of Mixing Time. Mixing time also affected regeneration efficiency as shown in Figure 7. The test was

Figure 7. Effect of mixing time on regeneration efficiency using HSS preloaded 69% OH Aliquat of 1 kmol/m3 (298 K, 1:1 phase ratio, and 4 kmol/m3 NaOH).

carried out using two mixing times of 5 and 10 min. NaOH concentration, temperature, and phase ratio were set at 4 kmol/ m3, 298 K, and 1:1, respectively. Preloaded HSS 69% OH Aliquat was used for regeneration. An increase of mixing time from 5 to 10 min increased the ability of the NaOH to better regenerate the used extractant. The effect was more pronounced for thiosulate and sulfite. A longer mixing time basically provided more contact time of the ion-exchange process of HSS and OH−. 3.6.3. Effect of Temperature. Figure 8 shows the effect of regeneration temperature with 4 kmol/m3 NaOH, mixing time of 10 min, and phase ratio of 1:2. The preloaded HSS Aliquat similar to the previous section was used to study the effect of temperature. Temperatures tested were 298, 313, and 323 K. The regeneration of used Aliquat was found to be temperature dependent in which all HSSs followed the Arrhenius relationship as shown eq 9 as follows: Regeneration efficiency = A e−Ea / RT [HSS−Aliquat]n 5015

(9)

DOI: 10.1021/acs.iecr.6b00035 Ind. Eng. Chem. Res. 2016, 55, 5006−5018

Article

Industrial & Engineering Chemistry Research

Figure 8. Effect of temperature on regeneration efficiency using HSS preloaded 69% OH Aliquat of 1 kmol/m3 (10 min mixing time, 1:2 phase ratio, and 4 kmol/m3 NaOH).

Figure 9. Effect of phase ratio on regeneration efficiency using HSS preloaded 69% OH Aliquat of 1 kmol/m3 (298 K, 10 min mixing time, and 4 kmol/m3 NaOH).

efficiencies of sulfite and acetate were also reduced and were the lowest among volume ratios tested in this study. The only benefit of using the 1:2 phase ratio was for glycolate, for which it showed the highest regeneration efficiency. Figure S12 also confirms the effectiveness of 1:1 ratio. The regeneration efficiency was obtained also by using the optimum conditions found for the other parameters (i.e., 4 kmol/m3 NaOH, 10 min mixing time, and 323 K temperature). The 88% OH Aliquat preloaded with HSS was used to test the optimum conditions. The result is compared with the run of 1:2 phase ratio of 69% OH Aliquat to NaOH. It is clear that 1:1 phase ratio worked more effectively than the rest of the conditions. 3.7. Extraction of Non-Ionic Degradation Products. Though it was not the objective of this study, the extraction technique developed for HSS was also applied to the nonionic degradation products of MEA. The 5 kmol/m3 MEA solution spiked with 1000 ppm of major degradation products, namely, imidazole, N-(2-hydroxyethyl)acetamide, 2-oxazolidone, N-(2hydroxyethyl)succinimide, and 1-(2-hydroxyethyl)-2-imidazolidinone. The extraction was carried out with 1 kmol/m3 88% OH Aliquat and 1:1 phase ratio of Aliquat and aqueous amine and at room temperature. Table 3 shows concentrations of all

where A is the preexponential constant (units depend on n), Ea is the activation energy (J/mol), R is the gas constant (8.314 J/ (mol·K)), and T is the regeneration temperature (K). [HSS− Aliquat] is concentration of the HSS and Aliquat complex before regeneration (kmol/m3). Finally, n is the reaction order with respect to the HSS−Aliquat complex. It must be noted that [HSS−Aliquat] of all HSS in all temperatures were constant. Thus, this term in eq 9 was a constant and was thus combined with A to become a single constant (i.e., k′ as shown in Figure S11) for the regression to show the Arrhenius based temperature dependency as a straight line plotted between ln k′ and 1/T in Figure S11 with an average R2 value being virtually 1. Results showed that a higher temperature was required to better regenerate the used Aliquat. Regeneration was most effective using the highest temperature of 323 K showing over 50% efficiency for most HSS. Propionate and butyrate were the exceptions with only 12 and 6% efficiency. When regeneration at 313 K was used, the efficiency clearly dropped. A huge decrease was further observed when regeneration was carried out at 298 K. This indicates the endothermic nature of the regeneration process of Aliquat using NaOH. 3.6.4. Effect of Phase Ratio. The optimum Aliquat/NaOH phase ratio was determined using 69% OH Aliquat, 4 kmol/m3 NaOH, 10 min mixing time, and 298 K temperature. In Figure 9, regeneration using a phase ratio of 2:1 shows a small range of efficiency in which most HSS was extracted back into NaOH phase by only 30%. The highest percentage that this phase ratio produced was 38% for sulfite. A further test was carried out with 1:1 Aliquat to NaOH, which was found to be optimum. The reduction of Aliquat volume showed significant improvement on regeneration efficiency of many HSS. Thiosulfate, sulfate, oxalate, formate, and succinate showed over 100% increase in terms of regeneration efficiency compared to those obtained from the 2:1 phase ratio. Acetate was the only HSS for which the efficiency decreased with 1:1 phase ratio while sulfite was not affected by the phase ratio change. A test was further attempted by changing the phase ratio to 1:2. Figure 9 shows that an increase of NaOH phase volume had a negative effect by reducing the regeneration efficiency of most HSS to as low as those obtained by the previous 2:1 volume ratio. The

Table 3. Extraction of Non-Ionic Degradation Products in 5 kmol/m3 MEA Solution Using 1 kmol/m3 88% OH Aliquat at Room Temperature, 10 min Mixing Time, and 1:1 Phase Ratio degradation products

extraction efficiency (%, ±5)

imidazole 2-oxazolidone N-(2-hydroxyethyl)succinimide N-(2-hydroxyethyl)acetamide 1-(2-hydroxyethyl)-2-imidazolidinone

96 98 99 43 49

degradation products in MEA and the corresponding % extraction efficiency after extraction. The Aliquat also successfully extracted imidazole, 2-oxazolidone, and N-(2hydroxyethyl) succinimide from aqueous MEA solution with the extraction efficiencies respectively being 96, 98, and 99%. N-(2-hydroxyethyl) acetamide and 1-(2-hydroxyethyl)-2-imida5016

DOI: 10.1021/acs.iecr.6b00035 Ind. Eng. Chem. Res. 2016, 55, 5006−5018

Article

Industrial & Engineering Chemistry Research

8. The phase ratio of Aliquat volume to aqueous MEA volume of 1:1 was found to be optimum in terms of both HSS extraction efficiency and economic considerations. Since the effect of temperature in the studied range (i.e., 313−338 K) was found to be insignificant, the amine stream after the rich/lean heat exchanger could be extracted either before or after the cooling process. 9. Higher CO2 loadings (i.e., 0.1, 0.2, and 0.3 mol CO2/mol MEA) were found to decrease HSS extractability of OH Aliquat. To extract HSS effectively, as lean MEA as possible must be used with the proposed extraction technique. 10. The optimized regeneration conditions for used Aliquat were determined at 4 kmol/m3 NaOH, 10 min mixing time, 323 K temperature, and 1:1 phase ratio of OH Aliquat and NaOH. 11. OH Aliquat was also capable of removing 96, 98, and 99% of imidazole, 2-oxazolidone, and N-(2-hydroxyethyl)succinimide, respectively. N-(2-hydroxyethyl)acetamide and 1(2-hydroxyethyl)-2-imidazolidinone were also extracted with the respective efficiencies of 43 and 49%.

zolidinone were also extracted by Aliquat with the respective efficiencies of 43 and 49%. The high extraction efficiency of imidazole could possibly be explained by its resonance structure. The lone pair electron delocalization of acidic nitrogen atom into the ring induces various negatively charged sites on the ring. This could be favorable to the ion-exchange process with OH− from the modified Aliquat. 2-Oxazolidone and N-(2-hydroxyethyl)succinimide could also acquire resonance structure but giving only one negative site on the carbonyl oxygen, if this was the only site for the ion-exchange process with Aliquat. There must be other factors contributing to a very high extraction efficiency similar to that of imidazole which must be investigated further. The ring structure could have also contributed because it was the feature 2-oxazolidone and N-(2-hydroxyethyl) succinimide had in common with imidazole. The factors contributing to less extraction efficiency of N-(2-hydroxyethyl) acetamide and 1-(2hydroxyethyl)-2-imidazolidinone are still unclear and will be determined in our future work. The optimization of the technique needed to improve the ability of the extractant will also be done in our future work.



ASSOCIATED CONTENT

S Supporting Information *

4. CONCLUSIONS 1. The optimum concentration of TOA and Aliquat 336 used without MEA were respectively determined to be 0.6 and 1 kmol/m3. In water, TOA gave 58% to 96% HSS extraction efficiency while 31% to 91% was obtained from Aliquat 336. 2. The optimum concentrations of both extractants used in the MEA environment remained the same as those used in water. However, the extraction efficiency dropped to 44% to 86% (average %) and 40% to 88% (average %) respectively for TOA and Aliquat 336. TOA worked more effectively than Aliquat 336 in the absence of MEA while their extractability was equivalent when MEA was present in the system. 3. MEA had a significant effect on both TOA and Aliquat 336 as it reduced their extraction performance. The presence of MEA increased the pH of the aqueous phase and generated a strong bond formation with HSS, thus reducing the effectiveness of TOA and Aliquat 336 in HSS extraction mechanisms. 4. Further chemical modification of Aliquat using OH− to replace Cl− in its structure was able to extract over 90% of most HSS from aqueous MEA. 69% OH Aliquat extracted HSS better than the original 336 by 9% to 45%. The increase of the % OH in the Aliquat structure to 79%, 87%, and 88% did not give a significant increase in HSS extraction efficiency. 5. Two step extraction using 69% OH Aliquat to initially remove HSS followed by TOA as a second extraction step and mixed 69% OH Aliquat and TOA were found to perform equally and be capable of extracting over 90% HSS from aqueous MEA. Both techniques were superior to the use of only TOA. However, the performance of the two-step extraction and mixed extractants was equivalent to 69% OH Aliquat. 6. Aliquat conversion was absolutely necessary to reduce Cl− contamination. A higher % conversion introduced less Cl− into MEA solution. 7. In the system with Cl− alone, the two-step extraction using 87% OH Aliquat and TOA removed 73% Cl− as compared to 40% removal rate obtained by a mixture of the same extractants. This shows that Cl− contamination from modified Aliquat was manageable using these techniques.

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.iecr.6b00035. Figures S1−S4 show concentration effect of TOA and Aliquat 336 on extraction of HSSs in water and MEA solution. Figures S5−S6 compare TOA and Aliquat 336 at their optimum concentration on HSS extraction in water and MEA, respectively. Figures S7 and S8 show the effect of 69% and 88% OH Aliquat concentration on extraction of HSS in 5 kmol/m3 MEA. Figures S9 and S10 show the effect of % OH in Aliquat and comparison of different extraction steps. Figure S11 shows Arrhenius plot of HSS for Aliquat regeneration. Figure S12 compares regeneration efficiency of different phase ratios in 88% and 69% Aliquat systems. Table S1, S2, S3, and S4 summarize extraction conditions, Aliquat 336 chemical modification conditions, Langmuir correlation data of TOA and Aliquat extraction systems, and concentration of HSS in 1 kmol/m3 OH Aliquat before regeneration, respectively (PDF)



AUTHOR INFORMATION

Corresponding Author

*(R.I.) Phone: +1-(306) 585-4470. Fax: +1-(306) 585-4855. Email: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors would like to acknowledge Natural Sciences Engineering Research Council (NSERC) Canada and Clean Energy Technologies Research Institute (CETRI) for financial support.



REFERENCES

(1) Kertes, A. S.; King, C. J. Extraction Chemistry of Fermentation Product Carboxylic Acids. Biotechnol. Bioeng. 1986, 28, 269. (2) Wardell, J. M.; King, C. J. Solvent Equilibria for Extraction of Carboxylic Acids from Water. J. Chem. Eng. Data 1978, 23 (2), 144.

5017

DOI: 10.1021/acs.iecr.6b00035 Ind. Eng. Chem. Res. 2016, 55, 5006−5018

Article

Industrial & Engineering Chemistry Research (3) Tamada, J. A.; King, C. J. Extraction of Carboxylic Acids with Amine Extractants. 2. Chemical Interactions and Interpretation of Data. Ind. Eng. Chem. Res. 1990, 29, 1327. (4) Hong, Y. K.; Hong, W. H. Equilibrium Studies on the Reactive Extraction of Succinic Acid from Aqueous Solutions with Tertiary Amines. Bioprocess Biosyst. Eng. 2000, 22, 477. (5) Yang, S. T.; White, S. A.; Hsu, S. T. Extraction of Carboxylic Acids with Tertiary and Quaternary Amines: Effect of pH. Ind. Eng. Chem. Res. 1991, 30, 1335. (6) Reisinger, H.; King, C. J. Extraction and Sorption of Acetic Acid at pH above pKa To Form Calcium Magnesium Acetate. Ind. Eng. Chem. Res. 1995, 34, 845. (7) Bora, M. M.; Dutta, N. N.; Bhattacharya, K. G. Extraction of 7Aminocephalosporanic Acid with Secondary, Tertiary, and Quaternary Amines. J. Chem. Eng. Data 1998, 43, 318. (8) Kyuchoukov, G.; Marinova, M.; Molinier, J.; Albet, J.; Malmary, G. Extraction of Lactic Acid by Means of a Mixed Extractant. Ind. Eng. Chem. Res. 2001, 40 (23), 5635. (9) Li, Z.; Qin, W.; Dai, Y. Extraction Behavior of Amino Sulfonic Acid by Tertiary and Quaternary Amines. Ind. Eng. Chem. Res. 2002, 41 (23), 5812. (10) Yunhai, S.; Houyong, S.; Deming, L.; Qinghua, L.; Dexing, C.; Yongchuan, Z. Separation of Glycolic Acid from Glycolonitrile Hydrolysate by Reactive Extraction with Tri-n-octylamine. Sep. Purif. Technol. 2006, 49, 20. (11) Rasrendra, C. B.; Girisuta, B.; van de Bovenkamp, H. H.; Winkelman, J. G. M.; Leijenhorst, E. J.; Venderbosch, R. H.; Windt, M.; Meier, D.; Heeres, H. J. Recovery of Acetic Acid from an Aqueous Pyrolysis Oil Phase by Reactive Extraction using Tri-n-Octylamine. Chem. Eng. J. 2011, 176−177, 244. (12) Vitasari, C. R.; Meindersma, G. W.; de Haan, A. B. Glycolaldehyde Coextraction During the Reactive Extraction of Acetic Acid with Tri-n-octylamine/ 2-ethyl-1-hexanol from a Wood Based Pyrolysis Oil-Derived Aqueous Phase. Sep. Purif. Technol. 2012, 95, 39. (13) Supap, T.; Idem, R.; Tontiwachwuthikul, P.; Saiwan, C. Analysis of Monoethanolamine and Its Oxidative Degradation Products During CO2 Absorption from Flue Gases: a Comparative Study of GC-MS, HPLC-RID, and CE-DAD Analytical Techniques and Possible Optimum Combinations. Ind. Eng. Chem. Res. 2006, 45 (8), 2437. (14) Bord, N.; Crétier, G.; Rocca, J.-L.; Bailly, C.; Souchez, J.-P. Simultaneous Determination of Inorganic Anions and Organic Acids in Amine Solutions for Sour Gas Treatment by Capillary Electrophoresis with Indirect UV Detection. J. Chromatogr. A 2005, 1100, 223. (15) Hong, Y. K.; Hong, W. H. Removal of Acetic Acid from Aqueous Solutions Containing Succinic Acid and Acetic Acid by Tri-noctylamine. Sep. Purif. Technol. 2005, 42, 151. (16) Marcus, Y.; Kertes, A. S. Ion Exchange and Solvent Extraction of Metal Complexes; Wiley-Interscience: New York, 1969. (17) Juang, R.-S; Huang, R.-H. Comparison of Extraction Equilibria of Succinic and Tartaric Acids from Aqueous Solutions with Tri-noctylamine. Ind. Eng. Chem. Res. 1996, 35, 1944.

5018

DOI: 10.1021/acs.iecr.6b00035 Ind. Eng. Chem. Res. 2016, 55, 5006−5018