Volumetric Method for Free and Total Glycerin Determination in Biodie

Scientific Journal
of 7
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  Volumetric Method for Free and Total Glycerin Determination in Biodiesel M. L. Pisarello, B. O. Dalla Costa, N. S. Veizaga, and C. A. Querini*  Instituto de In V estigaciones en Cata´lisis y Petroquı´mica, INCAPE (FIQ, UNL-CONICET),Santiago del Estero 2654, 3000 Santa Fe, Argentina In the past few years, biodiesel has gained considerable attention due to the increasing concern for theenvironmental problems associated with the use of fossil fuels. The quality control of this renewable fuelinvolves numerous analyses, several of them requiring the use of analytical instruments. According to ASTMand EN standards, the analyses of free and total glycerin which are of particular relevance to biodiesel qualitymust be carried out by GC analysis. This work presents an alternative volumetric method, which does notneed expensive equipment, and is therefore particularly useful for quality control in small facilities. Anotheradvantage of the proposed method is that it also overcomes the shortcomings of the GC procedure, such asdetection limits and type of raw material used to produce biodiesel. While the GC analysis is restricted tobiodiesel obtained from soybean, rapeseed or sunflower oil, the volumetric method has no limitations in thissense. 1. Introduction Biodiesel production has significantly increased in the pastfew years. Several plants with large production capacity havebeen built and several others are under construction in manycountries. These plants mainly use rapeseed oil (Europe),soybean oil (USA, Brazil, Argentina), or palm oil (Asia) as rawmaterials. On the other hand, the use of alternative raw materialshas also received considerable attention due to the increasingconcern about the use of edible oil for fuel production. Amongthe oils used for biodiesel production other than soybean,rapeseed, or sunflower oil, we can mention oils of jatropha,algae, castor, linseed, or safflower oil. These raw materialsrepresent an option for small farmers and communities locatedfar from urban centers to produce the fuel they need for farming,heating, and transportation. The use of very small productionfacilities to produce biodiesel for self-consumption has alsospread throughout many countries. For these latter cases, thequality control of biodiesel is a major concern.The quality of biodiesel is specified in standards such asASTM D-6751, 1 or EN 14214. 2 Among the numerous propertiesthat must be controlled in order to meet these specifications,total and free glycerin are two of the most important ones, sincethey are related to the extent of unconverted triglycerides andthe effectiveness of the purification procedure, respectively.According to ASTM or EN standards, these determinations mustbe carried out by chromatographic analysis. Samples have tobe silylated, and two different internal standards must be usedin order to quantify free glycerin and mono-, di-, and triglyc-erides. The procedure that must be followed for these analysesis described in ASTM D-6584 3 or in the EN 14105 4 and EN14106. 5 There are other methods to determine the mono-, di-, andtriglycerides content in biodiesel, such as size exclusionchromatography, 6 viscosity, 7 TGA, 8 reversed-phase high-performance liquid chromatography (RP-HPLC), 9 free glycerolin biodiesel by capillary electrophoresis, 10 nonaqueous reverse-phase HPLC with a UV detector, 11 infrared spectroscopy, 12 - 14 Raman spectroscopy, 15 or NMR to identify intermediate andfinal products during the transesterification. 16 Different GCmethods have been analyzed by Mittelbach et al. 17 Each technique has advantages and drawbacks, and theselection of the analytical procedure will be finally madedepending on whether it is for quality certification, pass or failedcontrol, or process monitoring. In addition, quality, simplicity,cost, and duration of the analysis including possible samplepretreatments are very important aspects.The gas chromatographic analysis as described in each of the above-mentioned standards (ASTM D6584 3 and EN 14105 4 )has limitations since it has to be used only for biodiesel obtainedfrom soybean, rapeseed, or sunflower oils; besides, it cannotbe used with other oils, such as coconut, or animal fats, sincepeak overlapping might occur. This is a severe limitation of the method since one of the reasons why biodiesel is gainingacceptance is because it allows the use of alternative rawmaterials such as chicken, pork, or cow fat, in addition to manyother vegetable oils such as coconut, cotton, algae, or palm oil.A volumetric method, described in the AOCS Ca 14-56, 18 involves saponification with an alcoholic solution of KOH,followed by the addition of chloroform and acetic acid,separation, washing, addition of periodic acid first and then KI,and finally iodine titration with sodium thiosulfate. However,even though this method does not require any complicatedequipment, it implies the handling of several harmful substancesand is not quite suitable for the determination of total and freeglycerin at low levels with high precision. Besides, it includesmany chemical steps.In this work, we present a different volumetric procedure,which has been tested for several years already. This methodhas no limitations regarding the raw material used to producebiodiesel or the level of combined or free glycerin; it has a veryhigh precision and repeatability, and does not require the useof expensive and/or sophisticated equipment. In fact, severalbiodiesel industries in Argentina, Paraguay, and Chile have beenusing this procedure for some time, as well as small farmers.This method does not allow the individual determination of mono-, di-, and triglycerides. 2. Fundamentals Total glycerin content is obtained after all the glycerides, thatis, mono-, di-, and triglycerides are quantitatively transformed * To whom correspondence should be addressed. Tel.:  + 54-342-4533858. Fax: + 54-342-4531068. E-mail: Ad-dress: Santiago del Estero 2654, 300) Santa Fe, Argentina.  Ind. Eng. Chem. Res.  2010,  49,  8935–8941  8935 10.1021/ie100725f   ©  2010 American Chemical SocietyPublished on Web 08/30/2010  into methyl esters and glycerin, by transesterification. Afterward,the glycerin is extracted first with acidified water, and then withwater. Finally, the glycerin is titrated according to standardprocedures.This method is very sensitive, and it is possible to performthe total glycerin determination with 50 g of sample or less.Free glycerin is determined just by extraction with water,followed by titration, as indicated in the ASTM D-1615, 19 IRAM 41089, 20 BS-5711, 21 or AOCS Ea 6-94. 22 ASTMD-1615 19 and IRAM 5571 23 refers to the determination of glycerin in alkyd resins and was discontinued in 2004. BS-5711is specific for glycerin analysis and has also been discontinued.Nevertheless, since it has not been replaced by any otherstandard, international commerce is based on BS-5711.In all these standards glycerin titration is based on itsoxidation by sodium periodate. This reaction leads to theformation of formic acid when more than 2 hydroxyl groupsare present in the molecule. Primary hydroxyl groups reactforming formaldehyde upon oxidation and secondary hydroxylgroups lead to formic acid production, according to the followingreaction:Formic acid is then titrated with sodium hydroxide. This methodcan be applied to samples free from sugars or other organiccompounds with more than two adjacent hydroxyl groups.To consume the excess of sodium periodate, ethylene glycolis added when reaction I is completed. The reaction that takesplace is 3. Results and Discussion3.1. Total Glycerin Determination.  To obtain reliablequantitative results, the total conversion of the glycerides presentin the sample must be assured. This is achieved using a largeexcess of methanol and catalyst. This procedure cannot be carryout during normal biodiesel production, because of the severepurification problems that arise when large amounts of catalystare used. Besides, the energy requirement involved in methanolrecovery would also be a major problem. In our case, since theobjective is not the production of biodiesel according tostandards, using a large excess of catalyst is not a problem.After the reaction, the system is neutralized with an aqueoussolution of HCl 5 wt %, followed by two additional washingsteps, in order to recover all the glycerin present in the biodieselphase. In this work, we have checked that under these condi-tions, no mono-, di-, or triglycerides are found on the finalproduct, and therefore the glycerin formed under these condi-tions exactly reflects the amount of unreacted glyceridessrcinally present in the sample.Figure 1a shows an example of a biodiesel that contains 0.24wt % of total glycerin, as determined by EN14105, 4 before andafter the first step of this procedure, that is, the reaction withmethanol and NaOH as catalyst. It can be clearly seen that nodi- and triglycerides are left after the reaction. However a verysmall peak that elutes at the same time as the monoolein is stillpresent after this reaction step. According to our results, it isvery likely that this peak is not a monoolein, as discussed below.Therefore, as above-mentioned, this means that the glycerinformed during this step of the procedure is an exact representa-tion of the amount of unconverted glycerides. Figure 1b showsanother example of a GC analysis of a biodiesel sample, andthe same sample after the first step of this procedure. In thiscase, the biodiesel contains 1.08 wt % of total glycerin(EN14105 4 ). Again, all the di- and triglycerides are converted,and a very small peak comes out at the same retention time asthe monoolein. In both cases (Figure 1 panels a and b), thispeak represents an equivalent of total glycerin content lowerthan 0.01 wt %, if assumed to be monoolein. In any case, thisis a very small value, and as discussed below, it is well belowthe repeatability and reproducibility of the total glycerin contentdetermination by the GC-technique.Even though this procedure does not allow the determinationof the individual amounts of mono-, di-, or triglycerides, itprovides a reliable and relatively fast determination of the totalglycerin content. It should be remarked that the ASTM standarddoes not specify the individual content of each type of glycerides, while the EN standard does.The detailed procedure that must be followed in order toobtain good results is presented in Appendix A. 3.1.1. Glycerin Titration.  In many standards related toglycerin analysis (ASTM D-1615, IRAM 5571, IRAM 41089,AOCS Ea 6-94), a blank must be performed. However, we havechecked that when following the procedure described in IRAM41089 or IRAM 5571, in all cases only one drop of 0.1 N NaOHsolution was used. Nevertheless, each lab must check whetheror not with the procedure being followed, the blank has aninfluence on the final calculation. Other standards, such as BS5711 carry out the final titration at an acidic pH and, therefore,in these cases it is necessary to perform the blank determination.In our procedure we used the glycerin analysis as described inIRAM 5571, in which the blank experiment was alwaysnegligible. This procedure is similar to that described in ASTMD-1615. However, the latter is designed to determine glycerin,ethylene glycol, and pentaerythritol in alkyd resins. In this latterstandard, the solution is not boiled prior to the titration.Therefore the blank could be more relevant since the CO 2 absorbed from the air during the sample handling is not strippedby boiling the solution, and consequently certain amount of thetitrating reactant (NaOH) will be used to neutralize the carbonicacid. 3.1.2. Analytical Results.  A set of experiments was carriedout, by adding a known amount of triglycerides (refined soybeanoil) to a biodiesel which was previously analyzed by GC inorder to determine the total glycerin content.Table 1 shows the results. In this table, examples of thedeterminations that were carried out are presented. A plot of the theoretical total glycerin amount as a function of themeasured amount (not shown) displays a general trend exactlyon the line at 45 ° , which indicates that the procedure has nobias either at low or high total glycerin content. The absolutedifference between the theoretical value and the measure valueof approximately 0.1 wt % is the larger deviation found at hightotal glycerin content, and a difference of approximately 0.05wt % is the larger error at total glycerin contents below 0.25wt %. It has to be emphasized that this set of results correspondsto experiments in which additions of given quantities of vegetable oils were carried out, and the exact amount of totalglycerin that corresponds to this addition is not known due tosmall uncertainties, such as for example, the oil molecularweight.The volumetric procedure was also used with a sampleemployed in an interlaboratory test, organized by CEMITEC(Spain), in which 9 European laboratories took part. Table 2shows these results, as well as a comparison of several othersamples analyzed by the method presented in this work, andC 3 H 8 O 3  +  2NaIO 4 f 2HCHO  +  HCOOH  +  2NaIO 3  +  H 2 O(I)C 2 H 4 (OH) 2  +  NaIO 4 f 2HCHO  +  NaIO 3  +  H 2 O (II) 8936  Ind. Eng. Chem. Res., Vol. 49, No. 19, 2010  by the GC methology. It can be seen that our results are ingood agreement with the average of other laboratories that usedthe EN 14105 procedure.Regarding the repeatability, ASTM D 6584 establishes thatfor total glycerin determination, the repeatability is  r  ) 0.009.The EN 1405 gives the following equation for repeatability:  r  )  0.0687  x   +  0.004 (where  x   is the mean value), and forreproducibility it provides the following information:  R  ) 0.4472  x  - 0.01. Therefore, in the latter case, it can be expectedthat for a total glycerin content of 0.15 the reproducibility is0.057 wt %, which represents a difference of 38% between twomeasurements carried out in different laboratories. Our deter-mination in this case differs from the mean value in 0.03 wt %,which represents a deviation of 20%. Therefore, the volumetricprocedure results to be well within the quality standardsregarding the statistics criteria. The  z -criterion establishes that  z  should be less than 2, to have a reliable result,  z  being definedasThe volumetric measurement is well within the average of theother nine laboratories, with  z ) 0.55, being  σ  ) 0.055. A valueof   z  less than 2 is acceptable and means that the measured valuediffers from the real value less than two times the standarddeviation, and this would occur in 95% of the cases for a normaldistribution. 3.1.3. GC Analysis: Influence of Free Fatty Acids.  Asabove-discussed, the GC analysis of a biodiesel sample reactedwith excess of methanol and catalyst, shows a small peak thatcomes out at the same retention time as the monoolein (seeFigure 1a,b).We explored the possibility that it could correspond to a fattyacid. The fatty acid elutes at longer retention times than thecorresponding methylester. As indicated in the example givenin the ASTM D-6584 standard, the C24 methylester comes outin the chromatogram right before the main peak that correspondsto the monoglycerides. Therefore, the small peak shown inFigure 1a,b could be related to the C24 fatty acid. Figure 2shows GC analyses of a biodiesel sample before and afteraddition of a mixture of free fatty acids. It can be clearly seen Figure 1.  GC analyses of biodiesel samples, before (A) and after (B) the first step of the volumetric procedure: (a) sample with 0.24 wt % total glycerin,(b) sample with 1.08 wt % total glycerin.  z  )  x   -  x  j  /  σ  Ind. Eng. Chem. Res., Vol. 49, No. 19, 2010  8937  that the peak assigned to the monoolein increases after theaddition of free fatty acids, which indicates that in the GCanalysis there might be an interference of the free fatty acid inthe total monoglyceride analysis, most probably due to the C24fatty acid. We carried out many reaction tests, using differentmethanol to oil ratios, catalyst amount, reaction times, andperforming also up to three reaction cycles without being ableto eliminate this peak, what strongly suggests that it is not anonconverted monoolein. Either a free fatty acid or another typeof compound could be responsible for this small peak. Never-theless, we did not try to further identify it, since whatever itis, it is included in the analysis of monoglycerides as indicatedby the EN14105 or ASTM D-6584. On the other hand, the peak represents a total glycerin content of 0.01 wt % or less. This isthe reason why even after two or three reactions carried outwith a large excess of methanol, starting with a raw materialwith low total glycerin content, a small peak can still beobserved at approximately the same retention time as themonoolein peak. Therefore, either the acidity of the sample orother nonidentified oil component could lead to an overestima-tion of the total glycerin content when using the GC method.This problem is not found when using the volumetric procedure. 3.1.4. Repeatability.  Figure 3 shows results obtained whenreplicated determinations were carried out in the same sample. Itcan be observed that the difference between two determinations isseldom larger than 10%. Only in one out of 14 samples was thedeviation between two measurements larger than 10%, includingexperiments carried out by different operators and experimentscarried out on different days. This is an excellent result comparedwith the expected repeatability of the instrumental proceduredescribed in EN14105 or ASTM D6584. It is important to notealso that many of the samples included in Figure 3 have totalglycerin content in the range required by the standards.The titration of the glycerin collected in water after the threewashings, is very reliable. Results for replicated titrations of this glycerin are shown in Table 4. The difference between twoanalyses is very small. 3.1.5. Influence of Each Step Length.  The chromatographicprocedure as described in the above-mentioned standardsrequires a total analysis time of approximately 1.5 h, takinginto account sample preparation with two internal standards,silanization, and GC analysis. The time involved in each stepof the volumetric procedure as indicated in the Appendixes, arevery conservative, and if needed, they can be substantiallyreduced, as follows. Glycerin Titration Times.  The time required to oxidizeglycerin with NaIO 4  according to the ASTM standard is 30 min,followed by 20 min in darkness after ethylenglycol (EG) is added.We compared results obtained using these times (as prescribed inthe standards) with results obtained using 5 min in each interval.It was found that the differences between these results are withinthe experimental error of this procedure (not shown). Reaction.  Experiments with times shorter than 2.5 h werecarried out, varying also the methanol/sodium methoxideproportion loaded to the reactor. Table 5 shows the results. Fromthese data, it can be concluded that using 40% methanol referredto the volume of oil loaded to the reactor, with 35 g NaOH/Lof methanol and a reaction time of 30 min, it is enough to ensuretotal conversion of mono-, di-, and triglycerides. Extraction Steps.  To determine whether three washing stepsare necessary to fully extract the glycerin from the reactionmedia, the aqueous phase coming from each of these washingsteps was analyzed to determine the glycerin content. It wasfound that the first extraction recovered around 97% of the totalglycerin formed during the reaction, the second extraction 2%,and the third 1% or less. These results indicate that in order tomake sure that all the glycerin is extracted, three washingsshould be carried out. On the other hand, it can also beconcluded that a very small error could be introduced if thethird washing is not carried out. 3.2. Results for Free Glycerin Determination.  The freeglycerin determination is carried out by extracting it from thebiodiesel sample with three consecutive washings. The first of them is carried out with an aqueous solution of HCl to avoidthe formation of a stable emulsion in those cases in which thesample is basic, that is, when the biodiesel sample is not properlypurified during the production. The detailed procedure isdescribed in Appendix B.Several experiments were carried out by adding glycerin tosamples of biodiesel that were previously washed several times,in order to make sure that no free glycerin was left on the sample.Table 3 shows the results. It can be seen that there is a verygood agreement between the amount of glycerin added to thesample and the experimental determination. Table 2 shows thevalues that correspond to the interlaboratory test, in whichthe free glycerin determination by the method described in thiswork corresponds exactly with the mean value of the other ninelaboratories, as determined by the EN 14105.The ASTM D-6584 renders the repeatability of this deter-mination to be  r  ) 0.001 wt %, while the EN 14105 establishes: Table 1. Total Glycerin Analyses, After Addition of Vegetable Oil toa Biodiesel Sample glycerin, wt % triglycerideadded wt %calculated  x  c measured  x  m Absolutedifference wt %error %((  x  c  -  x  m )/   x  c )  ×  100 1.028 0.112 0.154 0.042 37.51.713 0.184 0.203 0.01 5.44.310 0.462 0.462 0.000 0.00.430 0.049 0.060 0.011 22.40.965 0.105 0.113 0.007 6.71.463 0.162 0.148 0.015 9.50.970 0.732 0.750 0.017 2.30.730 0.002 0.32.010 0.842 0.900 0.058 6.90.840 0.001 0.13.000 0.946 0.870 0.076 8.00.830 0.116 12.20.940 0.949 0.880 0.069 7.31.930 1.053 1.000 0.053 0.050.450 0.897 0.950 0.052 5.80.630 0.916 0.860 0.056 6.11.530 1.011 0.940 0.071 7.02.800 1.145 1.040 0.105 9.20.600 0.863 0.860 0.003 0.31.100 0.916 0.930 0.014 1.51.550 0.963 1.000 0.036 3.7 Table 2. Comparison of Volumetric Method with the EN14105 analysis volumetric EN14105 commentstotal glycerin 0.18 0.15 Interlab control, nine Europeanlaboratories0.88 0.82 different European controllaboratories0.16 0.150.33 0.310.26 0.24 biodiesel production plantlaboratory0.29 0.26free glycerin 0.01 0.01 Interlab control, nine Europeanlaboratories0.0067 0.01 different European controllaboratories0.019 0.02 8938  Ind. Eng. Chem. Res., Vol. 49, No. 19, 2010
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!