PERFORMANCE OF EXTRACTION IN MINI-FLUIDIC CHANNELS

Compositional analysis of the raw 18/12 EE feedstock (Table 6.1), typical EPA/DHA contents in the final hexene fractions (Table 6.2) and estimates of wt.% recoveries (Table 6.3) are provided for both the mini-fluidic reactor and batch-wise experiments carried out in a vessel containing 50 ml AgN0₃ solution and 14.7 ml 18/12EE FOs (comparable sample volumes to what was obtained for each length from the mini-fluidic sample). The composition of 18/12EE FOs used as a raw material in the extraction studies.

TABLE 6.1 Composition Analysis of the 18/12EE Fish Oil Feedstock

TABLE 6.2 Weight Percent EPA DHA'co-3 in Final Hexene Fractions Collected at Different Contact Times from the Mini-Fluidic (Left) and CSTR (Right) Experiments

TABLE 6.3 Approximate wt.% Yields of EPA DHA co-3 Based on the Actual Weight of Each Collection in the Hexene Fraction Relative to the Expected Quantity Fed

The analytical results provided from DSM for the various samples tested are provided in the appendices for future reference, as well as the raw mass balance data used in determining the wt.% yields given. The mini-fluidic flow reactor took minimum contact time of ~36 seconds, extreme in excess of 15 seconds necessary to meet the equilibrium condition in previously reported on an amino-fluidic system of Kamio et al. (2010, 2011). This was, in part, due to lower-than-anticipated pumping rates used in the actual experiment due to the below-specification performance of the syringe pumps employed. As a result, almost all of the compositions and yields are comparable to the contact times tested. The performance of the mini-fluidic system and CSTR are also comparable in terms of equilibrium conditions, with an approximate increase of 260% co-3 content and a conservative recovery estimate approaching 75 to 80%.

The recovery estimates are to be viewed as conservative estimates, as they represent only the actual mass of co-3’s collected after sample losses compared to a theoretical amount added. For the CSTR, the amount added was well defined due to the batch-wise operation, limiting potential errors in yield calculations to losses during handling. For the mini-fluidic system, an unanticipated flow pattern within the channels could have potentially affected the relative flow rates of the two species in question (and the contact times), creating additional uncertainty in the yield. In the interest of defining an upper bound for recovery calculations, the co-3 content in the oil residual and hexane fractions are also presented in Tables 6.4 and 6.5 for the mini-fluidic and CSTR experiments. It is worthwhile noting that the typical sample masses obtained for the oil fractions were 5.5 to 7.5 grams, while the hexane fractions after solvent evaporation were 0.5 to 1.5 grams. This is compared to a mass of 2 to 4 grams of concentrated oil obtained following decomplexation and solvent removal. Based on these values, the maximum recovery of EPAand DHA based on mass retained in the residual oil would approach just over 90% (based on 8 grams of residual oil, 1.2 wt% EPA content, vs. 13.2 g of fed oil at 15 wt.% EPA).

TABLE 6.4 Weight Percent EPA/DHA'co-3 in the Hexane Fraction Obtained During De-Emulsification

Note: These values are determined from peak areas and interpolated calibrations from the raw oil and hexene fraction calibrations, introducing some error.

TABLE 6.5 Weight Percent EPA'DHA/co-3 in the Residual Oil Fraction Initially Separated from the Sample