THERMOGRAVIMETRIC APPROACH FOR ASSESSING THE OXIDATION LEVEL OF A BIODIESEL SAMPLE

Transcription

1 Quim. Nov, Vol. 41, No. 5, , 2018 THERMOGRAVIMETRIC APPROACH FOR ASSESSING THE OXIDATION LEVEL OF A BIODIESEL SAMPLE Artigo Luis Díz-Bllote, *, Kren Gómez-Hernández, Elsy T. Veg-Lizm, Miguel A. Ruiz-Gómez, Luis Mldondo nd Emnuel Hernández b Deprtment of Applied Physics, Center for Reserch nd Advnced Studies of the Ntionl Polytechnic Institute, A. P. 73 Cordemex, Mérid, Yuctán, 97310, México b Deprtment of Se Resources, Center for Reserch nd Advnced Studies of the Ntionl Polytechnic Institute, A. P. 73 Cordemex, Mérid, Yuctán, 97310, México Recebido em 21/09/2017; ceito em 04/01/2018; publicdo n web em 08/02/2018 Biodiesel, unlike diesel, is highly susceptible to utoxidtion. This utoxidtion is mjor concern for biodiesel producer. Vrious methods hve been developed to estimte the degree of oxidtive stbility of biodiesel. These methods re useful to compre nd estimte the oxidtive stbility between different biodiesel smples. However, n bsolute vlue of the oxidtive stte will be more useful. In this study, we describe method of estimting figure tht cn be used s n indictor of the oxidtion level. This indictor ws determined from the therml decomposition curve of biodiesel, s obtined from the thermogrvimetric test. Biodiesel from wste cooking oil ws oxidized t 80, 100 nd 120 C using different exposure res (19.6, 63.6, nd cm 2 ). The percentge of oxidtion, mesured from low (80 C, 19.6 cm 2 ) to highly ggressive conditions (120 C, cm 2 ), ws in the rnge of 0.4 ± 0.1 nd 23.5 ± 3.3. Kinemtic viscosity ws lso mesured for the oxidized smples, nd strong correltion (R 2 =0.96) ws observed between the percentge of oxidtion products nd kinemtic viscosity. In ddition, the oxidtive effect of temperture nd re of exposure on the biodiesel smples ws confirmed by ultrviolet spectroscopy. Keywords: oxidtion of biodiesel; thermogrvimetric nlysis; wste cooking oil, oxidtive stbility, biodiesel. INTRODUCTION Biodiesel is green fuel lterntive tht is considered to be renewble, environmentlly friendly, biodegrdble nd non-toxic. 1,2 This fuel is derived from vegetble oil, or niml ft. It hs low sulfur content, low prticulte emissions nd neutrl contribution of CO 2 to the environment. 3 Therefore, this lterntive fuel is cndidte for totl or prtil substitution of petrodiesel in diesel engines. Biodiesel is commonly obtined by the trnsesterifiction rection, where vegetble oil or niml ft is mixed with short chin lcohol, like methnol or ethnol, nd ctlyst (KOH or NOH). 4 The biodiesel must be purified, nd it must comply with qulity specifictions listed in relevnt stndrds such s ASTM D or EN 14214, 6 before this green fuel is redy for use in compression engines. Oxidtion of biodiesel results in remrkble chnges in the physicl nd chemicl properties 7 of the fuel, nd in generl, the chnges hve negtive effects on the biodiesel qulity. Therefore, the degree of oxidtion of biodiesel is mjor concern for biodiesel producers. The min driving force of the oxidtion process is the presence of double bonds 8 in the chemicl structure of the mono-lkyl esters. The oxidtion process begins with the loss of hydrogen toms from the chemicl structure tht subsequently leds to formtion of lkyl-rdicls. Afterwrd, the newly formed rdicls rect with oxygen forming peroxides tht, in turn, rect gin nd give rise to the formtion of more lkyl-rdicls or hydroperoxides. 9 This cycle consumes esters nd ends with the formtion of stble compounds such s ketones, ldehydes, ethers nd lknes. The reltion between oxidtion nd qulity hs encourged development of vrious methods for evluting the oxidtive stbility of biodiesel. The most common methods in use re: the Rncimt method, the oxygen bomb test, nd the ctive oxygen method (AOM). In ddition, the use of novel methods such s ner-infrred (NIR) *e-mil: luisdiz@cinvestv.mx spectroscopy, 10 Fourier trnsform infrred (FTIR) spectroscopy, chemiluminescence, 14 UV-Vis bsorption nd thermogrvimetry hve been proposed. 15 In the Rncimt method, smll smple of oil or biodiesel (2 3 g) is plced in glss tube, nd heted up to 120 C. Next, ir is pssed through the smple nd collected in deionized wter. The oxidtion strts with the formtion of cid, voltile products tht re trpped in the wter increse its conductivity. From plot of time vs. conductivity, the induction period (IP) is determined nd used s n indictor of the oxidtive stbility. The oxygen bomb method uses 50 g of ft in metl continer which is seled nd pressurized with oxygen. Next, the bomb is heted to 99 C. In this method, decrese in pressure is used to determine the oxidtive stbility. The ctive oxygen method (AOCS method Cd 12-57) requires bubbling dry ir through 20 g of oil t constnt temperture (98 C). The peroxide vlue is monitored every 2 h. The limit of the peroxide vlue is 20 meq/kg per AOCS method Cd On the other hnd, in our previous study, 16 we demonstrted tht the residul mss exhibited by thermogrvimetric curve of the biodiesel therml decomposition cn be used s n indictor of its oxidtion level. Wheres, in this study, simple method bsed on the residul mss is used to determine quntittive figure for the oxidtion level of biodiesel smple. This method gives n estimte of the bsolute oxidtion stte, unlike other methods tht provide the reltive oxidtion of the biodiesel. With the index proposed, the biodiesel smple cn be rejected or ccepted on non-subjective bsis, depending on the oxidtion level. EXPERIMENTAL Biodiesel from wste cooking oil ws provided by biodiesel producer in México (Energí Girón. S. A.). Biodiesel ws produced following typicl lkline-ctlyzed trnsesterifiction procedure. The wste cooking oil ws used fter step of filtrtion to remove residul food. Then the rection ws crried with 6:1 molr rtio

2 Vol. 41, No. 5 Thermogrvimetric pproch for ssessing the oxidtion level of biodiesel smple 493 of methnol to wste cooking in the presence of KOH s ctlyst. Severl prmeters relevnt to the oxidtive stbility for fresh biodiesel (FB) re listed in Tble 1. The purity of the biodiesel ws determined from the weight lost due to voltiliztion of biodiesel in thermogrvimetric curves. 17 Kinemtic viscosity ws mesured t 40 C using n Ubbelohde viscometer following the procedure described in ASTM D-445. The density ws determined by mesuring the specific grvity with hygrometer (ROBSAN-1107, México) with rnge of 0.80 to 0.90 nd Polyscience 9106A11B circulting bth following the ASTM D-1298 stndrd. Tble 1. Properties of fresh biodiesel Property Vlue Purity (ester content) 97.5 ± 0.9 (mss %) Kinemtic viscosity, 40 C 4.8 ± (mm 2 s -1 ) Density, 15 C ± (g cm -3 ) Vlues re the men of triplicte mesurements ± the stndrd error. Biodiesel with different oxidtion levels were produced with modifictions to method described by Knothe. 18 Smples of biodiesel (30 ml) were poured into petri dishes with dimeters of 5.0 cm (A 1 =19.6 cm 2 ), 9.0 cm (A 2 =63.6 cm 2 ) nd 14.0 cm (A 3 = cm 2 ). A smple without oxidtion ws used s blnk nd coded s FB (fresh biodiesel). The petri dishes contining the biodiesel smples were heted in n oven (Isotemp vcuum oven, Model 281 A, Fisher Scientific, Pittsburgh, PA) to specific constnt temperture (80, 100 or 120 C). All ccelerted oxidtion tests were performed t lest in triplicte, t fixed exposure time of 24 h. Figure 1 shows scheme to illustrte the ccelerted oxidtion process which ws performed with different res nd tempertures in tmospheric ir. Thermogrvimetric nlysis ws crried out using thermoblnce (Discovery series, TA Instruments, New Cstle, DE). Mesurement conditions were: verge mss of the smples 10 ± 0.5 mg, rnge of temperture from 50 to 600 C, heting rte of 10 C/min nd nitrogen flow of 50 ml/min. Kinemtic viscosity mesurements were mde t 40 C with n Ubbelohde type viscometer, size 1 with rnge from 2 to 10 mm 2 s -1 per the ASTM D445 method. UV-visible (UV Vis) spectr of oxidized biodiesel smples were recorded from 190 to 300 nm, using model EPP2000 spectrophotometer (StellrNet, Tmp, FL). The smples were diluted (1:500 v:v) with hexne regent grde (Sigm-Aldrich). RESULTS AND DISCUSSION Thermogrvimetric nlysis Thermogrvimetric nlysis (TGA) is technique where the mss lost is recorded s function of incresing temperture in controlled environment. The technique provides dt concerning the therml decomposition of compound to its voltile form. TGA is used to study pure compound or mixture. Although biodiesel is mixture of mono-lkyl esters, most of them decompose in reltively nrrow nd defined rnge of tempertures. This chrcteristic llows differentition between biodiesel nd its contminnts. For exmple, in mono-lkyl ester nd vegetble oil mixture, the mount of ech component cn be estimted becuse both hve different decomposition temperture regions. 19,20 The bility of TGA to determine the biodiesel purity hs been compred with nucler mgnetic resonnce (NMR) spectroscopy nd the result is n cceptble discrepncy of ±1.5%. 17 A similr principle cn be used to determine the percentge of oxidtion products present in biodiesel smple, s described below. Figure 2 shows typicl therml decomposition curve of FB smple used in this study. It cn be observed tht the therml decomposition of the biodiesel smple is in the rnge of 198 to 251 C. The observed weight loss is ttributed to the ester content which is nerly 97.8% in mss. A FB cn be used s n initil reference to estimte the percentge of oxidtion or nother option is to use the Europen stndrd UN tht specifies minimum purity vlue of 96.5% in mss. Clerly, in this cse the FB complies with this specifiction. The temperture intervl for the FB decomposition is similr to the temperture intervl reported for the therml decomposition of Jtroph oil methyl esters ( C). 21 Typicl thermogrvimetric (TG) nd derivtive thermogrvimetric (DTG) curves of biodiesel smple oxidized t 120 C with n exposition re of 63.6 cm 2 (A 2 ) is presented in Figure 3 to illustrte the pplicbility of TGA in the oxidtion of biodiesel. The TG curve shows the weight loss s function of temperture. Clerly two significnt steps re observed nd confirmed by the DTG curve, which hs two evident peks. The first step is ttributed to the esters tht compose the biodiesel. The second step is ttributed to the oxidtion products. The key tempertures re T mx (FAME) t 216 C nd T mx (OxP) t 315 C, which corresponds to the mximum rte of weight loss of ftty cid methyl esters (FAME) nd oxidtion Figure 1. Scheme of the ccelerted oxidtion process of biodiesel derived from wste cooking oil

3 Díz-Bllote et l. 494 Quim. Nov Figure 2. Typicl thermogrvimetric curve of fresh biodiesel (FB) Figure 4. TG curves of biodiesel smples oxidized t different tempertures nd exposition res (A1<A2<A3) Tble 2. Averge percentge of oxidtion Percentge of oxidtion Are of biodiesel exposed to ir (cm2) 80 ( C) 100 ( C) 120 ( C) 19.6 (A1) 0.4 ± ± ± (A2) 1.3 ± ± ± (A3) 1.6 ± ± ± 3.3 vlues reported re men of triplicte mesurements ± stndrd error of the men. Figure 3. Typicl TG (red trce) nd DTG (green trce) curves of biodiesel oxidized t 120 C nd A2=63.6 cm2 products (OxPs). The difference between these tempertures llows differentition of the FAMEs from OxPs. Figure 4 shows typicl TG curves of biodiesel smples t different tmospheric ir exposure res (A1, A2, nd A3) s well s different heting tempertures (80, 100, nd 120 C). The effect of the re on the mount of oxidtion products is very smll t 80 C in comprison with the residul mss of FB. Due to the poor mount of oxidtion products formed t 80 C, these conditions re not recommendble for n ccelerted oxidtion test. When the temperture is incresed to 100 C, the percentge of oxidtion products increses nd re detectble for res A2 nd A3. The level of oxidtion is remrkble on the smple oxidized t 120 C when the biggest re (A3) is used. The ctlytic effect of temperture on the oxidtion process of biodiesel is well known. An increse in the contct re between biofuel nd tmospheric ir llows mjor oxygen exchnge t the biodiesel/ir interfce. The mount of oxidtion products increses with incresing temperture nd re. The percentge of oxidtion (in mss %) is clculted using eqution (1), s PO (%)= ROP - RFB (1) where PO: Percentge of oxidtion, % m/m; ROP: Percentge of residul mss lost from oxidized biodiesel t Tmx(OxP), % m/m; RFB: Percentge of residul mss lost from fresh biodiesel t the sme Tmx(OxP), % m/m. Tble 2 shows the verge PO tht is formed fter the ccelerted biodiesel oxidtion tests. The tests re conducted t 80, 100 nd 120 C with three different tmospheric exposure res (A1<A2<A3). The mount of oxidtion products increses with incresing the temperture nd re of biodiesel exposed to tmospheric ir. The vlues in Tble 2 re obtined using the PO of fresh biodiesel t Tmx(OxP) s referenced. However, even without fresh smple of biodiesel, rough estimtion of the PO cn be determined using the purity (96.5%) specified in the UN stndrd s reference. Kinemtic viscosity The effect of the oxidtion on the kinemtic viscosity of biodiesel smple is shown in Figure 5. A smll increse in viscosity is observed t 80 C, though none of the smples exceed the upper viscosity limit (6.0 mm2 s-1) specified by ASTM D6751. An increse in temperture to 100 C cuses n increse in the formtion of oxidtion products. The only smple tht complies with the stndrd specifictions is the smple oxidized with the smllest re (A1). When the oxidtion temperture is rised to 120 C, notorious increse in viscosity, regrdless of the size of the re used for the oxidtion process, ws obvious. The increse in viscosity with incresing oxidtion of biodiesel smple hs been reported in previous studies.22 The increse in viscosity hs been relted to high level of biodiesel oxidtion due to the presence of oligomers of high polrity nd moleculr weight forming undesirble deposits.11 In ddition, viscosity hs been ssocited with the methyl ester content nd the products of the trnsesterifiction rection;23 s consequence, chnges in biodiesel purity, due to oxidtion, cuse chnge in viscosity. UV-Vis spectroscopy To illustrte the oxidtion process nd void signl sturtion, the smples of biodiesel oxidized t 100 C were used s representtive

4 Vol. 41, No. 5 Thermogrvimetric pproch for ssessing the oxidtion level of biodiesel smple 495 Figure 7. Kinemtic viscosities s function of the verge percentge of oxidtion products Figure 5. Kinemtic viscosity of oxidized smples of biodiesel t three different tempertures (80, 100, nd 120 C) nd three different res (A1<A2<A3) of exposure to tmospheric ir for UV-vis nlysis. The results re shown in Figure 6. The min bsorption bnds, in the wvelength rnge of nm, re centered pproximtely t 232 nd 270 nm nd these re ssigned to the formtion of conjugted dienes nd secondry oxidtion products. As observed, the pek t 232 nm grows s function of exposure re, wheres the pek t 270 nm increses slightly with the exposure re. This result indictes tht n increse in conjugted dienes, s result of degrdtion of unsturted mono-lkyl esters,24 confirms higher levels of oxidtion with incresed exposure re. Accordingly, the ccelerted oxidtion test chieves the objective nd provides different oxidtion sttes. CONCLUSIONS Biodiesel from wste cooking oil ws oxidized t vrious tempertures nd tmospheric ir exposure res. Smples of oxidized biodiesel were studied by thermogrvimetric nlysis, nd the therml decomposition curves obtined were used to quntify the percentge of oxidtion products formed during the test. In ddition, the effect of the ccelerted oxidtion test on the biodiesel ws confirmed by UV-Vis spectroscopy. From the experimentl results, the following conclusions could be drwn: Under the experimentl conditions of this study, the percentges of oxidtion products mesured t 80 C for ll exposure res were not enough to increse viscosities beyond the upper limit specified by the stndrd ASTM D6751. In this study, the percentge of oxidtion products, s well s viscosities, becme importnt t 100 nd 120 C, prticulrly when the exposure res A2 (63.6 cm2) nd A3 (153.9 cm2) were used. A strong reltionship between the mount of oxidtion products nd viscosities (R2=0.96) ws found, suggesting the potentil use of both prmeters to estimte the oxidtion level of biodiesel smple. Unlike conventionl methods tht only provide mesure of reltive oxidtive stbility, the percentge of oxidtion (PO) provides simple wy to determine n estimtion of the bsolute oxidtion stte of biodiesel smple. ACKNOWLEDGMENTS Figure 6. Typicl UV-visible spectr of biodiesel smples oxidized t 100 C with three different tmospheric ir exposure res (A1<A2<A3) Reltionship between PO nd viscosity Viscosity of liquid is mesure of its resistnce to flow. It is minly ffected by the size of the molecule. Therml degrdtion leds to the formtion of insoluble high moleculr weight polymers7 tht hve more resistnce to flow. Therefore, n increse in oxidtion products results in incresed viscosity. In Figure 7, the viscosities of ll smples re grphed vs. percentges of oxidtion products, without considering the oxidtion conditions. Liner fit between PO nd viscosity (R2=0.96) suggests tht either the mount of oxidtion products or viscosities cn be good quntittive pproch to the bsolute oxidtion stte. This result grees with studies tht show strong correltions between the mono-lkyl esters (biodiesel purity) nd viscosity or mss loss in thermogrvimetric nlysis.17,23,25 The uthors thnk the Ntionl Council of Science nd Technology (CONACYT) of México for the finncil support s well s PhD scholrship of E.T.Veg-Lizm (CVU ), Cátedr-Concyt Projects 1710 (M. Ruiz-Gómez) nd 1568 (E. Hernández-Nuñes). Thermogrvimetrics nlysis were performed t Lbortorio Ncionl de Nno y Biomteriles (LANNBIO), under grnts FOMIX-Yuctán nd CONACYT LAB No REFERENCES 1. M, F.; Hnn, M. A.; Bioresource. Technol. 1999, 70, Fukud, H.; Kondo, A.; Nod, H.; J. Biosci. Bioeng. 2001, 92, Kiss, A. A.; Dimin, A. C.; Rothenberg, G.; Adv. Synth. Ctl. 2006, 348, Leung, D. Y.; Wu, X.; Leung, M.; Appl. Energy 2010, 87, Americn Society for Testing nd Mterils; ASTM D , Stndrd

5 496 Díz-Bllote et l. Quim. Nov Specifiction for Biodiesel Fuel Blend Stock (B100) for Middle Distillte Fuels, Europen Committee for Stndrdiztion; Europen Stndrd EN Automotive fuels Ftty cid methyl esters (FAME) for diesel engine Requirements nd test methods, Dunn, R. O.; J. Am. Oil Chem. Soc. 2002, 79, Meir, M.; Quintell, C. M.; dos Sntos Tnjur, A.; D Silv, H. R. G.; Fernndo, J. D. E. S.; d Cost Neto, P. R.; Pepe, I. M.; Sntos, M. A.; Nscimento, L. L.; Tlnt 2011, 85, Knothe, G.; Fuel Process. Technol. 2007, 88, Zhng, W.-B.; Renew. Sustinble Energy Rev. 2012, 16, Fng, H. L.; McCormick, R. L.; Spectroscopic study of biodiesel degrdtion pthwys, Report , SAE Technicl Pper, Furln, P. Y.; Wetzel, P.; Johnson, S.; Wedin, J.; Och, A.; Spectrosc. Lett. 2010, 43, Svi, E. L.; Herculno, L. S.; Luksievicz, G. V.; Torquto, A. S.; Besso, M. L.; Astrth, N. G.; Mlcrne, L. C.; Energy Fuels 2017, 31, Mthäus, B. W.; J. Am. Oil Chem. Soc. 1996, 73, Dwivedi, G.; Shrm, M. P.; Egypt. J. Pet. 2016, 25, Veg-Lizm, T.; Díz-Bllote, L.; Hernández-Mézquit, E.; My Crespo, F.; Cstro-Borges, P.; Cstillo-Atoche, A.; González- Grcí, G.; Mldondo, L.; Fuel 2015, 156, Chnd, P.; Reddy, C. V.; Verkde, J. G.; Wng, T.; Grewell, D.; Energy Fuels 2009, 23, Knothe, G.; Eur. J. Lipid Sci. Technol. 2006, 108, Sous, F.; Lucino, M.; Ps, V.; Fuel Process. Technol. 2013, 109, Ounji, F.; Nchid, M.; Kcimi, M.; Liott, L. F.; Puleo, F.; Ziyd, M.; Chin. J. Chem. Eng. 2016, 24, Rshid, U.; Anwr, F.; Jmil, A.; Bhtti, H. N.; Pk. J. Bot. 2010, 42, Bnnister, C.; Chuck, C.; Bounds, M.; Hwley, J.; Proc. Inst. Mech. Eng., Prt D 2011, 225, De Filippis, P.; Givrini, C.; Scrsell, M.; Sorrentino, M.; J. Am. Oil Chem. Soc. 1995, 72, Oliveir, I. P.; Souz, A. F.; Lescno, C. H.; Cires, A. R. L.; Muzzi, R. M.; J. Am. Oil Chem. Soc. 2015, 92, Usmnov, R. A.; Mznov, S. V.; Gbitov, A. R.; Miftkhov, L. K.; Gumerov, F. M.; Musin, R. Z.; Abdulgtov, I. M.; J. Chem. Eng. Dt 2015, 60, This is n open-ccess rticle distributed under the terms of the Cretive Commons Attribution License.