https://doi.org/10.1016/j.petlm.2016.11.002
”
Bi–solvent blend first proposed by Chakravarty et al. is aimed at maximizing the potentials of the individual amine solvents and also limiting their individual problems [98]. Generally, bi–solvent blends involve mixing of a high CO2 absorption capacity solvent and a highly reactive amine solvent. In most cases, the high CO2 absorption capacity solvent (having an equilibrium CO2 loading of at least 1 mol CO2/mol amine) is either a tertiary or sterically hindered amines because of their ability to form bicarbonates (HCO3−) during amine–CO2 reactions. The reactive amine solvent can either be reactive monoamines (primary and/or secondary amine) or polyamines that contain primary and/or secondary amino groups (Fig. 4) which have the tendency of forming stable carbamates while tertiary amines only involve the direct formation of bicarbonates. Sterically hindered amines also involve an initial formation of an unstable carbamate which quickly hydrolyses to bicarbonate. Reactions for primary and secondary amines, tertiary amines, and sterically hindered amines are shown in Eqs. (3), (4), (5), (6)–(7), respectively [52], [123], [124], [125], [126].
2.1.1.1. Primary and secondary amines
(3)2RNH2+CO2↔RNHCOO−+RNH3+(4)2R2NH+CO2↔R2NCOO−+R2NH2+
2.1.1.2. Tertiary amines
(5)R3N+H2O+CO2→R3NH++HCO3−
2.1.1.3. Sterically hindered amines
(6)2RNH2+CO2↔RNHCOO−+RNH3+(7)RNHCOO−+H2O↔RNH2+HCO3−
In a bi–solvent blend containing different types of amines, a combination of their reactions including those shown in Eq. (3)–(7) will occur during the amine–CO2 interaction.
Idem et al. used two pilot plant (coal and natural gas fired) studies to confirmed that the bi–solvent blend of 4 kmol/m3 MEA–1 kmol/m3 MDEA could offer huge reductions in the energy of regeneration compared to single solvent 5 kmol/m3 MEA [84]. Mangalapally and Hasse studied bi–solvent blend of AMP–PZ (CESAR1) and their pilot plant results showed that the CESAR1 solvent required lower liquid flow rates (45%) and regeneration energy (20%) when compared to single solvent MEA [127]. Other researchers have studied various bi–solvent blends like MDEA–PZ and MDEA–DEA who have all reported improved CO2 capture capabilities compared to that of MEA [128], [129], [130], [131], [132], [133], [134], [135]. Bi–solvent blends containing AMP have also shown to be a potential alternative to single solvent MEA [87], [91], [133], [136], [137], [138]. Bruder et al. stated that the bi–solvent blend of 3 kmol/m3 AMP–1.5 kmol/m3 PZ showed higher cyclic capacity (120%) than 5 kmol/m3 MEA [88]. However, precaution must be taken with AMP–PZ blend due to precipitation problem potentially triggered by use of high amine concentration and high CO2 loading. Bruder also reported that bi–solvent blends involving high concentration of AMP–PZ could form solid precipitate and experimentally confirmed it when several concentrations of AMP–PZ formed solid precipitates with and without CO2 loading [88], [105].
It is also important to note that the concentration of blended amine solvents is more difficult to control than the single solvents. This is because, in the make–up unit as seen in Fig. 6, adding disproportionate amount of water and/or amine will affect the concentration of the individual amine solvents in the aqueous amine solution, which thereby reduces the CO2 capture efficiency. The 90% CO2 capture efficiency depends on the optimal concentration of the aqueous amine solution [17]. For instance, an in–house experimental study carried out on 3 kmol/m3 AMP–1.5 kmol/m3 PZ, 3 kmol/m3 AMP–2 kmol/m3 PZ and 3.5 kmol/m3 AMP–1.5 kmol/m3 PZ at 40 °C, atmospheric pressure and 99.99% CO2 confirmed that both aqueous amine solutions did not form any solid precipitate after reaching amine–CO2 equilibrium. However, based on visual inspection alone it is speculated that 3 kmol/m3 AMP–2 kmol/m3 PZ and 3.5 kmol/m3 AMP–1.5 kmol/m3 PZ still looked more viscous (sticky) which implied that precipitation could potentially take place, possibly after further use of the solvent. Upon cooling the three solutions at 20 °C for 200 h, as shown in Fig. 7, a solid precipitate was noticed in both 3 kmol/m3 AMP – 2 kmol/m3 PZ and 3.5 kmol/m3 AMP–1.5 kmol/m3 PZ blends. Their CO2 absorption capacity also followed this trend 3 kmol/m3 AMP–2 kmol/m3 PZ > 3.5 kmol/m3 AMP–1.5 kmol/m3 PZ > 3 kmol/m3 AMP–1.5 kmol/m3 PZ. This indicates that slight increase in the amine concentration can change the amine properties and CO2 capture efficiency of the amine solution which in particular, could lead to precipitation. Amine solution precipitation is very undesirable in a non–precipitating CO2 capture process because it can plug process lines and equipment as well as increase corrosion.
Fig. 7. Precipitation and non-precipitation of CO2 rich solution of the AMP–PZ bi–solvent blends after they were cooled at 20 °C for 200 h.
However, there are studies that have reported precipitating CO2 capture process [139], [140]. This involved the application of amino acid based absorbent which claimed to reduce the solvent regeneration energy. However, concerns’ regarding modification of the absorber, pumps, regenerator and other equipment’s to handle solid or slurry formed due to reaction of the solvent and CO2 still exist which can be a big burden to those who want to switch to the precipitating process. These can impact on the overall operating cost and complexity of the CO2 capture plant. Further studies need to investigate the viability of using precipitating amino acids for CO2 capture.
“