1.6. STATE OF THE ART OF BIODIESEL PROCESSING
1.6.2 COLLOID CHEMICAL ASPECTS
Colloid chemistry constitutes the basis of the improved biodiesel production technology presented here. Dorado [103] did touch the subject by reporting that feasibility of biodiesel production was “notoriously” dependent on FFA content of the feedstock. The notorious influence on feasibility term relates to loss in biodiesel product yield and increase in specific consumption, directly accredited to FFA contents.
Losses associated to soap formation in which the catalyst is consumed are generally accounted. Accordingly, on this basis vendors generally ask for an FFA content of not more than 5%. Additional costs associated to the need for more difficult and complex refining of the G-‐phase for cleaner grade glycerol production have not been accounted.
The necessary amount of catalyst is calculated in practice by summing the amounts needed for soap formation and catalysis. Any feedstock component that is not
converted into biodiesel is contributing to loss of profit in biodiesel production, as the most important cost category is the price paid for the most valuable asset, the feedstock [104]. To acknowledge the importance of feedstock the ultimate and accurate cost was if the purchase price is corrected to the rate of final yield to feedstock. Such account attracts attention to importance of feedstock and to good manufacturing practice alike.
Kwicien [105] concluded, without referring to colloid chemical phenomena, that the presence of salt contributes to loss of product into the glycerol phase, as concluded on the ground of experimental yield of G-‐phase. It is to report that in many samples of G-‐phase taken from conventional technology units the biodiesel content in the G-‐phase has been in the range of 10-‐20%. This corresponds to a total loss of 1-‐2% of biodiesel on feedstock basis. This loss is due to phase characteristics and by such it cannot be avoided in conventional technologies.
The view of Noureddini and Zhu [106] on trans-‐esterification, as being a system of reversible reactions, is that there is a time span needed to get to equilibrium conditions and this process is controlled by mass transfer characteristics. Unfortunately the discussion does not touch the matter that this mass transfer takes place between phases of a disperse system. Mittelbach [107] made a step toward colloid chemical description of the system, by reporting that soap formation is a necessary process initiator in alkaline catalyzed trans-‐esterification. It is needed to kick start the reaction in the disperse oil-‐methanol-‐catalyst system, to draw the methanol and catalyst into contact with the substrate oil. This is a clear description of an interfacial reaction system, both the forward and the backward reactions take place between the continuous substrate and the disperse reagent-‐reactant globules phases. May be because of the lack of translating this colloid chemical observations into the language of trans-‐esterification chemistry debates are still open about the mechanism and kinetics matters. Negy [108] stated that the knowledge of mass transfer and reaction mechanism is still far from complete. Not even is known where exactly the reaction takes place, at the interface or in one or both the phases, in bulk or in film or both.
When revisiting and defining trans-‐esterification from engineering and colloid chemical viewpoints the objectives are to shift the reversible reaction toward methanolysis and to remove mass transfer limitations. Barrier to phase transfer at an interface can be overridden by freeing the interface by the use of a co-‐solvent that brings reaction partners into a single phase [109]. Boobcock employed and recommended the use of tetra-‐hydro-‐furan, a characteristic polar solvent on the ground of ease of recycle by distillation (boiling point: 66 °C). Guan [110] proposed the use of dimethyl ether that has even lower boiling point: -‐24 °C. This is making me to question the credibility of the results published4. By the use of these solvents the apparent reaction rates become much faster and time to equilibrium much shorter. Accordingly a suitable polar solvent keeps the entire system in solution, including the starting stocks and all the reaction partners and products. The entire system stays in a single homogenous phase. Such polar solvents are known in petroleum refining practice as selective refining solvents for aromatics and lubricating oil fractions, because of
4 My ethical questions beside of forgetting that the polar solvent used was recommended a decade earlier by Boccock are the following: it is stated that the reaction of trans-‐esterification occurs in the methanol phase without knowing which phase is the disperse and which is the continuous one, if the reaction mixture remained homogenous how could have been the limit of reversible reaction lifted?)
preferentially dissolving polar constituents into the extract phase and rejecting the apolars into the raffinate phase. Under those specified phase conditions published breaking into two phases is avoided and operational conditions are selected to maintain a single phase along the reactions in procedures that employ polar co-‐solvent technique. Accordingly, by maintaining a monophase system, selectivity of the solvent has been ignored and not profited. Reactions in both directions are equally possible and promoted by kinetic means. The net gain in employing this non-‐selective polar co-‐
solvent technique results in reaching close to equilibrium conditions in much shorter time. Another important handicap of using polar solvent to promote trans-‐esterification is the need for intermediary product and solvent recycling in a system that employs close to 6 times the stochiometric amount of methanol. The equilibrium limitation applies and the trans-‐esterification must be done in two successive steps with intermediary solvent recycling and separation of glycerol byproduct.
Shifting the equilibrium toward completion can be achieved by applying very high excess of methanol reagent. This approach can ruin feasibility of the process. Under normal industrial conditions methanol:oil ration is usually 6:1 (double to stochiometry) and the equilibrium condition is close to 80%. The increase of excess rate to quadruple shifts this conversion figure to close to 90%, see figure 1.8 [111]. Even under such conditions reaching the equilibrium conversion needed about 1 hour of reaction.
Relatively mild conditions in trans-‐esterification favour methanolysis, for hydro-lysis higher temperature (180-‐260 °C) and pressure (3-‐7 MPa) and presence of water are needed [112]. The cited work has a merit of attracting attention to work in counter current mode of operation to attain the highest possible conversion and productivity in hydrolysis. This conclusion was mainly based on colloid chemistry aspects by expressing that the chemical reactions involved are very slow. This view was in fact a promoting reason to work on bringing the batch biodiesel process to counter current mode.
FIGURE 1.8. KINETICS OF SUNFLOWER TRANSESTERIFICATION AT MEOH:OIL=12:1 WITH DIFFERENT CATALYSTS
(NaOH catalyst: lowest, KOH: high, CsOH: highest)
Removal of either reaction products can be an efficient tool in shifting equilibrium toward completion and chances are given for making it in a single reaction contact step.
For such scope selectivity of the solvent must be reinstalled. Barring the reverse reaction by removing one of the products from reaction media could only be done within the frame of economic operations by the proposed use of apolar solvent. The present work addresses colloid chemical basis on this basis. Phase transfer, glycerol rejection and trans-‐esterification kinetics experiments were carried out to de-scribe the phenomena of shifting trans-‐esterification toward completion in a single phase.
When revisiting colloid chemical aspects micro-‐emulsions must also be visited and considered. Micro-‐emulsions with short chain alcohols can yield clear, thermodynamically stable liquid fuel with viscosities in the range of diesel fuel specifications [113]. The special benefit of water and oxygenates containing diesel fuel consists in freezing NOx chemistry by lowering the flame temperature and by such reducing the emissions of this regulated compound [114]. Similar emissions reduction can be achieved with blending water in fossil diesel fuel too [115]. Revisiting this aspect can be reasoned on the ground of ease in atomization of the (micro)-‐emulsified fuels similarly to biodiesel. This effect compensates for the slightly lower heating value in operational diesel engines.
It is to report that the patent related to the present thesis was submitted in 2001 and full grants have been obtained since. This is worth for another ethical comment, Shi and Bao used my technique (and my motives) of hexane dilution without referring to my works. The authors have not even observed a series of advantages of using the hexane solvent and passed next to finding facts, such as: a) necessary reaction time is less than 10 min Shi and Bao used 120 min., b) settling occurs instantaneously and there is no need for settling over 2 h of time, c) ignored the fact that if the crude biodiesel is simply washed with distilled water emulsion formation must be present, d) there is no need for employing 9 times stochiometry, even 2 times stochiometry is excessive under such conditions [116].