Helsinki. University of TechnologyFaculty of Process Engineering and Materials ScienceDepartment of Chemical EngineeringLaboratory of Biotechnology and Food Engineering

FIN-02150 Fspoo, Finland

Technical Biochemistry Report 2/1995

Environmentally Friendly Products Basedon Vegetable Oils

Kerja Lamsa

Dissertation for the degree Doctor of Philosophyto be presented with the due permission

for pUblic examination and debatein Aucfltorium Kem 2

at Helsinki University of Technology (Espoo,Finland)on the 11th of December, 1995, at 12 o'clock noon.

ESpoo, November, 1995.

TO Jyrki, Inka and Riku

CONTENTS

AbstractPrefaceList of publicationsIntroduction1.1. Historical background of chemical esters used in lubricants 11.2. Preparation of fatty acid esters 21.2.1.Monoesters offatty acids 21.2.2.Diesters of fatty acids 41.2.3.Polyol esters of fatty acids 61.3. Lipase catalysed fatty acid ester production 101.3.1.Definition of Iipases 101.3.2.Structure of Iipases 111.3.3.Properties of lipases 111.3.4.Specificity of Iipases 131.3.5.Production of lipases 161.3.6.Reactions of Iipases 181.3.7.Effect of water on lipases 211.3.8.lmmobilisation of lipases 211.4. Aims of the present work 232. Materials and methods 242.1. Vegetable oils 242.2. Alcohols 252.3. Chemical catalysts 252.4. Reagents 252.5. Solvents 252.6. Enzymes 262.7. Analytical methods 262.8. Immobilisation of lipases 282.9. Chemical preparation of 2-ethyl-1-hexylester of rapeseed oil 292.10.Enzymatic preparation of 2-ethyl-1-hexylester of rapeseed oil 292.11.Chemical preparation of rapeseed oil methylester 292.12.Chemical preparation of soybean ethylester 292.13.Chemical preparation of coconut oil methylester 302.14.Chemical preparation of trimethylolpropane ester from tallow oil fatty acids 302.15.Chemical preparation of rapeseed oil trimethylolpropane ester 312.16.Enzymatic preparation of rapeseed oil trimethylolpropane ester 313. Results and discussion 323.1. Chemical synthesis of 2-ethyf-1-hexylester of rapeseed oil 323.1.1.Choice of catalyst 323.1.2.Effect of molar ratio 333.1.3.Effect of temperature and pressure 343.2. Enzymatic synthesis of rapeseed oil 2-ethyf-1-hexylester (I) 383.2.1.Choice of lipase (I) 38

3.2.2.Effect of substrate molar ratio (I) 393.2.3.Effect of lipase quantity (I) 393.2.4.Effect of added water (I) 413.2.5.Effect of temperature (I) 423.3. Small pilot scale enzymatic preparation of 2-ethyl-1-hexylester of rapeseed

oil (II) 433.4. Chemical synthesis of rapeseed oil methylester 443.5. Chemical synthesis of soybean oil ethylester 473.6. Chemical synthesis of coconut oil methylester 473.7. Chemical synthesis of trimethylolpropane ester from fatty acids of tallow oil 483.7.1.Choice of catalyst 483.7.2.Effect of catalyst quantity on esterification of tallow oil fatty acids 493.7.3.Effect of substrate molar ratio on esterification of tallow oil fatty acids 503.7.4.Effect of temperature, time and pressure on esterification of tallow oil

fatty acids 513.8. Chemical synthesis of TMP-ester from rapeseed oil methylester and

trimethylolpropane (III) 543.8.1.Choice and quantity of catalyst (III) 543.8.2.Effect of substrate molar ratio on RME transesterification (III) 553.8.3.Effect of temperature and pressure on transesterification of RME (III) 563.9. Small pilot scale production of TMP-ester from rapeseed oil methylester

and TMP (III) 593.10. Enzymatic preparation of rapeseed oil TMP-ester (IV) 623.10.1.Choice of lipase (IV) 623.10.2.Effect of lipase quantity and its stepwise addition (IV) 633.1 0.3.Effect of substrate molar ratio on enzymatic transesterification of RME (IV) 643.1 0.4.Effect of added water on enzymatic transesterification of RME (IV) 663.1 0.5.Effect of temperature on enzymatic transesterification of RME (IV) 663.10.6. Effect of pressure on enzymatic transesterification of RME (IV) 673.11. Immobilization experiments for the preparation of TMP ester from rapeseed

oil methylester 683.12. Applications and uses of rapeseed oil esters 703.12.1.Methyl-, ethyl-, and 2-ethyl-1-hexylesters of rapeseed oil 703.12.2.Hydraulic fluids made from rapeseed oil esters (V,VI) 713.13. Characterization of rapeseed oil ester based hydraulic fluids (V,VI) 733.13.1 Viscosity (V,VI) 733.13.2.Filterability (V,VI) 743.13.3.Standards of purity and impurity particles (V,VI) 753.13.4.Foaming (V,VI) 753.13.5.Colour (V,VI) 763.13.6.Cold stability (V,vl) 763.13.7.Friction and wear (V,VI) 773.13.8.Oxidation stability (V,vl) 783.13.9.Corrosion stability (V,VI) 793.13. 1O.Biodegradability (V,VI) 803.14. Conclusion of laboratory tests (VVI) 804. Condusions 81

References 83

Uimsa, Merja. Environmentally Friendly Products Based on Vegetable Oils.Espoo 1995. Helsinki University of Technology. Technical Biochemistry Report 2/1995.87p+ app.65p.

UDCISBNKeywords

ABSTRACT

579.66..663.1 ..582.28951-22-2770-3Vegetable oils, transesterification,alcohols, Iipases, biosolvents,biolubricants, biofuels.

The scope of this work was to study and develop new methods and processes for theproduction of vegetable oil esters as raw materials in the manufacture of biodegradablelubricants and solvents. Different esters were produced either chemically orenzymatically. In the chemical process, catalysts were either bases such as alkalimetaloxides or alcoholates or acidic such as phosphoric, sulphUric or p-toluenesulphonic acid.In the enzymatic reactions lipases used were produced by Candida rugosa (formerlycyJindracea), Micor miehei, Pseudomonas Buorescens or Micor sp. The vegetable oilsused were Finnish rapeseed oil, soybean oil, coconut oil and their esters. Alcoholsemployed were mono-, di- and polyols. The produced esters were methyl-, ethyl-, 2-ethyl­1-hexyl- and trimethylolpropane esters.The mono esters were prepared to be used as solvents or fuels or as raw materials forfurther synthesis, preparation of raw materials for hydraulic fluid use. Polyol esters wereprepared to be used as raw materials for lubricants. In the chemical methodstemperature varied from 50°C to 140°C and pressure from ambient to 1.3 MPa.ln theenzymatic methods temperature range was from 30°C to 68°C and pressure limit fromatmospheric to 2.0 MPa. In both cases different new routes were found to producedesired esters with very good yields.In the chemical synthesis the conversion variedbetween 85 % and 99%, depending on the ester. In the enzymatic reactions conversionwas between 82 % and 100 %.2-ethyl-1-hexyl and trimethylolpropane ester were synthetized for the first time fromrapeseed oil. In the enzymatic preparation of the 2-ethyl-1-hexyl ester of rapeseed oilconversion was 99.8%. Chemically, the conversion of 2-ethyl-1-hexyl ester of rapeseedoil was 97.6%. Chemically from rapeseed oil methyl ester was succesfully synthetizedwith trimethylolpropane trimethylolpropane ester with a conversion of 99%. Enzymaticallythe conversion was 98%.All of the esters have already been tested for industrial use either on the laboratory scaleor in field tests. The results were encouraging in respect of lubrication or solventproperties.2-ethyl-1-hexyl ester of rapeseed oil from chemical and enzymatic preparation have beenin field tests e.g.by the detergent industry to replace traditional organic solvents. Theseesters are used today for example in carshampoos. Trimethylolpropane esters ofrapeseed oil (TMP ester), chemically and enzymatically synthetized, have been used asstarting materials for preparing hydraulic fluids for laboratory tests.TMP-ester based hydraulic fluid resisted oxidation better than rapeseed oil or commercialsynthetic ester based fluids. Also cold stability properties of TMP ester were better thanrapeseed oil based or commercial synthetic ester based fluids. The products are alsoreadily biodegradable, non-toxic and do not bioaccumulate in nature.

PREFACE

This work was carried out dUring 1991-1995 at the Laboratory of Oil Milling Division,Raisio Group and at the Laboratory of Biotechnology and Food Engineering, Departmentof Chemical Engineering, Helsinki University of Technology.

My warmest thanks are to Professor Pekka Linko and Docent Yu-Yen Linko, whoseenthusiasm and fine co-operation made this work possible. I am deeply grateful to themfor their valuable advice and the expertise they have put to my disposal.

I express my graditude to Anne Huhtala, Esa Uosukainen,Sari Kiviniemi and TomiVirtanen, all of whom have done excellent work during this long process.

I also want to thank with my whole heart Mari Heikkila, Hannele Sulonen and BirgitVainio, whose practical work was very important in this project.

I could not have succeeded with the English language without the help of M.Sc. AnneliKarjalainen, to whom my sincere thanks are not enough.

To my employer Raisio Group, Oil Milling Division and to my superior, director MattiSoupas, warm thanks for making it possible to carry out this project.

I would also like to give my deepest thanks to my husband Jyrki, and to my parents Ritvaand Pekka for their caring, love and support during these years. Without them I would nothave succeeded in all this. The sweetest and lovecaring thanks to my little sunshinesRiku and Inka, who have given me the strength to continue.

LIST OF PUBLICATIONS

This thesis is based on the following publications, referred to as I to VI in the text.

Linko Y.-Y., Uimsa M., Huhtala A. and Linko P.,Lipase catalyzed transesterification ofrapeseed oil and 2-ethyl-1-hexanol, J.Am.Oil Chem.Soc. 71 (1994) 1411-1414.

II Lamsa M., Huhtala A., Linko Y.-Y.and Linko P., 2-Ethyl-1-hexanol fatty acid ester fromrapeseed oil by transesterification, Biotechnol.Techn.8 (1994) 451-456.

III Lamsa M., Process for preparing a synthetic ester from a vegetable oil, Patentnumber SF 95367/1995.

IV Lamsa M., Linko P., Linko Y.-Y. and Uosukainen E., An enzymatic process forpreparing a synthetic ester from a vegetable oil, Patent number SF 95 395/1995.

v Lamsa M., Ecologically acceptable synthetic hydraulic fluids based on vegetable oils,Proceedings of the 4th Scandinavian International Conference on Fluid Power,September, 1995, Tampere, Finland.

VI Lamsa M., Vegetable oil based lubricants, Finnish Tribol. 14 (1) 1995 39-45.

Additional new results are also presented.

1

II In tiny, tiny drop, You can fit the whole world"Aaro Hellaakoski

1.1NTRODUCTION

1.1. HISTORICAL BACKGROUND OF CHEMICAL ESTERS USED INLUBRICANTS

The development of synthetic esters in lubricant use (1) started in the 1930s in theUnited States and in Germany. In Germany this was mainly due to the second worldwar and shortage of petroleum. Diesters and polyol esters were developed inGermany and first used as aircraft turbine lubricants. Their benefits were good lowtemperature properties and resistance to thermal oxidation. Since 1960 the aviationindustry has used neopentyl polyol esters as lubricants for jet engines (2). Theirimportant feature is the flexibility of physical and chemical properties achieved bychanging the chemical structure of starting materials such as chain length,branching, number of carbon atoms and type of alcohol.In the 1960s synthetic esters for lubricants were further developed for arcticconditions, functioning as crankcase, transmission and gear oils, and hydraulic fluidsand greases. During the 1970s motor oils (3) came into the picture, first semi­synthetic oils, including diesters. Their most important benefits were good lowtemperature properties and low volatility. The first synthetic motor oil esters cameonto the market in 1977 being mostly polyalfaolefins (PAO). The first two-strokeoutboard engine oils based on synthetic esters came onto the market in 1982 inSwitzerland and southern Germany (4). Their main base fluid was neopentylpolyolester of branched fatty acids.The first laws concerning biodegradable lubricants were published in Portugal in1991 by the Ministry for Environment and Natural Resources (4). It obligated the usein outboard two-stroke engines oils, which biodegradability is a minimum of 66 % byCEC-L-33-T-82-test. This test is the first one developed for biodegradation. It wasdeveloped in 80s and has been used quite a long time and is a well acceptedlubricant test. OECD-tests (301-serie) have also been used since the beginning of90s (4).The next step from two-stroke engine oils in biodegradable products werebiodegradable chainsaw oils. They were launched in the middle of 80s. The firstproducts were based on natural esters, triglycerides of Finnish rapeseed oil (5).In the 80s and still at the beginning of the 90s the main raw material forenvironmentally acceptable products was rapeseed oil. Users and researchershave noticed however that applications for example in high temperatures, demand

2

more from their raw material. From this started the development of biodegradablesynthetic esters for lubrication use. Application areas are hydraulic fluids, metalcutting fluids, roller oils, turbine oils, industrial gear oils and compressor oils. Otherapplication areas are cosmetics, textile and fiber industry, heat transfer liquids andplastics industry (6). Fatty acids for the synthetic esters were from vegetable oils oranimal fats. Pure or purified fatty acids were obtained by splitting, fractionating,distilling, dewaxing, hydrating, cracking etc. Alcohols can vary from short chainalcanols to polyols (4,7). Mixtures of esters are used to improve the properties oflubricants. Mono-, di- and polyesters can be combined to achieve for example theright viscosity behavior. Non-hindered dimer esters, made from dicarboxylic acids,can be combined with polycarboxylic acid esters, prepared with alcohols containingside chains. These synthetic esters can be even combined with mineral oil ormineral oil based products ( PAO) (1,8,9).In 1989 the first biodegradable greases, which were based on neopentyl esterscame to the market (4). Also new gas turbine engine oils based on monoerytritol andtrimethylolpropane were invented in 1989 (10).

1.2. PREPARATION OF FATTY ACID ESTERS

1.2.1. Monoesters of fatty acids

In the production of simple esters from short chain alcohols such as methanol,ethanol or butanol and triglycerides from natural sources, most widely used catalystsare sodium or potassium hydroxide, metal alcoholates, - hydroxides, - carbonates or- acetates and different acids (11,12,13,14). An economic production of monoestersin industrial scale should fulfill at least the following criteria: good yield, few by­products, favorable energy balance, little impact on the environment and simpleprocess without complicated technical devices. Only few of the many differentpublished processes actually match with these requirements.

Figure 1: General molecular scheme of monoester synthesis fromtriglycerides and short chain alcohols.

CHz-O-gR

6H- 0-8-RII +

hHz-o-gR'·

triglyceride

3 R""OH

short chainalcohol

CHz-OH

l\catal~t 3 R"'"..o-8-RIIII+ hH- OH~ I

CHz-OH

ester of glycerolshort chain alcohol

3

The least expensive catalysts are sodium and potassium hydroxides. The quantityneeded is around 0.6 % (wlw) of total amount of reactants. The amount of alcoholshould exceed the theoretical value to force the reaction toward the products.Mittelbach et al (11) used in their method a 1.6-fold alcohol excess. In the literaturemany researchers have used acids to neutralize the alkali, but Mittelbach et al (11)found out that washing with 50-60°C pure water several times was a better way toavoid saponification problems.Kusy (15) used in the transesterification reaction 1.75 equivalent of ethanol and 0.5% (wlw) sodium methoxide in methanol. These figures were based on the patent ofBradshaw and Meoly (16). Reaction catalyst was deactivated by adding sulfuric acid(1:1) to reaction mixture, adjust the pH between 5-6. The acidification prevents,according to the authors, the reverse reaction towards the starting materials andhelps to separate the glycerol phase.If the catalyst was not deactivated by acid, Kusy found out that only half of theexpected glycerol was obtained. The product contained mono- and diglycerides,which remained in an inseparable emulsion, if only water washing was done. Thisemulsion was analysed by IR-spectrophotometry and proved to be mono- and di­glycerides.Consequently, Kusy used both neutralisation with acid and water washing with warmwater to remove the rest of the glycerides, salts of sulfuric acid and the rest of theused catalyst.Freedman et al (12) suggested that the optimal excess of alcohol is six times thetheoretical value. They pointed out that acoording to Bradshaw (16) 4.8-fold excesswas enough and jf alcohol was added in three or four subsequent portions, thiscould be reduced from 4.8 to 3.3. Freedman et al (12) studied also the effect of themolar ratio on ester yield, because different researchers had published so variableresults (Figure 2). When the molar ratio of short chain alcanol and triglyceride was6:1, 98 % ester conversion was obtained. When the molar ratio was decreased tothe theoretical level of 3:1, the ester yield decreased to 82 %, also the amount of di­and monoglycerides increased. They were at their lowest at the molar ratio 6:1 andincreased thereafter showing that the conversion to ester was incomplete. Theoptimal temperature was a few degrees lower than the boiling point of the alcoholused. The reaction time varied from a few minutes to few hours, depending on thealcohol used. Effect of temperature could be compensated by a shorter reactiontime. If tile reaction temperature was lower than a few degrees below the boilingpoint, an almost equal conversion of ester could obtained with a longer reactiontime. For example, in the methanolysis of soybean oil at 60, 45 and 32°C after onehour reaction time the conversion was identical to that at 60°C, 45°C and 32°C afterfOur hour reaction time, respectively.

4

Figure 2: Effect of molar ratio on ester yield, natural triglycerides and short chainalcohols, 0.5 % (wlw) catalyst, 60De, 1h).

Ester (_); triglyceride (0); diglyceride (.); monoglyceride (0).

100 ,.I

90 t80 t,

_ 70 +c 60+g I.~ 50!1; 40o I

o ~~~====§~==""'1o i 2 3

molar ",no ( short chain alcohol:vege18bJe oil)

Different catalysts have been also tested. Using 15 %(wlw) sodium hydroxide or 0.5% (wlw) sodium metoxide with a molar ratio 6:1 (alcohol: triglyceride), theconversions were identical after one hour reaction time. With alkaline catalysts, theresults were much better than with acidic catalysts. With acidic catalysts reactiontimes were much longer, required temperatures were much higher, and more excessof alcohol was needed. For example. in soybean oil ethanolysis with unsatisfactoryconversion of ester was achieved 1 % (wlw) sulphuric acid as catalyst, molar ratio20:1 (alcohol: triglyceride), after three and eighteen hours (11).The technical use of monoesters of fatty acids have lately, from the beginning of90s, been mainly fuels or solvents.

1.2.2. Diesters of fatty acids

Diesters are produced in the chemical reaction of two monohydric alcohols with adicarboxylic acid or a diole and a monocarboxylic acid (1,17). Since water is theother product in the equilibrium reaction, it is essential to remove the producedwater.

5

Figure 3: The general scheme of diesters

dicarboxylic acid diester

diol carboxylicacid

diester

The ester synthesis from dicarboxylic acids is more commonly used than esterproduction from diols.For example azelaic, sebatic, oleic, and malic acids and malic anhydride have beenused as the diacids. Most widely employed alcohols are 2-ethylhexanol, i-tridecylalcohol and a mixture of C8-C10-alcohols (1,12,13). One of the most commondiesters used in lubrication is di-2-ethyl-hexylazelate. This ester is prepared from 2­ethyl-hexanol and azelaic acid (1). Diacids have reacted with alcohols using diacid:alcohol molar ratios of 1:1 or 1:3. Catalysts have been either alkaline, e.g. sodiumhydroxide, carbonate or -phosphate or alternatively acidic types. The amount ofcatalyst varies from 0.1 to 0.25 % (wlw). The optimal reaction temperature isbetween 180-200°C, always near the boiling point of the used alcohol, and thepressure atmospheric. Reaction time varies from half an hour to five hours. Theproduced water is always removed to force the reaction toward the side of products(1 ).Diesters have been used in gasturbine engines, compressors, and both hydraulicand two-stroke engine oils. The first lubricants used in gasturbine engines weredioctylsebacates. Later these esters were replaced by azelates and adipates andmore recently by polyolesters (3,18).Diesters have exceptionally good low temperature properties and high viscosity­index, and they operate quite well as lubricants. As an example of diesters, 2­ethylhexanol diesters and their main properties are de~bed in Table 1.

6

Table 1: 2-ethylhexanol diesters and their properties (1,2,17,18,19).

acid viscosity viscosity VI pourpoint37.8°C 98.9°C

(mm2/s) (OC)

adipine 8.2 2.4 123 -562,2,2-tri-methyladipine 12.0 2.8 89 -54atselaine 11.8 3.1 144 -58sebasine 12.5 3.3 155 -51dodecane-dicarb. 15.2 3.9 168 -46nonadecane 23.5 4.8 141 -49ftalic 30.2 4.4 20 -42isoftalic 32.6 4.7 47 -43tereftal 30.5 5.1 100 -44

It is obvious from Table 1 that the use of branched chain alcohols and acids lowersthe viscosity-index more than the straight chain compounds. The use of cyclicsubstrates increase the viscosity markedly.For two-stroke engine oils esters, such as non-hindered dimer-ester can be used. Itis prepared from diacids such as oleic acid dimers combined with linearmonoalcohols (20). Dibasic acid esters are made from diacids like sebatic acid andalcohols (21). This type of esters has the advantage of uniform chemistry, uniformproperties like low temperature properties. The pourpoint can be as low as -55°C.

1.2.3. Polyolesters of fatty acids

Po/yol esters are produced in the reaction between polyhydric alcohols and mono- ordicarboxylic acids. They are also called hindered esters. This reflects the structure ofthe polyols, where the p.carbon atom does not have anylXf-hydrogen in it. The mostwidely used polyalcohols are trimethylol propane (TMP) (A)

7

and other trimethylol alkanes, neopentylglycol (NPG) (8) , pentaerythritol (PE) (C),ditrimethylolpropane (diTMP) and dipentaerytritol (diPE) (22,23,24,25). Some of themare shown in Figure 4.

Figure 4: Examples of nOiX<-hydrogen alcohols used in polyol ester synthesis.

<;:;H2-OHEt-C-CH2-0H

CH2-OH

(A)

MeHO-CH2-¢-CH2-OH

Me

(8)

9H2-0HHO-CH2-9-CH2-OH

CH2-OH

(C)

Acid composition varies widely, depending on the applications of the final polyolester. The acids used in polyol ester synthesis can be short-, /ong-, saturated-,unsaturated-, straight- or branched chainacids. Recently the tendency has been touse mixtures of different types of acids. Some examples of acids used in thesereactions are shown in Figure 6. The following features of the starting compoundsaffect the properties of the resulting ester: molecular weight, the size of the acylgroups, the functionality of polyols, and the method of preparation of the ester or themixture of esters (2,3).Figure 5 demonstrates how the functionality and po/yol chain length and branchingaffects the properties of the ester (26,27).

Figure 5: The effect of polyol chain length and side chains on the properties ofpolyol ester.

Influence of chain length

:>..lJ''';

UJaoUJ...;:>

length

.lJI::

''';a0.

\-l::la0...

I;"C~h~a""'~--n--;l--e--n--g--tLh-

Influence of side chains

:>..lJ...;rnaorn...;:>

8

Figure 6: Different acid types used in the preparation of polyol esters.

RR-y-COOH

R"

monocarboxylic acids

R• I

R-CH-COOH

a..branched acids

Me RMe-Q-(CH2)n-CH-(CH2)n-COOH (OH)x-R-COOH

Me

isocarboxylic acids acids cont. hydroxylic groups

Aviation gas turbine engines work with a polyester composition made by reactingpolyalcohols with monocarboxylic acids as the lubricant (28). An example of suchproducts is the commercial pentaerytritol esterified with a mixture of C5 to C9carboxylic acids (28).Car engine lubricants can be prepared by total esterification of TMP by a mixture ofsaturated aliphatic carboxylic acids, typically by using 134 g (1 mol), TMP, 36.5 g(0.25 mol) adipic acid, 130 g (1 mol) heptanoic acid and 465 g (1.5 mol) isostearicacid (29). The reaction temperature is close to the boiling point of TMP and thereaction time 8 hours. Ethyl Corporation has obtained a patent for producing TMP­ester with a mixture of aliphatic monocarboxylic acids containing 4 to 12 carbonatoms (30). An example of the reaction procedure is as follows: 67 parts of TMP,235 parts of an acid mixture containing 26 % hexanoic acid, 43 % octanoic acidand 31 % decanoic acid are dissolved in 17900 parts of xylene, using sodiumbisulphate (6 parts) as the catalyst and with a total reaction time of 8 h. Water isremoved from the reaction mixture during the entire reaction, and the mixture isfinally washed with 10 % caustic soda and with water until it is neutral.One possibility for the acid composition is to useo<-branched acids (31). This type ofesters have been used in automatic transmission fluids. The reaction could becarried out as follows to obtain an ester yield of 96 %: 25 parts of TMP, 2,5 parts of

a,branched chain acid and 0,5 parts of an acid with a lower molecular weight aremixed. The catalyst is tetra-n-butyltitanate (TNST), which decomposes in air, sonitrogen atmosphere is necessary as a protection. Reaction temperature is variedbetween 180 to 250°C to reflux reaction mixture. The total reaction time is about 8 h.Synthetic polyol esters can also be made by using a single purified acid together

9

with polyols (32). The acid can, for example, be a commercial pelargonic acid (4582parts), the other components TMP (1206 parts) and toluene (260 parts) as solventand p-toluene sulfonic acid (8 parts) as catalyst. The mixture is refluxed for 18.5hours with a maximum temperature of 216°C and water is separated continuously.The crude ester is obtained upon removing of the excess of alcohol and solvent bydistillation and purifying by calcium oxide water at 50-60°C. Water is then distilled offand, decolorising carbon added. The esterification can also be done stepwise. Thereare at least two different types of acids, which can be added. One example isdescribed by Metro, Hoffman and Matuszak using TMP as the alcohol (33).TMP (1 mol,134 g) and neo-octanoic acid (1 mol,150 g) are mixed with sodiumbisulfate (0.4 % (wlw)) as catalyst and heptane as the water-removing agent. Thereaction mixture is then refluxed at atmospheric pressure, until 1 mol of water hasbeen removed. Then pelargonic acid (12.2 mol,34 g) is added together withadditional catalyst ( 10.1 % (wlw)). The reaction mixture is heated to reflux so that 2moles of water are separated.An improved lubricant composition has been achieved by using a mixture of acidsincluding straight chain acids (C6 to C10) and iso-acids (C6 to C10) for example iso­nonanoic acid (10). An example of a reaction procedure of this type is the following:66.7 parts of TMP, 20.2 parts of straight chain acids and 13.1 parts of iso- acids isused as reactants, tinoxidate as the catalyst at 238 to 243°C at 3.3 kPa with nitrogenatmosphere. Water has to be separated during the whole reaction time. The totalreaction time is controlled by the hydroxyl value and the reaction is stopped, whenthe hydroxyl number is sufficiently low.

Table 2: Properties of different acid containing polyol esters(19,28,29,34,35).

Type of aCId Alcohol Viscosity VI Pour-40'C point

(mm'ts) ('Cl

non -hydrogen TMP 23.9-35.0 -40-59

-branched TMP 32.6 137 -37

straIght chainmonocartloxylic di-TMP 27.9-28.1 135-140 -57-59

straight chainmonocartloxyhc 70% TMP 24.8 144

30% di-PEmixture of straight

-50chain and isochain TMP 23.9

mixture of straight-54chain and isochain PE 23.8

mixture of mono-and dicartloxylic TMP 135-152 -20-40

hydroxyl containing TMP 20.5-35-2 106-167

hydroxyl containing PE 43.7-142.3 85-186

unsatisfactory PE 22.0-46.0 180-200 -30-40

10

1.3. LIPASE CATALYZED FATTY ACID ESTER PRODUCTION

1.3.1. Definition of lipases

Lipases (triacylglycerol acylhydrolases EC 3.1.1.3 ) are esterases widely found in theanimals, micro-organisms and plants. Their main biological function is to catalyse thehydrolysis of triacylglycerols to fatty acids, diacylglycerol, monoacylglycerol andglycerol (Figure 7). The reaction is reversible, so under certain conditions theenzymes can also catalyse ester synthesis such as the formation of acylglycerolsfrom glycerols and free fatty acids (36-39).

Figure 7: Lipase reaction

Triglyceride

HDiglyceride + free fatty acid

HMonoglyceride + free fatty acid

HGlycerol + free fatty acid

The natural substrates of Iipases are triacylglycerols of long chain fatty acids. Theyhave a very low water solubility but the lipases can quickly catalyse the hydrolysis ofester bonds at the interface between an insoluble substrate and water. The ability oflipases to catalyse the hydrolysis of water inspluble fatty acid esters separates themfrom other esterases, which catalyse the hydrolysis of water soluble esters. Lipasescatalyse a large number of the hydrolysis of water insoluble fatty acid esters,although the best substrates are acylglycerols.

11

. 1.3.2. Structure of Iipases

Lipases studied consist of 270 to 641 amino acids and they have a certain degree ofsimilarity, regardless of their origin. Some researchers have proposed, that Iipaseshave an especially high proportion of hydrophobic amino acids and this results in theinteraction of hydrophobic amino acids and substrates (36,37). However, theresearch of pUblished amino acid compositions clearly shows that as a group Iipasesare no more hydrophobic than other enzymes. The strong interaction withhydrophobic compounds is propably due to the hydrophobic patches on lipasesurface. In the lipase from pancreas and MiJcor mieheifor example, the activecentres contain a structurally analogous Asparagine- Histidine-Serine triad (41). Inthe Geotrichum lipase a catalytic triad of a similar structure Serine- Histidine-Glutinewas found. For each of these three lipases, the catalytic site has been found to becovered by a helical lid, which opens at the lipid-water interface or in an organicsolvent. The lipases thus undergo a conformational change before becomingcatalytically active. The molar mass can vary from 27 000 g/mol ( lipase ofPenicillium cyclopium) to 500 000 g/mol ( lipase in human saliva) (36,37).Several Iipases from different sources have been pUrified and crystallized. Mostpurified Iipases contain 2- 15 % carbohydrates with usually glucoside residuemannose. Galactose, xylose, arabinose and hexoseamine have also been reported.The role of the carbohydrate moiety has not yet been clarified and its importance tothe activity has been questioned (36,37,41).

1.3.3. Properties of Iipases

Lipases are generally quite stable in neutral water solutions at room temperature.Microbial lipases are usually more thermostable than animal or plant Iipases (42).Lipases produced by Aspergillus niger, Rhicopus japonicus and Chromobacteriumviscosum are stable even at 50°C. Humicola lanuginosa produces a lipase, that isstable at 60°C. The Pseudomonas Buorescens enzyme used in heat-treated dairyproducts is only partly inactivated in these processes (36).Generally Iipases are active in a wide pH range with the highest activity at a pH 6 to8. Extracellular Iipases produced by Aspergillus niger, Cromobacterium viscosum andPhizopus arrhizus are extremely active at a low pH and Pseudomonas nitroreducanslipase is active at pH 11 (36).Microbial lipases have an exceptionally high surface activity at air-water interfacewhen compared to other proteins. This high surface-activity leads to a strongadsorption of Iipases to hydrophobic surfaces and can explain the high activity ofIipases with insoluble substrates (37,43). Table 3 lists some properties ofcommercially available lipases (36,37,43).

12

Table 3: Some properties of commerciallipases (36,37,43)

Species Molecular Carbohydrate Specific Averageweight content activity+ hydrophob.(0/0)

Aspergillusniger 25000 10 2400 1 058CandidarugosH> 120000 4.2 1 140 1 150

Chromobacterium 30000 0.0 5780 986viscosU11f'Geotrichum 54000 7.0 447 1 135candidwrfHumicola 27500 0.0 1 490 1 079lanuginosae

Afucor 21 000 2.6 1 322javanicuS£Pseudomonas 32000 0.0 4200 1 001Duorescen#Rhizopus 43000 6.7 9300 1 097arrhizu1'Rhizopus 41 300 4000 1 270delemarJ-

+= From purified lipase ( lipase units! mg protein)a= M.P.Tombs, private communicationb= Tomizuka et ai, 1966c= Isobe and Sugiura, 1977; Horiutu and Imamura, 1977d= Tsujisaka et aI, 1973e= Liu et aI, 1973f= Ishihara et ai, 1975g= Sugiura and Oikawa, 1977; Sugiura et aI, 1977h= Semeriva et ai, 1969; Laboureur and Labrousse, 1968j= Chiba et ai, 1973; Iwai and Tsujisaka, 1974

13

1.3.4. Specificity of lipases

The specificity of Iipases can be divided into five main groups: Iipid-, regio- and fattyacid specificity, stereochemical specificity and a combination of these. The degree ofspecificity varies widely from distinctly specific to weakly or even non-specific. Lipidspecific Iipases have a high activity in hydrolysis of mono-, di- and triglycerides. Forexample Penicillium cyclopium lipase has the highest activity to monoglyceridesubstitutes (37,38,44).Regiospecificity can be devided into three groups: 2-specific, 1,3-specific and non­specific (31,37). Figure 8 shows some regiospecific reactions. Non-specific lipasescatalyse the total decomposition of triglycerides to free fatty acids and glycerolswith di- and mono-glycerides as products. When a non-specific lipase is used, theresultant fatty acid composition is similar to that obtained with chemicaltransesterification. If there is a double bond or a bulky substituent in the fatty acidnear to the carbonyl group, there is a resistance to the attack by enzyme. Suchcomponents are poor substrates for lipase for steric reasons. Extracellular Iipasesproduced by Candida rugosa, Cromobacterium aenes, Propiobacterium aenes,St3phyloroccus aureus and Geotrium cadicum show no marked regiospecificity(36,37,44).The second group of Iipases are regiospecific, releasing fatty acids from 1- and 3­positions of triacylglycerols. 1,3- Specified lipases produce free fatty acids, 1,2(2,3)­diglycerides and 2-monoglycerides.as reaction products. Because the last two arechemically unstable, under longer reaction times a total migration of fatty acids takeplace. Regiospecific Iipases are produced for example by Aspergillus niger, Mucorjavanicus and Phiedzopus arrhizus. Also lipase Mucor miehei has been claimed to be1,3-specific (44,45).

14

Figure 8: Products formed by lipase catalyzed hydrolysis of triglycerides.

1) Non-specific lipase

R COO-y H2 9H2-0HR'COO-CH~RCOOH + R'COOH + R"COOH + CH- OHR"COO-CH2 CH2-OH

2) 1,3-specific lipase

RCOO- yH2 yH2-OH GH2-OHR'COo-CH;=R COO-CH + RCOOH~R' COO-CH + R"COOHR"COo-CH2 R"COO-CH2 CH2-OH

3) Fatty acid specific lipase

RCOO- yH2 R'COO-QH2 HO-QH2 R'COO-9H2R'COO-CH + RCOO- CH~2 RCOOH + R'COO-CH + CH-OHR"COO-CH2 R"COO-CH2 R"COO-CH2 R"COO~CH2

In Figure 9 the reactions of non-specific and 1,3-specific microbial Iipases have beendemonstrated.

15

Figure 9: Positional specificity of microbial lipases.

1.Reaction catalysed by non-specific Iipases

2. Reaction catalysed by 1,3-specific lipases

CH2-o-gR CH2-OHo I 0 I 0 !

R-t-O-CH ~ R-~-O-CH + RCOOH~R-t-O- CH + 2RCOOHI 9 ! q !CH2 -O-C-R CH2-O-C-R HO-CH2

2-specificity is extremely rare. For example Geotrichum viscosum lipase is claimed tohydrolyse the fatty acid in the 2- position in oleic and linoleic acids (44).One group of Iipases also express specificity towards a certain group of substrates.Specificity can be for the chain length of carbonatoms in substrate, for a quantity ofunsaturated bonds or for the place of a double bond in the carbon chain (45). Forexample lipases of Candida rugosa, Aspergillus niger, Geotrichum candicum catalysewell the release of fatty acids of 18 carbom atoms, which have a cis- double bond inposition nine, as in oleic and linolene acid (36,46,47,48). The lipase in wheat embryois selective for monosubstituted fatty acids, and pancreatic lipase hydrolyses mostactively C4 saturated substitues of fatty acid chains.Stereochemical specificity of some lipases has been reported for straight chainsecondary alcohols, optically active esters of acetinide and butyric acid,cyclohexanals, 2-benzyl and glyceral ethers, sugaralcohols and enantiomeric estersof ibuprofen (44). The human and rat saliva Iipases hydrolyse enantiomeric estersfrom the 3-position. Many lipases are also stereo specific, which has been utilized inoptical resolution and in chiral synthesis. For example Candida rugosa lipase exhibitspreference toward (L)(-)-menthol (49).

16

1.3.5. Production of lipases

Lipases are produced by plants, animals and micro-organisms (39,45). The majorityof commercially employed Iipases are of fungal origin, and at least 20 are availablein commercial quantities (50). The porcine pancreatic lipase has been widelystudied. Also milk contains Iipases (51) and according to Hills and Mukhe~ee

rapeseed is a promising source for Iipases (52).The formation of lipase is usually supported by the presence of mono- ordisaccharides, glycerol and/or lipids in the growth medium (40).Carbon sources used in the production of extracellular Iipases are mainlypolysaccharides, such as starch, triglycerides or fatty acids. Common nitrogensources are soybean meal, yeast extract or com steep liquor. The most widely usedassay of Iipases is based on emulsions of insoluble triglycerides like olive oil. Anexample of microbial lipase producers is shown in Table 4 (39,42).

17

Table 4: An example of microbial lipase producers (39,42).

Manufacturer Country Organism

Amano Pharmaceu- Japan Aspergillus nigertical CO,Ltd Rhizopus niveus

Pseudomonas Buorescens

Biocatalysts Ltd United Aspergillus nigerKingdom Candida rugosa

Chromobacterium viscosumMiJcor mieheiPseudomonas Buorescens

Gist-Brocades NV Holland MiJcor miehei

Hughes and Hughes United Rhizopus arrhizus(enz.) Ltd Kingdom

John & E Stuge Ltd United Aspergillus nigerKingdom

Meito Sangyo CO,Ltd Japan Candida rugosa

Novo Industry AfS Denmark Aspergillus nigerlYfncor miehei

Osaka Saikin Japan Rhiropus japonicusKenkyusho

Sapporo Breweries Japan Pseudomonas fi-agi

Toyo Jozo Co, Ltd Japan Chromobacterium viscosum

18

1.3.6. Reactions of Iipases

In addition to the hydrolysis of fats and oils transesterification is one of the mostcommon reactions performed by Iipases. Different types of reactions are shown inFigure 10. Transesterifications without solvent have several benefits. Substrates canbe used in high concentrations. If substrates are used in stoiciometry, the finalproduct can easily be separated, for example by filtration. There are also fewerbyproducts and waste treatment is easier. The process is safer, because reactiontemperature and pressure is relatively low (43,51,52,53).Transesterifications have been used in oil and fat industry in order to modify thecomposition of fats and oils and to change the physical properties of triglycerides tothe desired direction (38). One major difference between chemical and enzymaticinteresterification reactions is that one can obtain acyl migration with 1,3-specificenzymes, which is impossible to achieve with chemical catalysts.

19

. Figure 10: General reactions by Iipases.

I Hydrolysis of ester

R-Ko-R + Hp -- R-8-0H + HO-R

II Synthesis of ester

R-KoH + HO-R R-Ko-R + Hp

III Transesterification

III A Acidolysis

R1-g·O-R + Rf gOH --Rz-g·O-R + R1-gOH

III B Alcoholysis

R-g.O-R 1 + HO-Rz ---- R.gO.Rz + HO-R1

III C Interesterification

R1-g-O-R1 + Rzg·O-Rz---R1-go-Rz + Rz-gO-R1

1110 Aminolysis

R.g-O-R1 + HzN-Rz --- R-8-NH-Rz + HO-R 1

20

Alcoholysis of long-chain fatty acids has been investigated by quite a fewresearchers. One example is celite immobilized Pseudomonas Duorescens lipase ascatalyst in the alcoholysis of tripalmitine, trioleine and olive oil together with ethanoland isopropanol (54,55). Lipozyme (a Milcor miehei lipase, immobilized onphenolformaldehyde resin; Novo Nordisk, Denmark) has been used for example intransesterification of trioleine and stearyl alcohol. Also di- and mono-oleins wereproduced (43). Enzymatic ester hydrolysis is described in general in Figure 11 (55).The acyl-enzyme intermediate is similar to that of serine protease.The removal of water during the reaction clearly affects the final yield of ester. Anexample of this is the reaction of oleic acid with oleyl alcohol. The optimum yield of85 % reported, when no water was removed during the reaction. If the reaction wascarried out in vacuum and the produced water was continously eliminated, theesterification was forced to completion (55).

Figure 11: Enzymatic ester hydrolysis.

HOH» R'OHo Enz 9 EnzRC-OR'~ RC-Enz---:;;;;=:

R OH HOH

R'-OH> HOHo Enz 9 Enz

RC -OR'~ RC-Enz---:.;;;;;:::R OH HOH

qRC-OH

R~-OH

Miyoshi Oil and Fat Co of Japan have used Candida rugosa lipase to hydrolyse oilsfor soap production (56). Enzymes were used instead of the conventional chemicalprocess to achieve better odour and colour of the products. It has also beenclaimed that the enzymatic process is overall more economic than the chemicalmethod (56).An example of an esterification reaction with lipases is the production of geranyl andmenthyl esters from butyric acid and geraniol or lauric acid and menthol,respectively. The final products are high-valued chemicals, where the chemicalprocess is expensive and complicated (49).It has been discovered recently that lipases also catalyse the formation ofperoxycarboxylic acids (41). An example is the epoxidation of cyclohexene with

21

lipase, hydrogen peroxide and a catalytic amount of a long or medium chaincarboxylic acid.

1.3.7. Effect of water on Iipases

Water is essential for enzyme catalysed reactions. Usually a low water activity isfavourable for lipase catalysed ester synthesis, while excessive water favorshydrolysis (57,58). Water is essential for lipase, which will become inactive due tostructural changes when all water is lost (42). It is assumed that in an environmentof very low water content, the conformational space of the enzyme is restricted. Incompletely waterless environment, there is no space for comformational changesnecessary for substrate binding. Lipases from moulds (like Rhimpus and Penici11umsp.) seem to be more tolerant to low water actiVity than bacterial Iipases forexample Pseudomonas sp. (44,59). The water content is also important, affectingreaction rate, product yield, selectivity and operational stability. For example whenpig pancreas lipase is wet (3.6 % (wlw) , the reaction speed for transesterification isclearly faster than for "dry" lipase (0.48 %(wlw» (60). Water quantity affects alsosubstrate specificity. Many Iipases have an optimum range for water quantity. Forexample commercial Candida rugosa lipase shows little or no activity without addedwater, but by adding water reaction rate rises markedly until it reaches themaximum. Recently the emphasis in lipase studies has been on the measuring andcontrol of water activity in lipase catalyzed reactions (61). One way is to continuouslyadjust the water activity in the headspace above the reaction mixture. This can bedone by operating a reactor with a saturated water solution in contact with thereaction mixture.Another possibility is to use pairs of salt hydrates acting as buffers of water activity.

1.3.8. Immobilization of Iipases

Because Iipases are active at the interface between water and oil, the total surfacearea has a pronounced effect on the reaction rate. The surface can be increasedeither by emulsifying or by immobilising the enzyme to a proper carrier. The purposeof immobilisation is also to bring the lipase into a form, in which it can be physicallyseparated from the substrate or product for reuse. Additionally, with this techniquereactions can be carried out continously or repeatedly, and the thermo and storagestability of lipases can be improved.The carrier should be chosen in such a way that the essential water in lipase isretained. The carrier should also have sufficient mechanical strenght, microbiologicaldurability, chemical resistance and functionality, hydrophobic or hydrophilic characterand the ability to regenerate easily (44).There are two main categories for immobilisation, the chemical methods, in which

22

covalent bonds are created between the lipase and carrier and the physicalmethods, in which weaker interactions in lipase are used. The most commonimmobilisation method is adsorption to a proper carrier, like Celite, cellulosederivatives, porous glass or ceramic pearls, aluminum- or titanium oxide andionexchange resins. Lipases can also be immobilised by entrapment in agaros incellulosic matrix, calcium alginate or polyacrylamide. However, there is no generalimmobilization technique available for every process and an optimal applicationshould always be developed for each specific case (62).

23

1.4. AIMS OF THE PRESENT WORK

The main purpose of this work was to find new biotechnical and chemical ways toproduce in high yields vegetable oil based esters. Another purpose was to inventnew uses of vegetable oil based esters. This was planned to do by laboratory- andfield tests.The alcohols used were chosen by the end use of the desired ester. They wereshort, straight chain alcohols such as methanol, ethanol or long, side chain alcoholslike 2-ethyl-1-hexanol or polyols like trimethylolpropane.Rapeseed oil, soybean oil and coconut oil or their esters, and tallow oil fatty acidswere used as the acid donor.The transesterification reaction was to prepare monoester from vegetable oil andshort, straight chain alcohol (11,12,13,14). Reaction conditions were tried to beimproved, so that, for example, industrial scale production of rapeseed oil methylester (RME) could be started. RME was also planned to be used as a solvent.Because RME caused swelling and brittling of some gum materials, a more suitablealcohol (2-ethyl-1-hexanol), was succesfully found from literature (17,18,19).The 2-ethyl-1-hexyl ester of rapeseed oil was planned to prepare both chemicallyand enzymatically. It was ment to compare the preparation methods, their costs andconversion and purity of the wanted ester. 2-ethyl-1-hexyl ester of rapeseed oil wasplanned to be tested as solvent in for example cleaning printing ink rollers or as toreplace organic solvents in carshampoos.Different polyolesters were considered as raw materials for lubricants, especially

hydraulic fluids. Their raw materials were as acid donor tallow oil fatty acids or RMEand as alcohol trimethylolpropane (TMP). These reactions were planned to do eitherchemically or enzymatically for the same reasons as for esters mentioned above.From obtained esters hydraulic fluids were meant to prepare with additives. Theseesters were tested in laboratory scale by standardized methods.

24

2. MATERIALS AND METHODS

2.1. VEGETABLE OILS

Rapeseed oil

Rapeseed oil was distil/ed and raffinated by Raisio Group, Oil Milling Division and wasentirely based on Finnish rapeseed.The medium fatty acid composition was:oleic acid 57 %Iinolic acid 22 %linolenic acid 12 %palmitinic acid 4 %stearic acid 1 %eicosenic acid 2 %erucic acid 1 %others 1 %

The medium molar mass of rapeseed oil was 880 g/mol.

Soybean oil

Soybean oil was also distilled and raffinated by Raisio Group, Oil Milling Division.The medium fatty acid concentration was:linoleic acid 55 %oleic acid 22 %palmitinic acid 11 %linolenic acid 8 %stearinic acid 4 %

The medium molar mass of soybean oil was 885 g/mol.

Coconut oil

Coconut oil was a commercial product imported by Raisio Margarine.The medium fatty acid concentration waslauric acid 46 %myristic acid 18 %palmitic acid 10 %oleic acid 7 %caprylic acid 6.5 %isodecanoic acid 6.5 %stearic acid 3 %linoleic acid 2.5 %caproic acid 0.5 %

The medium molar mass of coconut oil was 795 g/mol.

25

Tallow oil

Commercial tallow oil was purchased by Raisio Chemicals.The medium fatty acid concentration was:oleic acid 41.5%stearinic acid 15 %myristic acid 1.5 %palmitoleic acid 1 %eicosenic acid 1 %lauric acid 1%others 39 %

The medium molar mass was 275 g/mol.

2.2. ALCOHOLS

Methanol (in the laboratory p.a., J.T. Baker, The Netherlands; at the factory Neste ResinsLtd., Finland)

Ethanol Aa (Alko Ltd., Finland)2-Ethyl-1-Hexanol (J.T. Baker, The Netherlands or Fluka Chemie Ag, Switzerland)Trimethylolpropane (J.T. Baker, The Netherlands)

2.3. CHEMICAL CATALYSTS

Potassium hydroxide (Merck, Germany)Sodium hydroxide (Merck, Germany)Sodium methoxide (J.T. Baker, The Netherlands or Merck-Schubert, Germany)Sodium ethoxide (J.T. Baker, The Netherlands)Phosphoric acid (Merck, Germany)Sulphuric acid (Merck, Germany)p-Toluenesulphonic acid (Merck, Germany)

2.4. REAGENTS

Hydrogen chloride (Merck, Germany)Sulphuric acid (Merck, Germany)Sodium sulphate (Merck, Germany)

2.5. SOLVENTS

Toluene (J.T. Baker, The Netherlands)Heptane (J.T. Baker, The Netherlands)Acetone (Merck, Germany)Acetone p.a. (Riedel de Haen, Germany)

26

2.6. ENZYMES

Lipases from the following micro-organisms were usedCandid1l rugosa (formerly cylindracea) (Biocatalyst Ltd., Great Britain, batches 1911021,292173, activity 42500 U/g, batch 9922238, activity 80000 U/g; Meito Sangyo Co Ltd.,Japan, batch L2502, activity 360000 Ulg)Chromobacterium viscosum (Biocatalyst Ltd, Great Britain, batch 21690, activity 13300Ulg)Mucor miehei (Biocatalyst Ltd, Great Britain, batch 200906, activity 7200 Ulg; Lipozyme1M 20, Novo Nordisk, Denmark, immobilised to porous, weakly alkalic anion exchangeresin, activity 380 Ulg)Pseudomonas fJuorescens (Biocatalyst LTd., Great Britain, batch 24090111, activity 11900Ulg; batch 10922342, activity 7700 Ulg)MUcor sp. (Lipase M-AP 10, LMK 10529, Amano, Japan, activity 1700 U/g)

2.7.ANALYTICAL METHODS

Thin-layer-chromatography (TLC)

Thinlayer Kieselgel 60 F 254 (Merck, Germany) plates were used.Standards:All standards made in Raisio Group, Laboratory of Oil Milling Division, were purified withpreparative TLC. The standards were:rapeseed oil (Raisio Group)trioleine (Raisio Margarine or Sigma, USA)dioleine (Raisio Margarine or Sigma, USA)mono-oleine (Raisio Margarine or Sigma, USA)methylester of rapeseed oil (Raisio Group)TMP-oleate (Unichema, The Netherlands)TMP-ester of rapeseed oil (Raisio Group)2-ethyl-1-hexylester of rapeseed oil (Raisio Group)

Eluents

Hexane (Merck, Germany)Diethylether ( Merck, Germany)

A41

ratioB91

C964

DHexane (Merck, Germany) 40Diethylether (Rathburn, Great-Britain) 10Acetic acid (Riedel de Haen, Germany) 1

27

colouring solutions ratio

I Ethanol Aa ( Alko Ltd., Finland) 50Sulphuric acid ( Merck, Germany) 50

II Acetic acid( Riedel- de Haen, Germany) 100Sulphuric acid ( Riedel- de Haen, Germany) 2Anisealdehyde ( Merck, Germany) 1

III 2-7-Dichlorofluorocene ( Aldrich-Chemie, Germany) 0.1 % inethanol Aa (Alko Ltd, Finland)

High pressure liquid chromatography (HPLC)

Method 1.pump: Waters 501 HPLCdetector: Waters differential refractometerintegrator. Merck-Hitachi D 2000column: Ultrastyregel 500 A (Waters-Millipore)

Ultrastyregel 100 A (Waters-Millipore)eluent: Tedrahydrofuran HPLC-quality (Merck, Germany), flow rate 0.5 mm3/min.standards: rapeseed oil (Raisio Group)

methylester of rapeseed oil (Raisio Group)trioleine (Raisio Group)dioleine (Raisio Group)mono-oleine (Raisio Group)

Samples were diluted to tedrahydrofuran (HPLC-quality) and filtered through 0.451Jmdisposable filter.

Method 2.pump: Beckman system gold 126detector. CUNOW light scattering detector DNL 21column: Spherisorb (C 18) (Phase separations) reversed phase, particle size 3lffn,

length 15 em, diameter 4.6 mmgradient: 30 % A and 70 % B to 70 % A and 30 % B, time 0.5 h

(A: CI2CH2+CICH2CH2CI (50:50);B: CH3CN)

28

Method 3.

pump: Perkin-Elmer series 4 pump moduledetector: Hewlett-Packard 1047 A-R-refractometerintegrator: Perkin-Elmer 316 or SI-316 satellite integratorcolumn: Nova-Pak C 18 (Nordion) or Ultrastyregel 500 Aand 100 A (Waters-Millipore)eluent: Tedrahydrofuran (HPLC-quality)standards: 2-ethyl-1-hexylester of rapeseed oil (Raisio Group)

rapeseed oil (Raisio Group)methylester of rapeseed oil (Raisio Group)TMP ester of rapeseed oil (Raisio Group)trioleine (Raisio Group)dioleine (Raisio Group)mono-oleine (Raisio Group)

Samples were diluted to tedrahydrofuran (HPLC-quality) and filtered through 0.451Jmdisposable filter.

Gas chromatography

Samples were methylated by IUPAC 2, 301-method and analysed using aHewlett-Packard 5890 gas chromatograph.column: NB 351 (Nordion), length 25 m, diameter O. 32J.lm, film strength 0.21Jm driving­

program: 1 min 70°C, 10°C/min to 240°C.

Infrared spectrophotometry

Perkin-Elmer 883 spectrophotometerPerkin-Elmer FTIR 16 PC spectrophotometercuvette: NaCII 0.025 mm.tablet: KBr

2.8. IMMOBILlSATJON OF L1PASES

Reagents: Na2HP04x2HP (Merck, Germany)NaH2P04xHP (Merck, Germany)

Carriers:Carriers were washed with hot, deionised water. The lipase solution was made byslurrying 6 g lipase in 100 cm3 0.05 M sodium phosphate buffer (pH 5.8), mixed for 2hours and filtered. Buffered carrier (40 g) and enzyme solution (60 cm3

) were mixed for 6hours at 26°C. Total mixture was filtered and cold-dried (20°C) for 30 hours to dry-solidcontent 99 %.

29

2.9. CHEMICAL PREPARATION OF 2-ETHYL-1-HEXYL ESTER OF RAPESEED OIL

Reaction conditions were as follows: 3.5 9 (0.4 mol) rapeseed oil, 2.1 9 (1.6 mol) 2-ethyl­1-hexanol and 0.5 % (wlw) catalyst. Rapeseed oil and alcohol were measured to a 50cmo three neck flask, equipped with a thermometer, stirrer and condenser. Catalyst wasadded and stirring was started. Mixture was heated to 60-70°C and mixed until thecatalyst was melted into the reaction mixture. Reaction mixture was then heated to 190­200°C in oil bath, so that alcohol was refluxing. The progress of the reaction wasfollowed by TLC. Mixture was neutralized by HCI-water or NaOH-water (1:1) dependingon the catalyst. The excess alcohol was distilled at 120°C/2.6 MPa for two hours.Glycerol was separated and this phase was discarded. The product was washed threetimes with warm water (50°C) and dried over sodium sulfate.

2.10. ENZYMATIC PREPARATION OF 2-ETHYL-1-HEXYL ESTER OF RAPESEED OIL

Reaction conditions were: Rapeseed oil 0.2 g (0.28 mol), 107 dmo (0.68 mol) of 2-ethyl-1­hexanol (molar ratio 1:3) and 3.0 % (wlw) added water were mixed in capped 13 mmo

test tubes equipped with a magnetic stirrer at 200 rev min,1 at 37°C. The reaction timevaried from 12 to 72 hours. The lipase was then separated by centrifugation for 5 min at2000 rev min-1. The supernatant was pipetted into Eppendorf tubes and stored at -20°Cfor later analyses.

2.11. CHEMICAL PREPARATION OF RAPESEED OIL METHYL ESTER

Rapeseed oil (0.3 mol) was weighed into a 100 cmo three neck flask, equipped with athermometer, condenser, stirrer and sample adapter. Stirring was started and methanol(2.0 mol) was added. Reaction mixture was heated to 60°C and the alkaline catalyst usedwas added 0.5 % (wlw). After four hour reaction time all rapeseed oil had reacted byTLC. Reaction mixture was washed by acidic water. Glycerol was separated and theexcess alcohol was distilled. The reaction mixture was analysed by HPLC. The RME­content varied from 95 % to 99 %.

2.12. CHEMICAL PREPARATION OF SOYBEAN OIL ETHYLESTER

Soybean oil (0.25 mol) was weighed into a 100cmo three neck flask, equipped with athermometer, condenser, stirrer and sample adapter. Stirring was started and ethanol(1.5 mol) was added. Reaction mixture was warmed to 80°C and 0.5 %(wlw) of alkalinecatalyst was added. Stirring was continued for two hours. The reaction was monitored byTLC. Upon completion the reaction mixture was neutralised by acid water and washedwith warm water. The excess alcohol was distilled and glycerol separated. Reactionmixture was analysed by HPLC. A typical conversion was 89 %.

30

2.13. CHEMICAL PREPARATION OF COCONUT OIL METHYLESTER

Reaction conditions were similar to those of rapeseed oil esterification; amount ofsubstrates were coconut oil (0.34 mol), methanol (2.0 mol) and an alkaline catalyst0.6 % (wlw). Reaction time was five hours. Conversion was typically 97 % asdetermined by HPLC.

2.14. CHEMICAL PREPARATION OF TRIMETHYLOLPROPANE ESTER FROMTALLOW OIL FATTY ACIDS

Two methods were used.

A. Reaction conditions were: Fatty acids (161 g, 0.59 mol) were melted, warmed to60°C and TMP (25 g, 0.19 mol) was added under mixing. TMP was allowed to meltand mix properly. Solvent (45 mm3

) and catalyst (0.5 % (wlw» were then added andmixed properly. Temperature was raised to 120°C and kept there for seven hours.The reaction progress was monitored by TLC. After the completion of the reactionabout 11 mm3 of water was separated. Heptane was separated. Reaction mixturewas neutralized using sodium hydroxide water and washed twice using warm water(50°G).

B. Reaction conditions were: Fatty acids (161 g, 0.59 mol) were melted, warmed to60°C and TMP (25 g, 0.19 mol) was added under mixing. TMP was allowed to meltand mix properly. Solvent (45 mm3

) and catalyst (0.5 % (wlw» were then added andmixed properly.Reaction temperature was elevated to 70-75°C and a slight vacuum was introduced,so that reaction mixture refluxed and water was separated. Reaction temperaturewas kept around 70°C during whole reaction, but the pressure was reducedsmoothly. The amount of water separated was 155 mm3

. The reaction mixture wasneutralized using sodium hydroxide solution and washed with warm (50°C) water.HPLC was used for analyses. The highest conversion was 87.3 % and 6.0 % fattyacids were left in the product.

31

2.15. CHEMICAL PREPARATION OF RAPESEED OIL TRIMETHYLOLPROPANEESTER

RME (205.8 g, 0.7 mol) was weighed into a 500 cm3 three neck flask equipped witha thermometer, condenser, stirrer and sample adapter. Methylester was heated to60°C and TMP (25 g, 0.2 mol) was added and efficient stirred. When TMP wasmelted and well-stirred, 0.5 % (wlw) catalyst was added and mixed. Reaction wascontinued under reduced pressure (3.3 MPa) and the mixture was heated up torefluxing. Samples were taken hourly with a total reaction time of 8 hours. Aftercooling, the mixture was neutralized with alkaline or acidic water, washed with warm(50°C) water, and dried over anhydrous sodium sulfate. Analyses were carried outby TLC during the course of the reaction and by HPLC for the final products.

2.16. ENZYMATIC PREPARATION OF RAPESEED OIL TRIMETHYLOLPROPANEESTER

Reaction conditions were TMP 0.6 g (4.5mmol), 10 % (wlw) added water, RME 4.0 g(13.6 mol) and 40 % (wlw) lipase. Distilled water was added to TMP and once TMPwas in solution, RME and lipase were added. The capped test tubes were kept in awater bath (37°C) and stirred magnetically 250 rev min-1

. The supernatant waspipetted into Eppendorf tubes and stored at -20°C for later analyses.

32

3.RESULTS AND DISCUSSION

3.1. CHEMICAL SYNTHESIS OF 2-ETHYL-1-HEXYLESTER OF RAPESEEDOIL

3.1.1 Choice of catalyst

The effect of different catalysts on the degree of conversion in the chemicalsynthesis of 2-ethyl-1-hexyl ester of rapeseed oil was studied at a temperature 182­185°C. The catalysts tested for the esterification reaction were selected on the basisof previous experience (11,12,15,28,63). They were alkaline sodium hydroxide,potassium hydroxide, sodium methoxide and sodium ethoxide or acidic sUlphuricacid and phosphoric acid. Reaction conditions were as described in 2.9. (page 29).A sample was taken for HPLC analysis. The reaction proceeded slowly andconversions were quite low. Best results are shown in Table 3. The highestconversion, 68 %, was obtained with sodium methoxide. Sodium hydroxide was thenext best catalyst, which is an advantage for industrial production, because sodiumhydroxide is quite a cheap catalyst. With acidic catalyst only a 30 % conversion wasachieved. According to other literature and patent search, this type of reaction hadnot been done before and these results can only be compared to shorter alcoholreactions. There the best catalyst is sodium hydroxide (11,13,14) and conversionsup to 100 % have been reached at temperatures near the boiling point of thealcohol. The reaction scheme is shown in Figure 12.

Table 3: Different catalysts in rapeseed oil and 2-ethyl-1-hexanol reaction, (0,5 %(wlw) catalyst, 195-200°C, 3h).

Catalyst

Sodium hydroxidePotassium hydroxide

Sodium methoxideSodium ethoxideSulphuric acid

Conversion(%)

5854683030

33

Figure 12: Chemical reaction between rapeseed oil and 2-ethyl-1-hexanol reaction.

CH2-0-8-R6H -o-8-RbH2-0-8-Rrapeseed oil 2-ethyl-1-hexanol

2-ethyl-1-hexylester ofrapeseed oil

9H2-0H+ yH- OH

CH2-OH

glycerol

3.1.2. Effect of molar ratio

The stoichiometry of this reaction requires a 1:3 molar ratio for rapeseed oil: 2-ethyl­1-hexanol (Figure 12). Molar ratio was varied in following reactions from 1:3 to 1:6,being 1:3.0, 1: 3.5, 1:4.0, 1:4.5, 1:5.0, 1:5.5, 1:6.0. Nitrogen atmosphere was used toprevent oxidation of fatty acids due to the high reaction temperature of 190-200°C.Reaction was carried out as described in 2.9.. The progress of the reaction wasfollowed by TLC analyses. After one or one and half hours no rapeseed oil wasdetected in TLC. Reaction mixture was cloudy in all cases. In table 4 the effect ofmolar ratio on the total conversion is demonstrated. In all cases fatty acids could bedetected by HPLC analysis, the amount depended only a little or not at all on themolar ratio. The main reason for the formation of fatty acids was the high reactiontemperature. Fatty acids in final products may cause problems with corrosion andoxidation. Best molar ratio found was 1:5.0, which was used in subsequentexperiments.

34

Table 4: The effect of molar ratio (rapeseed oil: 2-ethyl-1-hexanol) on the conversionof their ester, (0.5 % (wlw) catalyst, 4 h, nitrogen atmosphere).

Molar ratio Conversion(%)

ester RO FA

70.0 22.074.0 18.081.6 9.887.0 3.788.2 7.385.5 3.582.0 9.3

Gly

RO= unreacted rapeseed oilFA= fatty acids produced in reactionGly= glycerol left in product mixtureester= the ester percentage in the product calculated from rapeseed oil (from

HPLC-gram, method 1, page 23 )

3.1.3 Effect of temperature and pressure

Due to the high temperature and in spite of the nitrogen atmosphere, rapeseed oiloxidized during the process and free fatty acids were found in the reaction mixture.This reduced the degree of conversion to the desired ester, and the presence of freefatty acids lowered the quality of the final product.To avoid this, reactions were made under reduced pressure. The molar ratio ofrapeseed oil: 2-ethyl-1-hexanol was in this case 1:5.0 and 0.5 % (wlw) of alkalinecatalyst. Reaction was carried out at 95-120°C/10.6 MPa, 90-115°C/8.0 MPa, 85­110°C/5.3 MPa, 80-105 °C/2.7 MPa, 80-100°C/2.0 MPa. The progress of the reactionwas followed by TLC. Total reaction time was four hours.Table 5 demonstrates the total conversions versus temperature and pressure.

35

Table 5: Effect of temperature and pressure on the conversion of 2-ethyl-1­hexylester of rapeseed oil,(molar ratio 1: 5.0, 0.5 % (wlw) catalyst, 4 h).

Temperature/Pressure(OC/MPa)

95-120/10.690-115/8.085-110/5.380-105/2.780-100/2.0

Conversion(%)

88.489.692.097.693.2

Conversion was much higher (90%-99 %) at reduced pressure than at normalpressure and no oxidation of rapeseed oil was discovered. The highest conversion of97.6 % was obtained when temperature was 80-105°C and pressure was 2.7 MPa.2-Ethyl-1-hexYlester of rapeseed oil was subsequently produced using a molar ratio1:5.0 of rapeseed oil:alcohol, 0.5 % (wlw) alkaline catalyst,at a reaction temperatureof ao°c to 105°C at 2.7 MPa reduced pressure. Figure 13 shows the HPLC-gramfrom a small pilot scale procedure of 2-ethyl-1-hexyl ester of rapeseed oil.Conversion was 97.6%. Figure 14 the IR-speetrum and its interpretation. It can beconfirmed from the IR-speetrum, that the product synthetized was 2-ethyl-1-hexylester of rapeseed oil.

36

Figure 13: HPLC-gram of rapeseed oil and 2-ethyl-1-hexanol reaction,(molar ratio1:5.0,0.5% (wlw) catalyst, 90-105°C, 2.7 MPa, 4 h).

CSl...,<1"~

'"CIl

"-'"'"""-lfl0& ...

01(S> ...

<N

1IlU.U.<:)

M

...G:I-

~

(S> c-J

""!<"I'".... V1, ZV'!~ w

'"c.>

I I I I I I I I I I I I I " I I I I I I I I I I I I I I II I I I I I I I I• ~ ~ ~ $ ~ ~ ~~ ~ ~ N ~ ~

<.J

37

Figure 14: IR-spectrum of 2-ethyl-1-hexyl ester of rapeseed oil and its interpretation.

Interpretation (em -1): 3470, broad, water in trace; 3020, (m), C-H-streehing, C=C-Hin R-group; 2940&2870, (5), C-H-streehing,-CH

2,-CH

3; 1750, (5), C-O-streching, R-C-O-; 1390, (m), C-H-symmetrical

deformation,-CH3

; 1250&1185, (5), C-O-streehing, R-C=O; 1125, (m), C-Q-streching,C-O-CH

2; 785, (w-m), C-H-out of plane deformation, HC=CH in R-group; 740, (m),

CH2-roeking, -CH2•

38

3.2. ENZYMATIC SYNTHESIS OF RAPESEED OIL 2-ETHYL-1-HEXYLESTER (I)

3.2.1.Choice of lipase (I)

Preliminary esterification tests with rapeseed oil and 2-ethyl-1-hexanol were carriedout in capped test tubes and the results analysed by TLC. Lipases produced byCandida rugosa (formerly cylindracea) , MJcor miehei, Pseudomonas Duorescens andChromobacterium v.iscosum were used as they were considered to be the mostsuitable for the esterification of fatty acids (64). Reaction conditions were asdescribed in chapter 2.10. (page 29). According to TLC, lipase from c.rugosa,produced most of the desired ester; all rapeseed oil had reacted and only a smallamount of alcohol was left. P. Duorescens and Ch. viscoseum also lipases catalyzedthe ester synthesis well, but more unreacted alcohol and by-products were found.Poorest results were obtained with M miehei lipase: little ester was formed and a lotof unreacted rapeseed oil and alcohol remained. Samples after 24 and 48 hourswere also analysed by HPLC and the results are presented in Table 6. Table 7shows the characteristics and prices of the different Iipases.The best lipase was from C. rugosa which also turned out to be the least expensive.

Table 6: Different lipases in rapeseed oil and 2-ethyl-1-hexanol reaction, (molar ratio1:3,3.5 % (wlw) lipase, 3.0 % (wlw) added water, 12-72 h).

C.rugosaMMieheiP.BuorescensCh. Yiscosum

Lipase24 h

98459696

Reaction time48 h

Conversion(%)

98879997

39

Table 7: Properties and prices of different Iipases used in reaction 3.2.1 ..

Lipase

C.rugosaCr.viscosumM.mieheiPS.f1uorescens

Activity(U/g)

4250013300720011900

Water(%wlw)

5,05,97,43,1

PriceUSDlkg USD/10U

292 72300 1722250 3132100 177

3.2.2 Effect of substrate molar ratio (I)

The theoretical molar ratio in rapeseed oil 2-ethyl-1- hexanol ester synthesis is 1:3rapeseed oil:2-ethyl-1-hexanol. Other reaction conditions were as in (2.10., page 29).Ester synthesis was investigated using the following molar ratios of rapeseed oil and2-ethyl-1-hexanol: 1:1.0, 1:1.5,1:2.5,1:2.6,1;2.7,1:2.8,1:2.9,1:3.0,1:4.0,1:6.0and 1:10.0. On the basis of both TLC and HPLC the best results were obtained withlittle or no excess alcohol. All rapeseed oil had reacted with a 98 % conversion tothe desired product, with no residual 2- ethyl-1-hexanol and no by-products. Therelative ester yield decreased when an excess amount of alcohol was used.Apparently, excess alcohol inhibited the lipase and the use of excess of alcohol wasnot further studied. Figure 2/1 demonstrates the effect of the molar ratio on theconversion of 2-ethyl-1-hexyl ester of rapeseed oil. Molar ratio 1:2.8 seemed to yielda little higher conversion in a shorter time than molar ratio 1:2.9 or 1:3.0 (in 1 h 58 %compared to 50 % and 40 %, respectively, or in 5 h 96 % compared to 95 % and 85%, respectively).

3.2.3 Effect of lipase quantity (I)

Figures 15 and 16 demonstrate the effect of lipase quantity on the conversion ofrapeseed oil to the desired ester. With a lipase concentration of between 0.3 to 1.7% (wlw) transesterification was slow after one hour. By increasing the enzymequantity to 2.3-6.4 % (wlw), conversion was increased to 60 % in one hour, andwith a lipase quantity of 14.6 % (wlw) to almost 100 %. However with 0.3 % (wlw)lipase, the conversion reached nearly 90 % in 7 hours and with 1.0-1.7 % lipase, theconversion was over 80 % already in five hours. Bearing in mind the cost of theenzyme, it is important to observe that almost a 100 % conversion could beobtained with less enzyme at an extended reaction time.

40

lipase quantity of 14.6 % (wlw) to almost 100 %. However with 0.3 % (wlw) lipase,the conversion reached nearly 90 % in 7 hours and with 1.0-1.7 % lipase, theconversion was over 80 % already in five hours. Bearing in mind the cost of theenzyme it is of importance to observe that almost a 100 % conversion could beobtained with less enzyme at an extended reaction time.

Figure 15: Effect of lipase quantity on the conversion of 2- ethyl-1-hexyl ester ofrapeseed oil, (molar ratio 1: 2.8, added water 3.0 % (wlw), 37°C, lipase0.3 % (wlw) (.),1.0 % (wlw) (0),1.7 % (wlw) (+) and 2.3 % (wlw)

(~».

100 T

90 t80 +

I

~ 70 t~~ 60 +g I.~ 50 + I

~ 4O+~/u I30t,20 + /

I10 T ",#

oo 5

time (h)

11

41

Figure 16: Effect of lipase quantity on the conversion of 2- ethyl-1-hexyl ester ofrapeseed oil, (molar ratio 1:2.8, added water 3.0 % (wlw), 37°C, lipasequantity 3.3 % (wlw) (0),4.9 % (wlw) (.), 6.4 % (wlw) (4> ) 14.6 %(wlw) (Y».

••

4

time (h)

•100 TI

90 t80 +70 1

~ I!L. 60 +g I'0 50 I

Qj T I /,i!: 40 + ,I §o : I I>u 30 + .,1 j

20 +!i10 tilo ".'----+---~---...----_+__---_--_j

1

3.2.4. Effect of added water (I)

The effect of added water on the transesterification reaction is demonstrated inFigure 5/1. When 0.5 % (wlw) added water was used, there were no separablephases in the reaction system. When water content was between 10.0 to 50.0 %(wlw), a distinct water phase was formed, into which the lipase and glycerolproduced, were dissolved. Only the organic phase was used for the ester analysis.The water content of lipase alone (about 5.0 % (wlw» was not sufficient for the estersynthesis. Conversion after seven hours was only 24 % without added water and anincrease in water concentration up to 0.5 % (wlw) did not improve the conversion.With 3.0 % (wlw) added water a complete conversion was reached in five hours.Additional water increases did not further improve conversion.

42

3.2.5. Effect of temperature (I)

The effect of temperature is demonstrated in Figure 17. There were no significantdifferences between 37°C to 55°C in seven hours: The conversion was almostcomplete. In 55°C, the conversion was 90 % already after two hours. The sameconversion was reached at 37°C in three hours. The temperature of 60°C inhibitedlipase and the conversion was only around 60 %. The results support theobservations of Mittelbach (11) and Hirata (65). Mittelbach (11) found that thetemperature limit for Candida sp. lipase catalysed sunflower oil transesterification is45 to 50°C. Hirata et al (65) reported 50°C to be optimal for transesterification oftributyrin and 1- octanol with c.cylindracea lipase. For economic reasons thetemperature 37°C was chosen for the future industrial application of 2-ethyl-1­hexylester of rapeseed oil.

Figure 17: Conversion of 2-ethyl-1-hexyl ester of rapeseed oil versus temperature,(molar ratio 1:2.8, lipase 3.3 % (wlw) , added water 3.0 % (wlw),7 h).

11

100 T

90 + b80t r-I--~

~ 70 + //~, IoU/~~ 60 + _------.0: /

;;SOt ///

~ 40+ j)' /. //8 I / /

30 +.,./ /420 t "10.i-

o .'--~--~--;---------+-----;o S 7.

time (h)

1-37C!I ----0- 4SC I!--- SO"CI '

1-0- SS'C i1--- 6O"C iI :

43

3.3. SMALL PILOT SCALE OF ENZYMATIC PREPARATION OF 2-ETHYL-1-HEXYLESTER OF RAPESEED OIL (II)

Figure 1/11 demonstrates the effects of different mixing rates on rapeseed oilconversion. All experiments were performed with a substrate ratio 1:2.8, 1.0 % (wlw)added water and 3.4 % (wlw) lipase. Magnetic stirrer (300 rev min-1

) was used in flatbottom flasks. Conversion increased to 65 % in a few minutes and was up to 80 %in one and half hours. After 30 minutes the solid lipase had collected to the wallsand the bottom of the flask. This seemed to be beneficial for the esterificationreaction. The solid lipase preparation also absorbed the glycerol produced in thereaction. After five hours the conversion was 87 %.In another experiment the speedof the magnetic stirrer was 700rev min-1

, in this case conversion was only 46 % forthe first five hours because the high speed dispersed the lipase to reaction mixtureand resulted in high shear forces. Lowering the stirring speed to 300 rev min-1

, didnot increase the conversion to higher than 61 %. Lee and Choo (66) have alsonoticed that C. rugosa is easily denatured when the mixing speed is higher than 75­150 rev min-1

. Clearly the activity of lipase decreases as a function of mixing rateand time. This is in accordance with the results of Goldberg et al (67) on C. rugosalipase catalysed heptyl octanoate synthesis. When mixing rate was increasedsufficiently, for complete suspending of lipase, the enzyme became inactivated.A few experiments were also carried out in which the lipase was reused afterdecanting from the reaction mixture. Conversion then varied from 3 % to 13 %.Figure 18 shows an example of such experiments. Because preliminary experiencehad suggested that the immobilization of the lipase might increase conversion,different carriers were also tested.

Figure 18: Conversion of 2-ethyl-1-hexyl ester of rapeseed oil (molar ratio 1:2.8,3.4% (wlw) lipase, added water 1.0 % (wlw) , 37°C, motor stirrer 300 rev ( •f!lin-1 lipase reused (0), magnetic stirrer 300 rev min-1

( +)Iipase reused(~».

7

44

However, no improvement could be seen with glass beads, polyether foam orhydrophilic Amberlite XAD-2 resin immobilized lipase. Figure 4/11 shows how XAD-7­resin with different amounts of lipases affects to 2-ethyl-1-hexylester of rapeseed oilconversion. The hydrophobic Amberlite XAD-7 resin gave the best results as acarrier. After five hours the conversion was 95 % with 50 g of rapeseed oil, 2.5 glipase, 25 cm3 2-ethyl-1-hexanol and 3.0 % (wlw) added water.Under the optimised conditions a 2 kg batch of rapeseed oil was then esterified, with1 dm3

, (829 g) 2-ethyl-1-hexanol, 3 % (wlw) added water, 100 g lipase with 300 gXAD-7 resin, speed of rotation 170 rev min-\ 37'C. The obtained conversion about90 % was nearly theoretical. Total reaction time increased, but the conversion wasabout 20 % higher than in earlier experiments with different carriers.

3.4. CHEMICAL SYNTHESIS OF RAPESEED OIL METHYL ESTER

Methylester of rapeseed oil (RME) has been produced by Raisio Group, Oil MillingDivision, since summer 1992, the total production of RME being 150-200 tons.Production in Europe is estimated at 900 000 tons per annum and the total worldproduction is 1 million tons per year (68).The preparation of RME on a laboratory scale was done as described in chapter2.11. (page 8). The reaction scheme is shown in Figure 20. As can be seen from theHPLC-gram (Figure 19), the total conversion was 97 %, the rest being unreactedrapeseed oil. This matches other laboratory results well (11,12,13,14,15). Table 8shows the most important biodiesel fuel demands according to c>NORM 1190/1991and the standards, that they are made with. It is important to notice that the cetanenumber is about 50, being clearly the best in comparison with alternative fuels (69).Cold stability properties for winter season can be improved by additives. Pourpointcan be decreased from -15 to -36°C. Thermal value is also at the same level ascommon diesel oil (69), but the distilling figure is totally different. Summer dieseldistills at 180-360°C while rapeseed oil methylester at 240-270°C (70).

Figure 19: HPLC-gram of rapeseed oil methyl ester (method 1, page 27).

N

"" ,.) ":N

,~

N M

'"u..u..

\

0 N

'",..., "'In

f-,...

f-

'" "" <S>

'" r

'" ,~

, r- u.."' , .... ("'01

"'~..... l.·~~ _..-. '0'

l~~~.,

z rl,•.'l .... .u

'"

f tIl' 1111111111111111' II fIlII 1111I 111I1 II 'II't? ~ fI) IS' U"J CSJ If) ($)

.... "'" ... ~ M ""') ..po

45

Figure 20: Reaction scheme of methyl ester synthesis.

CH2-0-8-R

6H- 0-8-R

6H2-0-gR

CH2-OHI

+CH-OHICH2-OH

rapeseed oil methanol methylesterof rapeseed oil

glycerol

46

Table 8: Specifications of Raisios RME for fuel use as modified from Biodiesel­project report (71).

Property unit limits Standard

Density g/cm3 0.87-0.90 DIN 51757Viscosity40°C mm2/s 4.5-5.0 DIN 5156220°C 8.5-9.0 DIN 51562O°C 15.5-16.0 ASTM 0445

-20°C 39.5-40.0 ASTM 0445Surface tension dyn/cm25°C 30.5Flashpoint COC °C 190 ASTM 092Pourpoint °C -15 ASTM 097Winter °C -36 ASTM 097Cetane number min 49Thermal value MJ/kg 37.2Carbon residue %M <0.02 ISO 10370(Conradson)Oistilling/350°C % >90Corrosion Cu 1a ASTM 0130Sulphur content mglkg <5 ICPPhosphor content mg/kg <5 ICPElement analysis %MCarbon 75.0Hydrogen 12.6Oxygen 11.1Nitrogen <0.1Methanol content % <0.03 OIN 51413,1Water content mglkg <100 DIN 51777,1TAN mgKOH/g <0.40 DIN 53402Colour <G 10 ASTM 01500Biodegradability % >90 CEC-L-33-T-82

47

3.5. CHEMICAL SYNTHESIS OF SOYBEAN OIL ETHYLESTER

Soybean oil ethyl ester was prepared as in chapter 2.12. (page 25). The reactionscheme is similar to that for rapeseed oil ester (Figure 20, page 39). As can be seenfrom Figure 21, conversion was typically around 90 % and the only other peak wasunreacted soybean oil.

Figure 21: HPLC-gram of the ethyl ester of soybean oil.

..co

<$>

"11'\ --.., .. '"'<" ~,... .. '"... '" =.....

I Z., ...... uJ ...

'"1'1 1IIIIflllll,IIIIIIII II 1

'11111111"'1111

n ~ n ~ ~. ~ ~ s~ ~ ~ ~ ~ ~ ~

3.6. CHEMICAL SYNTHESIS OF COCONUT OIL METHYLESTER

Coconut oil methylester was prepared as in chapter 2.13.(page 30). Also in this casereaction scheme is similar to that for rapeseed oil methylester (Figure 20, page 46).As can be seen from Figure 22, a two hour reaction time was not suffient. Theconversion after two hours was only 72 % and almost total conversion of 92 % wasachieved after 5 hours. It was more difficult to esterify coconut oil than rapeseed oildue to its fatty acid content. According to literature the most frequently used oils arerapeseed oil, soybean oil and sunflower oil (15).

48

Figure 22: A thin layer chromatogram from coconut oil and methanol reaction, (molar ­ratio 1: 6.0, 0.6 % catalyst (wlw) , 80°C, 5 h).

CO=coconut oil1=1h reaction2=2h reaction3=3h reaction

t t f t i • t4=4h reaction

t t~ CO+2=coconutf ,• oil+ 2h reaction

• • • • • 5=5h reactionco 3 CO CO+2 5

3.7. CHEMICAL SYNTHESIS OF TRIMETHYLOLPROPANE ESTER FROM FATTYACIDS

3.7.1. Choice of catalyst

The synthesis of trimethylolpropane fatty acid esters was carried out using threedifferent catalysts sulfuric acid, phoshoric acid and p-toluenesulfonic acid. Thesecatalysts have been used previously in similar esterification reactions (10,31,35).Reaction stoichiometry requires one TMP mole for three moles of fatty acids (Figure23).

49

Figure 23: The reaction scheme of the esterification of TMP with fatty acids of tallowoil.

CH2-OHq I

3 R-C-OH + CH3-CH2-C-CH2-OHICH2-OH

CH2-0-8-RA 1 9cata.!lst CH3-CH2-C·CH2-O-C-R + 3 Hp

6H2-0-8-Rfatty acids TMP TMP- ester

To ensure that the reaction would progress towards the right direction, an excess offatty acids was used. Tallow oil fatty acids were distilled so that only certain types offatty acids were included in the starting material. Fatty acid concentration isdescribed in chapter 2.1., page 25 and reaction conditions are shown in chapter2.14.A., page 30.According to qualitative TLC best results were achieved with p- toluenesulphonicacid, yielding most of the desired ester. Almost all of the alcohol had reacted andonly a small amount of fatty acids were left. Sulphuric acid and phosphoric acid ascatalysts also aided the production of the desired ester, but there were moreunreacted fatty acids and more byproducts. With phosphoric acid a much longerreaction time was needed or the reaction temperature had to be increased, whichresulted in the oxidation of fatty acids.

3.7.2. Effect of catalyst quantity on esterification of tallow oil fatty acids

Figure 24 demonstrates the effect of different amounts of catalysts on the conversionof fatty acids into esters. The amount of catalyst was varied from 0.1 % (wlw) to 2.0% (wlw). Best results with a conversion of about 65 % as determined by HPLC wereachieved by using 0.3-0.5 % (wlw) catalyst.

50

Figure 24: The effect of the amount of catalyst on the conversion of fatty acids toTMP-ester, (molar ratio 1:3, 120°C,7 h.)

70 T

60 .,.

20 t

"10 J.

1.20.5 0.7 0.9

amount of catalyst (%)

0.3

o~,-------------------0.1

3.7.3. Effect of substrate molar ratio on esterification of tallow oil fatty acids

The theoretical molar ratio for the TMP fatty acid ester synthesis is 1:3 TMP:fattyacids. The reaction conditions were as desribed earlier (2.14.), but the molar ratiowas varied from 1:3 to 1:6, as follows: 1:3.0 1:3.5, 1:4.0, 1:4.5, 1:5.0, 1:5.5, 1:6.0.Analyses were carried out by TLC and HPLC. The conversion could also becalculated approximately from the water quantity formed in the reaction. The bestresult with almost all of the TMP reacted, a 65 % conversion to the desired productand a little residual fatty acids, was achieved by using little or no fatty acid excess.This can also be seen in Figure 25, which shows a thin layer chromatogram of thereactions with different molar ratios. In further studies (next chapters) stoichiometricquantities of fatty acids were used.

51

Figure 25: TLC-pictures of different molar ratios in reaction of TMP with fatty acidsof tallow oil, (0.5 % (wlw) catalyst, 120°C, 7 h).

·:.:-:ir~;:· :.•':'.•: ~ ,~ l'< ~~. JAn

1 2 .3 .. 5

l=A 2= l:J 3~ 1:3.5 4= 1:4 5=1:4.5 6=1:5 7=1:5.5 8=1:6

3.7.4. Effect of temperature, time and pressure on esterification of tallow oil fattyacids

Reaction time varied from 2 to 11 hours, reaction temperature from 120 to 180°Cand pressure from 18.6 MPa to 11.6 MPa. The conversion varied from 65 % to 100% as calculated on the basis of water formed. The progress of the reaction wasfollowed by TLC. Highest conversion was obtained in 7 hours at 120 to 130°C.Because some unreacted, and also some oxidised fatty acids were detected in thereaction mixture, the following changes were made: The reaction temperature wasadjusted to 70 to 80°C and the pressure was reduced to 11.9 to 18.6 MPa. Table 9presents data of the different reactions.

52

Table 9: The effect of temperature and pressure on the conversion of fatty acidsTMP-ester controlled by the produced water quantity, (molar ratio 1:3.0,0.5% (w/w)p-toluenesulphonic acid, 2-9h)

Tempe­rature

(oC)

Pressure

(MPa)

Time

(h)

Waterquant.

(%)

Conver­sion(%)

120 3.0 65130 5.5 75140 4.0 85130-140 3.5 65150 2.0 90150-160 2.0 100160 3.5 100160 5.5 100170 3.0 90180 3.5 9070-80 18.6 3.0 68

17.2 4.0 7417.2 6.0 7615.9 6.0 8213.2 4.0 8513.2 6.0 8611.9 6.0 88

--------------------------------------------------------------------------------------------

The optimal reaction conditions were then employed in a larger scale experiment asdescribed in chapter 2.14., B.(page 30). The conversion was at best 87.3 %, with 6.0% fatty acids still left in the product. In Figure 26 a comparison is made betweentallow oil trimethylolpropane ester and a comparable commercial TMP-ester. Thecommercial TMP-ester is prepared from tallow oil fatty acids, which have beendistilled and purified to single fatty acids before used in esterification with TMP (4).This can clearly be seen in the HPLC-gram as a narrow line of peaks. In all of thereactions, unreacted starting materials, if only in a small amount, were always left inthe final product. Considerable difficulties were encountered in the neutralising,washing

53

and purification steps of the ester-product owing to soaping and foaming. Because ofthese problems, the transesterification of small alcanol esters of vegetable oils wasinvestigated in place of the esterification.

Figure 26: HPLC-gram (method 2) of tallow oil TMP-ester (above) and comparablecommercial TMP-ester (below).

E- "~

~

! l

!~~

!~J.; ~

I

11 -

L~

~ ~

..

f~~~i e'

LiL

~~Ll.Ir~f:

~LI !..

54

3.8 CHEMICAL PREPARATION OF TMP ESTER FROM RAPESEED OILMETHYLESTER AND TRIMETHYLOLPROPANE (III)

3.8.1. Choice and quantity of catalyst (III)

The most frequently used catalysts in this type of transesterification reactions aresodium and potassium hydroxides and alcoholates or p-toluenesulphonic acid(72,73,74).The quantity of catalyst has been generally varied from 0.2 % (wlw) to 2.0 % (wlw).In the present work preliminary trials were done under following reaction conditions:substrate ratio TMP:RME 1:3.5 and catalyst 0.5 % (wlw). Reduced pressure wasalready known to be essential (3.1.3 and 3.7.4). Consequently, the reactiontemperature was 80 to 120°C and pressure 3.3 MPa. The reaction temperature andpressure were chosen so that TMP would properly melt and the reaction mixturecould be refluxed. The molecular scheme is shown in Figure 27. The best catalystunder these reaction conditions seemed to be sodium methylate, although alsosodium hydroxide and potassium methylate gave satisfactory results.

Figure 27: Molecular schedule of rapeseed oil methylester and trimethylolpropane.

RME TMPof rapeseed oil

TMP-ester methanol

The amount of sodium methylate was then varied from 0.1 % (wlw) to 2.0 % (wlw) ,being 0.1 %,0.3 %, 0.5 %, 0.7 %, 0.9 %,1.2 %,1.4 %,1.6 %,1.8 % and 2.0 %(w/w). Other reaction conditions were as given in chapter 2.15., page 31. The resultsare shown in Table 10. The best results were achieved by using 0.7 % (wlw) sodiummethylate.

55

Table 10: The effect of the catalyst amount on the conversion of rapeseed oil methylester to TMP-ester, (molar ratio 1:3.5, 60-130°C , 3.3 MPa, 8 h).

------------------------------------------------------------------------------------------------------

Catalyst(% (wlw»)

0.10.30.50.7

0.91.21.41.61.82.0

Conversion(%)

70.676.680.285.276.574.472.672.370.369.8

---------------------------------------------------------------------------------------------------------

3.8.2. Effect of substrate molar ratio on RME transesterification (III)

The stoichiometry of the transesterification reaction requires per one mole TMPthree moles RME (Figure 31). To ensure that the reaction proceeds towards the rightdirection, a small excess of RME was needed. Molar ratios studied varied from 1:3.0to 1:4.0, being 1:3.0, 1:3.1, 1:3.2, 1:3.3, 1:3.4, 1:3.5, 1:3.6, 1:3.8, and 1:4.0.Other reaction conditions were as described previously (2.15., page 10). Thedifferences in conversions were small, as can be seen from Table 11. Neverthelessa slight excess of RME gave the highest conversion of 86 %, and was later used inthis study.

56

Table 11: The effect of molar ratio (TMP:RME) on the conversion to the desiredester, (0.7 % (w/w) catalyst, 80-120°C, 3.3 MPa, 8 hours).

Molar ratio(RME:TMP)

3.03.13.23.33.43.53.63.84.0

Conversion(%)

82.284.086.080.276.576.074.473.072.6

-------------------------------------------------------------------------

3.8.3 Effect of temperature and pressure on RME transesterification (III)

Both temperature and reduced pressure had a strong effect on the conversion to thedesired ester. The following conditions were investigated using a slight excess ofRME, TMP:RME (1 :3.2), 0.7 % (wlw) sodium methylate catalyst, other conditionsbeing as described in chapter 2.15. (page 26). Preliminary experiments were donemaintaining reduced pressure at 3.3 MPa. Temperature varied from 70 to 130°C, atlevels 70-90°C, 80-11 O°C, 100°C, 100-130°C and 130°C.

57

Table 12: Effect of the reaction temperature on the conversion of TMP-ester, (3.3MPa reduced pressure, TMP:RME 1:3.2, 0.7 % catalyst, 8 hours).

Temperature(0C)

Conversion(%)

RME, unreacted(%)

---------_._--_._--------

70-9080-110100100-130

86.090.589.087.3

14.09.5

11.012.7

/._-------.-->--_.

As can be seen from Table 12, the conversion was higher when the temperaturevaried during the reaction procedure. It could be seen clearly from TLC- and HPLC­analyses, that the fatty acids of RME reacted with TMP in two steps: First theshorter, straight chain, less substituted fatty acids and then the bigger molecules.The reaction profile resembled that shown in Figure 28. Therefore following reactionswere carried out at two different temperatures or temperature ranges. Reaction timein each case was decided on the basis of TLC runs dUring the reaction procedure.Results are shown in Figure 29.

Figure 28: Typical profile of RME and TMP reaction. (Molar ratio TMP:RME 1:3.2,0.7 % ( wlw) sodium methylate, 3.3 MPa).

120 T

100 'U ~

"or 80 .....------··

~ 60 ~Q;~ 40 +Q) ,

- 20 +o l....:----------------~:__:.

10otime (h)

100;

58

Figure 29A: The effect of the time on the conversion of TMP and RME ester, (molarratio TMP: RME 1: 3.2, at lower temperature area, mean value 90°C,3.3 MPa).

- T35 -;-..___________\ _~~~ ( 30 e-

Rao ~ ~ i25~- /0" ~_ ~'." i20~

.~ 60 .../' t 15 ~(;; 40 \~. +10'".,~ I8 20 r 5 -

L ~-_:_-_;;;-' 0

o i 2.5 3,5 4 4.5 5.5

time (h)

Figure 298: The effect of the time on the conversion of TMP and RME ester, (molarratio TMP:RME 1:3.2, at higher temperature area, mean value 110°C,3.3 MPa).

- a, 35

_

" ..~,-;--__::-=__--;;-_----i~ f 30 ~~ 801 ~ f251J.Jg 60! 120~~ I 15 ~~ 40 \ ~! 10-5g 20r ~ 15 .~

oLi ~===L__:.;0 -

2.5 3.5

time (h)

5.5

Conversion was at lowest with 66.5 % of ester, 33.5 % of residual RME, whentemperature was 105-120°C for 2.5 hours. It improved a little, when two temperatureareas were chosen, 80-100°C, 2.0 hours and 100-130°C, 2.5 hours, being 83 %. Itwas noticed that the progress of the reaction did not only require two reaction

59

temperature areas, but also longer reaction times for better conversion. The best. conditions observed for the temperature and pressure were: first 85-110°C for 2.5

hours and then 110-120°C for 8 hours in 3.3 MPa. Conversion was then 98.9 % ofthe TMP-ester.

3.9. SMALL PILOT SCALE PRODUCTION OF TMP-ESTER FROM RAPESEED OILMETHYLESTER AND TMP (III)

All reaction conditions that had earlier been optimal for the transesterification toTMP-ester (3.8.1. - 3.8.3.) were combined and a pilot scale production of TMP-esterwas carried out. Five kilos of RME (17.1 mol) was weighed to a reaction vesselequipped with a thermometer, a condenser and a stirrer. The methylester washeated to 60°C and TMP 716 g (5.3 mol) was added in five portions with adequatemixing. When all TMP had melted and the mixture was well-stirred, 40.0 g (0.7 %(wlw))sodium methylate was added and properly mixed. Pressure was reduced to3.3 MPa. Temperature was first raised to 85°C and then gradually elevated to 110°Cduring a period of two hours under refluxing. Thereafter the reaction temperaturewas kept between 110 and 120°C for 8.0 hours. Samples were taken every hour andanalysed with TLC. No RME could be seen on the TLC-gram after 10.5 hours, afterwhich the reaction mixture was neutralized by hydrogen chloride water (1:1) andsubsequently washed three times with warm (50°C) water followed by drying overanhydrous sodium sulfate. The final product was analysed by both IR and HPLCwith two different columns.Conversion to the TMP-ester of rapeseed oil was 99.0 % and no methylester ofrapeseed oil could be found in the final product. Figure 30 shows the HPLC-grams ofproduct, and Figure 31 the IR- spectrum and its interpretation. HPLC-gram of RMEand TMP ester is comparable to HPLC-gram of commercial TMP-ester. It can beseen, that the commercial TMP-ester contained fewer fatty acids. The main peaks,however, were in the same area. The molecular structure of the product combinationwas confirmed by IR-spectrum of rapeseed oil TMP-ester.The TMP-ester ofrapeseed oil thus obtained was used for producing and testing hydraulic fluids.

60

Figure 30: HPLC-gram of rapeseed oil TMP-ester (left) and comparable commercialTMP-ester (right) with method 2 (page 27).

.~!

;.1.;:~+~ ..

::i :

61

Figure 31: IR-spectrum of rapeseed oil TMP-ester (obtained)and its interpretation.

Interpretation (em-1): 3020, (m), C-H-streehing, H-C=C-H in R group; 2950&2860, (s),

C-H-streehing, -CH2,-CH3; 1750, (s), C-O-streehing, R-C-O-; 1470, (s), C-H­deformation, -CH2,-CH3 ; 1385&1320, (m),-H2C-C-CH2-; 1250&1170, (s), C-O­streehing, R-C=O; 1020, (m), C-O-streehing, R-C-O-C-; 785, (w-m), C-H-out of planedeformation; 740, (s), -CH-roeking, -CH2 .

62

3.10. ENZYMATIC PREPARATION OF RAPESEED OIL TMP-ESTER (IV)

3.10.1 Choice of lipase (IV)

The preliminary tests were carried out in capped test tubes and analysed by TLC.Lipases used were produced by Candida rugosa (ex. cylindracea) , JlJucor miehei,JlJucor sp. and Pseudomonas fluorescens. They were selected to be the mostsuitable for this esterification on the basis of prior experiments (3.2.1.).The Candida rugosa lipase appeared to produce most of the wanted ester accordingto qualitative TLC. With JlJucor miehei lipase hardly any TMP-esters were obtained ascan be seen in Figure 32. Because conversion in capped test tubes was small ornone, the balance of transesterification was tried to shift to the side of desiredproducts by elevating temperature to 58°C and with open testtubes. Conversion wasnot much improved so the next step was to carry out the reaction under reducedpressure. Candida rugosa was chosen to be the best lipase for further experiments.

Figure 32: Conversion of TMP-esters from rapeseed oil methylester by lipaseproduced by C. rugosa{Cr) and Mmiehei (Mm), (molar ratio 1:3, addedwater 15 % (wlw) , lipase 40 % (wlw) , 3rC).TMPE, Cr (_); RME, Cr (0); TMPE, Mm (+); RME, Mm (¢).

0.04 "\

0'035 T\L0.03.). \

i~+ \\ .)2) i . • : _---0--0--

a 0.02 + \b-- • -- • • • •.~ ! • ~--~-.--...a-----a

~ 0.015 + /8 . /0.01 + ! - _ _..--.__...... ---.~ -+

'! /0.005 + / /

o,;;~.~e:::::::-.-/-_------~-~-------o 12 24 36 48 60 72 84 96 108

time (h)

63

3.10.2 Effect of lipase quantity and its stepwise addition (IV)

The effect of the lipase quantity was tested under two different reaction conditions:(1) substrate molar ratio (1 :3), reaction temperature 3rC, reduced pressure 12.0MPa and reaction time 24 hours, and (2) 4rC, 5.3 MPa and 68 h. Substrate ratioswere the same as above. Lipase quantities tested were 10 %, 20 %, 40 % and 80 %(wlw). In the reaction (1) the highest conversion of about 80 % was obtained, whenlipase quantity was 40 %. At higher temperature and pressure some RME remainedand 20 % (wlw) lipase quantity, conversion was increased to 95 % in 68 hours.Figure 33 shows the results.

Figure 33: Conversion of rapeseed oil methyl ester with TMP to TMP-ester, withvarious quantities of lipase, (molar ratio 1:3, added water 15 % ( wlw)1. 3rC, 12.0 MPa, 24 h (.) and 2. 4rC, 5.3 MPa, 68 h (0».

100 ,

90

80 ..70

~ 60 ..C.~ 50 ..;;> 40 ~c0...

30 .,.20 ~

10 ..0

0 10 20

enzyme (0/0)

40 80

64

Enzyme was also added stepwise to the reaction mixture. Reaction conditions wereas in chapter 2.16. (page 26) with the total enzyme quantity of 40 % (wlw). Thereaction time was 46 hours with two and 68 hours with three enzymesupplementations. Figure 34 shows the conversion obtained with a number of lipaseadditions. The conversion was approximately the same with one or two additionsand slightly lower with three steps. So there seemed to be no benefit with stepwiseaddition of lipase.

Figure 34: Conversion of rapeseed oil methyl ester with TMP to TMP-ester, (molarratio 1:3, total lipase quantity 40 % added stepwise, added water 15 %(wlw) , 4rC,5.3 MPa, 46 (. ) or 68 (0) h).

III"enzyme additions

r~lII-'__...JlII'- ~ LI

3.10.3 Effect of substrate molar ratio on enzymatic transesterification of RME (IV)

In the first experiments the substrate molar ratio TMP:RME was always thetheoretical 1:3. In the following experiments the substrate molar ratio was variedfrom 1:1.8 to 1:9.0, actual ratios being of 1:1.8, 1:2.0, 1:2.2, 1:2.3, 1:2.6, 1:3.0, 1:3.5,1:4.5 and 1:9.0. Other reaction conditions were: temperature 3rC, reduced pressure5.3 MPa, 40 % (wlw) lipase and 15 % added water (wlw). With 12 hour reaction timeat substrate molar ratio 1:3.5 and 1:4.5 conversion was 95 % and 85 %, respective.In both cases some unreacted RME remained in the mixture. With substrate molarratios smaller than 1:3.5 or higher than 1:4.5 a lot of unreacted RME remained andmono- and di-substituted TMP-esters and by-products were found in the mixture. Notrisubstituted ester was obtained under these conditions. Consequently, the molarratios of 1:3.5 and 1:4.5 were chosen for further experiments. Figure 35 shows theeffect of the molar ratio on the total conversion.

65

Figure 35: Conversion of rapeseed oil methyl ester with TMP to TMP-ester, whenmolar ratio was varied, (lipase 40 % (wlw),added water 15 % (wlw),37°C, 5.3 MPa, 24 h).

94.53.52.3 2.6 3

relative quantity (TMP:RME)

2.22 .

l:H~ 70 i~-;; 60o.§ 50 t

I ;~ j'20

10

o.l---+---+----+--->---+-----+----+-----<

1.8

3.10.4. Effect of added water on enzymatic transesterification of RME (IV)

The amount of added water for the first reactions was chosen on the basis of theearlier esterification experiments (3.2.4.).The following reaction conditions were used: (1) substrate ratio 1:4.5, reactiontemperature 3rC, 12.0 MPa pressure and reaction time 24 hours, and (2) substrateratio 1:3.5, reaction temperature 47°C, pressure 5.3 MPa and reaction time 68 hours.Lipase quantity was 40 % (wlw) in both cases.The examined added water quantities were 8 %, 10 %,15 %, 30 %, 45 % and 60 %(wlw) of the total mixture. In the first reaction procedure the conversion improved,when the water content was increased from 8 % to 30 % (wlw). In the secondprocedure the conversion was about 90 % with 15 % (wlw) added water with a littleRME left and no by-products. The detailed results are given in Figure 36.

66

Figure 36: Effect of the added water quantity on the TMP-ester conversion;«1) molarratio 1:4.5, lipase 40 % (wlw) , 3rC, 12.0 MPa, 24h (.) (2) molar ratio1:3.5, lipase 40 % (wlw) , 4rC, 5.3 MPa, 68h, (0».

1~~I~ 80C 70

.~ ~~ t!II 40 tg 30 t(J 20 +

10o~---

10 15 30 50

added water (%)

3.10.5 Effect of temperature on enzymatic transesterification of RME (IV)

The original temperature of 37°C was chosen on the basis of the experience gainedfrom the preliminary experiments (3.2.5.). In the following experiments the pressurewas 5.3 MPa, reaction time 72 h, substrate ratio 1:4.5, lipase quantity 40 % (wlw)and added water 15 % (wlw). The temperatures used were 37,42,47, 52 and 58°C.In Figure 37 the results are shown. The total reaction time was 72 hours. Thehighest conversion of 97.5 % was obtained at 42°C in 72 hours.

67

Figure 37: The effect of temperature on the conversion of RME with TMP torapeseed oil TMP-ester, (molar ratio of 1:4.5, lipase 40 % (wlw), addedwater 15 % (wlw), and 5.3 MPa, in 72 h).37°C (.), 42°C (0), 47°C (+) and 52°C (0).

724824

time ChI

12

'./

80

70

60

~ 50

"~40~:>8 30

20

10

0r:------------ ~o

3.10.6 Effect of pressure on enzymatic transesterification of RME (IV)

As was observed already in the first experiments, the reaction needed reducedpressure (3.8.3., 3.10.1) to obtain the desired ester. The experiments under varyingpressure were done at 37°C, substrate ratio 1:4.5, 40 % (wlw) lipase, 15 % (wlw)added water, and reaction time 24 hours. The conversion was calculated only for themain product, trisubstituted ester. The pressures used in the test were ambient, 12MPa, 5.3 MPa and 2.0 MPa. The highest conversion of 70 % to tri-TMP-ester wasobtained with 2.0 MPa (Figure 38).

68

Figure 38: Effect of the pressure on the TMP-ester conversion, (molar ratio 1:4.5, 40% (wlw) lipase, 15 % (wlw) added water, 37°C, 24 h).Normal pressure (.); 12.0 MPa (0); 5.3 MPa (.);2.0 MPa (0).

/

4 5

time (h)

11 24 30

3.11. IMMOBILISATION EXPERIMENTS FOR PREPARATION OF TMP-ESTERSFROM RAPESEED OIL METHYL ESTER (IV)

Because previous experiments with different esters had clearly suggested thatconversion of RME to desired ester might increase by immobilising the lipase,different carriers for immobilization were tested. They were Celites R-630, R-640 andR-626, Diatomite, Duolite ES-762 (neutral adsorbtion resin), GCC, GDC 220 (weakalkalic anion exchange resin), HPA 25 (strongly alkalic anion exchange resin),silicagel and WA 30 (weak anion exchange resin). The reaction conditions werethose found optimal in earlier experiments (3.10.1- 3.10.6.): temperature 47°C,pressure 5.3 MPa, lipase 40 % (wlw) , added water 15 % (wlw), substrate ratio 1:4.5,and reaction time 42 hours. The results are shown in Figure 39. With WA-30 andDiatomite immobilised lipase no reaction could be seen. With Duolite ES 762, GDC,GCC and HPA-25 immobolised lipase no improvement in conversion could be seen.With silicagel immobilised lipase reaction ceased to mono- and di-ester of TMP. Theuse of Celite R-630 as a carrier gave the highest conversion of 95 % to the tri-esterin 22 hours.

69

A comparison experiment was done with commercial, immobilized MIcor mieheilipase (Lipozyme 1M 20) under the same reaction conditions, except that no waterwas added. A conversion of 92.5 % was obtained. The results show that byimmobilizing the Candida rugosa lipase to Celite R-630 conversion to the desiredester was at the same level as without immobilization (95%). The reaction time,however, was shorter compared to nonimmobilized lipase reaction (22 h comparedto 42 h or 68 h).

Figure 39: Effect of the immobilised lipase on the TMP-ester conversion, (molar ratio1:4.5, added water 15 % (wlw) , immobilised lipase 40 % (wlw) , 47"C, 5.3

MPa, 42 h).

-..- ..- - ..-~ 70

-;: 60o

00; 50

~ 40"8 30

2010

o

100

90

80

"' ~ u~ g a; .~ m

Ql u

'" Em'" '6 '" ;, ;," "iii

~~ " Z :: ~u

~'" ";;; is."

70

3.12. APPLICATIONS AND USES OF RAPESEED OIL ESTERS

3.12.1 Methyl-, ethyl- and 2- ethyl-1-hexyl esters of rapeseed oil

The methyl-, ethyl- and 2-ethyl-1-hexyl esters of rapeseed oil have been used forexample as pure solvents or as solvents to replace organic solvents in detergents.Pure solvent use is mainly in printing ink industry, for example to clean printing inkrollers. First detergents using vegetable oil esters have been on the market for abouttwo years. Table 13 demonstrates how these detergents remove a dirty, solid phase,which results from biodegradable hydraulic fluids dried on metal surface undersunshine at extended periods.Methyl- and ethyl esters of rapeseed oil are also used as fuel, biodiesel (71-76).They were developed for environmental reasons. The main purpose is to reduceexhaust gas emissions (Table 14).

Table 13: Different detergents in test. Test conditions (77): Steel plates (washedand dried by acetone), 100 mm3 tested oil, 1 min. The plates are underUV-Iamp, 45-47°C (max), t varies, depending on the tested oil, until thefilm formation takes place. Washing: with pure liquid, 5 %, 10 % and20% solutions, measured effiency and time.

Detergent Type Effect Time

Company +/+++++ (min)

I Environmentallyrecommended

Limo/Solmaster limonene +++ 15-20Pro-Clean/Ab Ind. env.friendly +++ 10-15tekniklSwed enEcoclean autol RME +++++ 5-10

Trans-Clean

II Traditional

MT-Exima 095/TK- 10-15+++

TK-123/Noiro/10-15Orion

+++

Goldfain2000/ ++++ 10-15KS-Chemitra Ltd.

71

Table 14: Exhaust gas emissions of normal summer diesel ( DIK), and fromrapeseed oil bio diesel (Raisio) (modified from 71).

Fuel Exhaust gas emissions

(gIkWh)

Carbon HC-part PM-30 CO HC NOxparticles

DIK 0.18 0.36 0.54 4.57 0.58 12.15

RME 0.04 0.27 0.31 2.91 0.45 13.36

RME+ 0.04 0.14 0.18 0.24 0.11 13.10catalysator

limit 11.2 2.4 14.4

3.12.2 Hydraulic fluids made from rapeseed oil esters (V,VI)

TMP-ester based lubricants made in this work were compared with pure comparablevegetable oil based and commercial synthetic ester based hydraulic fluids, usingcommercial mineral oil based hydraulic fluid as a reference.Main purpose for a hydraulic fluid is to reduce friction and decrease wear. It has tohave stability against heat, cold, oxidation and corrosion. It also has to be able tomaintain these qualities for extended periods. All of these properties can of coarsebe improved with different additives, but the basic fluid is of crucial importance.These types of synthetic esters such as pure vegetable oils have great advantage inthe reduction of wear and friction due to their molecular structure (Figure 40).

72

Figure 40: The molecular structures of vegetable oil, TMP-ester of rapeseed oil andmineral oil (78,79,80).

CH2-0-gRbH- O-gRbH2-0-8-Rvegetable oil

92HS

C2Hs-CH2-CH-CH2-«H-CH3C2Hs

CH2-0-gRCH3-CH2-6-CH- 0-8.R

6H2-0-g-RTMP-ester of rapeseed oil

mineral oil components

All esters include oxygen in their structure. Because of this oxygen these fluids canform a very stable, unimolecular film over a metal surface (78,81,82,83). A majoradvantage is their biodegradability and non-toxicity. They are not bioaccumulatingeither. In many countries these requirements are essential for lubricants (25). Ageneral composition of hydraulic fluids from rapeseed oil TMP-ester is describedbelow.

Component

TMP-ester of rapeseed oilantioxidantpour-point depresserantiwearantifoamer

Amount(%(wlw»

90- 980.5- 2.51.0- 5.00.4- 2.00.1- 0.5

With the same procedure also vegetable oil and commercial, synthetic ester basedhydraulic fluids were made.

73

The hydraulic fluids have to pass certain standard tests to fulfill necessary criteria tobe classified as lubricant. In the next chapters the main tests, their purpose andlimits will be discussed together with test results of the three different types ofhydraulic fluids.

3.13. Characterization of rapeseed oil ester based hydraulic fluids (V,VI)

The lubricants tested in the present work were commercial trimethylolpropane esterbased hydraulic fluids (RBS 46S (A) and 68S (B», commercial vegetable oil basedhydraulic fluid (RBS 32L (C», reference mineral oil based hydraulic fluid (0) andhydraulic fluid based on trimethylolpropane ester developed in this thesis, viscositygrade 32 (E). These fluids are mentioned and marked in following text by theirletters (A)-(E).

3.13.1. Viscosity (V,VI)

Viscosity indicates how f1uidy a lubricant is. It is very important that the viscosity of alubricant is as stable as possible with temperature, which means that viscosity valuedoes not change with time at certain temperatures. A capillary viscosity value isbased on Hagen-Poiseville"s law (84,85) according to (AA):

{};: K x t, where K;: constant (mm2/s2) depending on viscosimeters capillary

length and diameter( AA)

t;: time (s) needed for fluid to run a certain length in thecapillary

For hydraulic fluids the most common viscosity grades are 32, 46 and 68 mm2/s. Tobe within its grade, the viscosity of hydraulic fluids can maximally vary +/- 10 %.According to ASTM D 445 viscosity is generally measured at 40°C and 100°C and aviscosity index is calculated on the basis of the results. Viscosity index is a numbertelling how stable the fluid is with temperature. The higher it is, more stable the fluidis with temperature. Viscosity Table 1 in publication VI shows that viscosities are intheir grade, except RBS 46S, which is a little high, 50,4 mm2/s, when the highestlimit is 46,9 mm2/s. Viscosity index for vegetable oil and their ester based hydraulicfluids are typically higher than comparable mineral oil based products. A typicalmineral oil value is around 180, here 187 and for vegetable oil or its esters, thevalue is around 220.

74

3.13.2 Filterability (V,VI)

Good filterability is necessary for well functioning hydraulic systems. Particles inhydraulic fluids naturally increase wear. The filterability is always measured bothfrom the raw material and the final prodUct: hydraulic fluid.The method used is under standardization and under development at the RoyalTechnical University of Stockholm together with users (86). It is called percentage­method, because results are given in percents. The higher the value is (max 100 %),the better. Figure 41 demonstrates the filterability equipment and its basic function(85). Because the test is under development, there are no official limits forfilterability, but for the research and development period, a limit has been set at 80%. This limit is well accepted by production and customers. Filterability results varyquite a lot. Commercial synthetic ester based fluids have a filterability of only 63 %.This figure can improve, when oil has been in use for a while, but it can causeproblems before this. The main reason for such a low value are particles in the rawmaterial, which has not been purified as well as should have done. An other reasonis that the additive package used was not the right one for this type of raw material.The mineral oil based fluid had a little better result, 72 %, but needs also a furtherstudy of its particles. Vegetable oil and its esters all passed the demanded 80 %,being 80 % to 98 % as can be seen in Table 2NI.

Figure 41: The filterability equipment of the percentage-method and its basicfunction.

75

3.13.3. Standard of purity and impurity particles (V,VI)

Standard of purity and impurity particles are very meaningful characteristics inhydraulic use. Small particles can cause various problems in fine hydraulic systems,for example extremely high wear. Small particles can already be present in the basicfluids, in additives, or in both. Therefore, the amount of particles is calculated beforeuse. This is done according to ISO 4406-standard by microscopically calculatingparticles larger than 5 m, but smaller than 15 m, and particles larger than 15 mmicroscopically (87). The type of crystals affecting the filterability can also bedetermined microscopically.Requirements for the cleanclass of hydraulic fluid varies depending on its end use,being around 16/13. As can be seen from Table 3M, the cleanclass 13/8 ofvegetable oil and its esters based hydraulic lubricants fulfill very well these demands.

3.13.4 Foaming (V,VI)

Foaming can cause problems, in the machine operating and instability frombeginning in the hydraulic fluid tank. The less foaming, the better it is for the user.Foaming is measured by ASTM D892-standard (88). Results are given in time (s)and height (mm),that is how high the foam was originally and how long time it takesfor foam to break down. In this method air is bubbled at a certain pressure into acertain volume of tested oil for a certain period of time (87). The maximum values forfoaming are 600 mm and 240 seconds. The results are shown in Table 15.Differences in the foaming of different vegetable oil and its ester based hydrauliclubricants were not big, the time varies from 120 s to 175 s and the height from 20mm to 33 mm. For RBS 688, 40s and 8 mm, exceptionally good results wereobtained.

76

Table 15: Foaming (ASTM 0 892) and colour (ASTM 0 1500) of some hydraulicfluids.

Hydraulic fluid TMPE R8S 32L R8S 46S R8S 68S MO(E) (C) (A) (8) (0)

Foamingtime (s) 175 120 151 40 150height (mm) 33 20 28 8 30

Colour 4- 3+ 4 4

3.13.5 Colour (V,VI)

Colour is measured by the ASTM 0 1500-standard. In this method the yellowishcolour of the tested oil is compared to reference colours on a yellow-orange-brown­scale. The result is given in a scale from 1 to 20. The lighter the oil is, the smallerthe number (88). The colour is light in a new hydraulic fluid, but darkens quitequickly in use. This can be seen well in the field test results (Table 8, in V). Thisdarkening is mainly caused by antioxidants.

3.13.6. Cold stability (V,VI)

Cold stability of hydraulic fluid is very important for machines working out of doorsduring winter time, such as forest machinery. According to the ASTM 0 97 standard,pour point is one way to measure cold stability. It indicates the temperature, at whichthe fluid is still f1uidy (88).Cold stability can also be expressed as longterm cold stability by indicating howmany days the tested fluid stays f1uidy at a certain temperature (VTT-standard).Further, viscosity can be measured at cold temperatures according to ASTM 0 445standard (88). Viscosity should be and should remain at a certain level at a certaintemperature in order not to cause trouble e.g. in COld-starting, if hydraulic fluid is tooviscous. It is important that the pourpoint is lower than -30°C for winter use out ofdoors in Nordic countries. Machine producers recommend as a limit for cold startingthe viscosity of hydraulic fluid to be less than 5000 mm2/s. The results have been

77

collected in Table 16. As can be seen from Table 16, the viscosity of RBS 46S andRBS 68 S at cold temperatures was lower than 5000 mm2/s. No problems have,however, been noticed in field tests with these oils. With vegetable oil basedhydraulic fluids, the viscosity at cold temperatures is not a problem. It stays around3000 mm2/s at least for a week.

Table 16: Pourpoint (ASTM D 97) and viscosity at cold temperatures (ASTM D 445)of different hydraulic fluids.

Hydraulic fluids

Pourpoint (0C)

TMPE(E)

-41

RBS 32L(C)

-39

RBS 46S(A)

-36

RBS 68S(B)

-39

MO(D)

-40

Viscosity (mm2/s)Temperature (OC)o-10-20-30Time (days)137

428 372 5731030 829 1430 21601540 1628 1480 36703114 2940 4060 7170

305030503050

3.13.7. Friction and wear (V,VI)

As explained in chapter 3.12.2, all of the basic fluids studied here have goodproperties for reducing friction and wear, which is one of the main functions oflubricants (78,79,82,83,89,90). There are several standardised methods to measurethese properties at the laboratory level (75). One of them is the fourball test(standards ASTM D 2783, IP 239) (88,91), in which wear under loading andmaximum loading at which the lubrication fails and the lubricating film breaks aremeasured. Limits for hydraulic fluids are <0,5 mm (damage is small) and >1,0 mm(damage is large). The bigger the maximum loading is and the smaller the wear, thebetter. The test results are shown in publication V, Table 4. In all vegetable oil andits ester based hydraulic fluids wear was under 0,5 mm. Under maximum load thedifferences, from 2000 N to 3000 N, are due to different additive packages.

78

3.13.8. Oxidation stability (V,VI)

Poor oxidation stability has been claimed to be one of the weaknesses of vegetableoil based hydraulic fluids. With the right choice of raw material and additives this isno longer a problem. With TMP-esters of rapeseed oil, oxidation stability can beeven improved slightly. Table 17 gives the test results from two different oxidationtests with ASTM D 525, oxidation bomb test and DIN 51586 viscosity change test(78,82,83,88,89,92). In Figure 42 is shown the apparatus of ASTM D 525. In theoxidation bomb test oxygen is lead to a steel vessel, Which is filled with the testedoil. The steel vessel is kept in a hot water bath (98 +/- 2°C). Either the time for totaloxygen consumption or oxygen pressure drop within a certain time is measured. Thelonger the time or the smaller the pressure drop is, the better. In DIN 51586 theviscosity of the tested fluid is measured before the test and after 312 hours at 95°Ctemperature under air bubbling (10 dm3/h). The smaller the difference is, the better itis. The latest oxidation test modified for vegetable oil based products is the Baader­test, DIN standard 51 554 teil 3. The temperature is 95°C and the time needed 7days. The oil (60 mm3

) is mixed with a copper catalyst. The total acid number andviscosity at 40°C are measured before and after the test. The changes arecalculated, the smaller they are, the better. The limit for acid number change is < 0,6mg KOH Ig and for the viscosity < 20 %.

Figure 42: Oxidationbomb test (ASTM D 525-standard) equipment.

OXYGEN9 b~r

TESTOIL .

79

Table 17: Oxidation stability measurements (ASTM D525, DIN 51586-, DIN 51 554standards) from some hydraulic fluids.

Standard TMPE RBS 32L RBS 46S RBS 68S MO(E) (C) (A) (B) (D)

ASTM D (psi) 42 30 39 29 39

DIN 51586 (0/0) 12.4 28.8 20.3 24.1 15.8

DIN 51554TAN (mgKOH/g) 0.02 0.05 0.13 0

visco (0/0) 4.2 6.2 0.86 5.5

3.13.9. Corrosion stability (V,VI)

It is essential that the lubricant does not corrode the metal surfaces with which it isin contact. This can also be measured at the laboratory level. Widely usedstandardised test is the Cincinnati-Milacron test, which does not only measurecopper and steel corrosion, but also what happens to the lubricant in contact withmetals during heating: Does it oxidize? Does there begin chemical reactionscatalysed by metals?In this test viscosity and total acid number of a 200 mm3 oil sample with steel andcopper plates are measured before and after the test at 135°C for 168 hours and thedifference is calculated on percent. Total sludge formation is measured after the test(93).As can be noticed from the results (table 7, VI) TMPE developed in this researchgave the best results. The difference in total acid number was significantly smaller,0.17 mg KOH/g, as compared to vegetable oil and commercial esters, 1.01 mgKOH/g to 1.74 mg KOH/g. Also the total sludge formation, 1.1.mg/100 mm3

, is muchsmaller in the TMPE developed in this research compared to others, 28.8 mg/100mm3

. Total sludge is generated mainly with additive package. Other properties suchas viscosity change and copper and steel plates weight changes were similar with alltested fluids.

80

3.13.10. Biodegradability (V,VI)

Biodegradability can be tested with many different methods. Table 18 shows resultsas tested by the CEC-L-33-T-82-standard (89,94). These figures are compared withother test results with RB5 32L. Bioaccumulation and toxicity properties of lubricantsare equally important characteristics as biodegradability. Bioaccumulation measuresif the product or parts of it accumulates in nature in micro-organisms, animals, plantsetc. It can be measured for example by GECD standard tests. Toxicity should bemeasured together with bioaccumulation by GECD standards, because the productmay be toxic when disposed in nature.

Table 18 : Biodegradability of some hydraulic fluids.

Testmethod

CEC-L-33-T-82

. TMPE

>90 %

RB5 32L RB5 465 RB5 685 limit

>90 % >90 % >90 % 70 %

DIN 38412

GECD 301F

5 days

>85 %

14 days

60 %

3.14. Conclusion of laboratory tests (V,vl)

TMP ester developed in this thesis seemed to meet its requirements as a basic fluidfor hydraulic fluids very well. Laboratory testing showed that especially at highertemperatures TMP ester worked better than a vegetable oil based hydraulic fluid andas well as comparable, commercial synthetic ester based hydraulic fluids. Theoxidation stability of TMP-esters was better than that of a vegetable oil basedhydraulic fluid and as good as, or even better than, a comparable commercialsynthetic ester based hydraulic fluid. This difference could be best observed athigher temperatures. In decreasing friction and wear all of these three differenttypes of hydraulic fluids were excellent. Cold stability of the TMP ester was betterthan that of vegetable oil or commercial synthetic ester. The pourpoint was lowerand viscosity stability slightly better than with references. The next step is to finalizealready started field tests, to ensure the suitability of TMPE based hydraulic fluids forthe market.

81

4. CONCLUSIONS

The aim of this work was to find new ways to produce and use of vegetable oilbased esters.The alcohols used were chosen with the end use of the required ester in mind. Theywere short, straight chain alcohols such as methanol, ethanol, etc. or branched chainalcohols such as 2-ethyl-1-hexanol or polyols like trimethylolpropane. With shortchain alcohols esters were produced chemically with excellent conversions (95 -100%). These esters were also made on an industrial scale. They are mainly used asfuel or solvents. Customers have been very pleased with these environmentalacceptable products. However in use as solvent problems such as swelling orbrittling of gum- and seal materials arose.The next step was to use longer, branched chain alcohols, like 2-ethyl-1-hexanol.Such esters were developed for the first time from rapeseed oil both chemically andenzymatically. Good conversions, 96-100 %, were obtained using both methods. Inchemical synthesis the best results were obtained under reduced pressure andsufficiently long reaction time. Rapeseed oil was used in excess of stoichiometricamount, and the catalyst was alkaline sodium methylate. The conversion was 97.6%. In enzymatic synthesis best results were obtained by Candida rugosa lipase. Themolar ratio was near to stoichiometric, added water was needed, and the amount oflipase was 7,4 % (wlw). On the laboratory scale the conversion was 99.8%. Theenzymatic reaction was also examined on a pilot scale. The conversion was 87 %.The lipase was immobilized to different carriers, of which XAD-7-resin, gave in pilotscale 90 % conversion. For the chemical reaction, higher temperatures and reducedpressure was needed than for the enzymatic reaction. This partly compensates thefact, that lipases are more expensive than chemical catalysts, when bearing in mindthe costs of the reactions. No differences could be seen with the preparationmethods in either the conversions or in the purity of the wanted esters.Both esters synthetized by chemical catalyst or enzymatic biocatalyst have been infield tests e.g. in detergent industry use to replace traditional organic solvents. Thetest results have been promising and these esters are today used for example incarshampoos.The chemical synthesis was also applied in tallow oil fatty acids esterification withpolyols. The best result, the conversion of 87 % and 6 % of unreacted fatty acid wasobtained at reduced pressure and middle temperature at quite long reaction time, 70­75°C/2-9 h. The molar ratio was stoichiometric and the catalyst was acidic p­toluenesulfonic acid. Because there were several serious difficulties with thissynthesis, the next step was to use transesterification instead of esterification.Transesterification was done both chemically and enzymatically.Instead of fatty acids, the raw material was the methylester of rapeseed oil.Chemically from rapeseed oil methylester was succesfully synthetized with

82

trimethylolpropane trimethylolpropane ester of rapeseed oil with conversion of 99 %.. This reaction demands reduced pressure and at least two different temperature

areas. The methyl ester of rapeseed oil was used a little over stoichiometric quantityand alkalic sodium methylate was used as catalyst. The fatty acid composition of theproduct is similar to that of comparable,commercial synthetic esters. The productionof ester by the method developed in this work, is economically much cheaper thanwith the methods based on purified fatty acids. TMP-ester of rapeseed oil has alsobeen produced chemically in pilot scale. The yield was 99.0 %. Enzymatically TMP­ester of rapeseed oil was produced by Candida rl/gosa:.Jipase at a temperature, whichthis lipase could tolerate. The reaction was proceed at reduced pressure with longtimeschedule, 47°C, 5.3 MPa, 42 h. The methylester of rapeseed oil was used inexcess of stoichiometric to obtain the conversion of 98 %. The immobilizaton oflipase was also researched. The best carrier was Celite R-630 with a conversion of95 %. As a comparison, a commercial immobilized lipase, MIcor miehei-lipase(Lipozyme 1M 20), was used. The conversion was 92,5 %. In the chemical reactionslightly better conversion was achieved than in the enzymatic reaction. The reactiontime was much shorter in the chemical preparation, but at the same time thetemperature was much lower in the enzymatic preparation. In the chemicalpreparation also more triglyceride ester of TMP was achieved.Both of these esters were used as starting materials for producing hydraulic fluids forlaboratory tests. The esters were compared to commercial synthetic ester basedhydraUlic oils, and to rapeseed and mineral oil based hydraulic fluids. As aconclusion of the laboratory tests the trimethylolpropane ester of rapeseed oil basedhydraulic fluid resisted oxidation better than the rapeseed oil or mineral oil basedproducts, and equally well or even better than commercial synthetic ester basedfluids. This difference was even more noticeable at higher temperatures. The coldstability properties of trimethylpropane ester of rapeseed oil based fluids were betterthan those of the rapeseed oil,mineral oil and commercial ester based fluids, as wasevident by the viscosity stability at cold temperatures.Friction and wear properties were good for all vegetable oil based products, betterthan for mineral oil based fluids. The TMP-esters developed in this work are readyfor fieldtests, which are essential before commercialization.In this work new methods to produce vegetable oil esters were explored. Themethods have been succesfully used either in industrial or pilot scale productionwith good yields. The aims of this work were fulfilled extremely well.

83

REFERENCES

1. Boylan, J.B., Proceedings of the NLGI 53rd Annual Meeting, San Diego,CA, 1986pp. 185-195.

2. Niedzielski, E, Ind. Eng. Chem., Prod. Res.Dev. 15(1) (1976) 54-58.3. Szydywar, J., J. Synth. Lubr. 1 (1984) 153-169.4. Van der Waal, G. and Kenbeek, D., Proceedings of the Tribology 2000, 8th

International Colloquium,vol IV, 1992, 13.3.5. Vilenius, M., Proceedings of the Nordtrib 1984, Nordic Symposium on Tribology,

August, 1984, Tampere Finland, pp.15.6. Meffert, A., J.Am.Oil Chem.Soc. 61 ( 1984) 255-258.7. Deak, G. and Bartha, L., Proceedings of the Tribology 2000, 8th International

Colloqium, vol II, Technische Academie Esslingen, Germany, Abril, 1992, p.13.5.8. Henkel KGaA, German Patent 3 521 711,1986.9. Basf, German Patent 3 635 490, 1988.10.Carr, D.D. and DeGeorge, N., U.S. Patent 4 175047 (1979).11.Mittelbach, M., Worgetter, M., Pernkopf, J. and Junek, H., Energy in Agriculture 2

(1983) 369-384.12.Freedman, B., Pryde, EH. and Mounts, T.L., J.Am.Oil Chem.Soc. 61 (1984)

1638-1643.13.Mittelbach, M. and Tritthart, P., J.Am.Oil Chem.Soc. 65 (1988) 1185-1187.14.Mittelbach, M., Tritthart, P. and Junek, H., Energy in Agriculture 4 (1985) 207­

215.15.Kusy, P.F., Vegetable Oil Fuels- Proceedings of the International Conference of

Plant and Vegetable Oils as Fuels, Am.Soc.Agr.Eng.,St Joseph, 1982, pp. 127­137.

16.8radshaw, G.C. and Mealy, W.C., U.S.Patent 2360844 (1944).17.Wildersohn, M., Tribologie + Schmierungstechnik 32 (1985) 70-75.18.Boylen, J.B., Nat. Lubr. Grease Inst. Spokesman (1987) 185-195.19.Kiviniemi, S., Master of Science thesis, University of Oulu, Department of

Chemistry, (1993), pp.75.20.Jelitte, R. and Beyer, C., Proceedings of the Tribology 2000, 8th International

Colloquium"volll, Technische Academie Esslingen, Germany, Abril, 1992,p.13.4.

21.Groshart, E, Metal Finishing 87(3) (1989), 49-51.22.Schmid, K.-H., Fat.ScLTecho/. 89 (1987) 237-248.23.Nikorov, EM., Sazonova, N.S., Petrova, L.N. and Ataeva, O.V., Chem.Techo/.

Fuels Oils 18 (1982) 517-519.24.Malec, R.E, U.S.Patent 3 673 226, 1972.25.Stempfel, E.M. ans Schmid, L.A., Proceedings of the Tribology 2000, 8th

International Colloquium, vol I, Technische Academie Esslingen, Germany, Abril,1992, p.10.1.

26.Van der Waal, G.B., Proceedings of the Nat.Lubr.Grease Inst.55th AnnualMeeting 53 (1989) pp. 359-368.

84

27.Gunderson, R.C. and Hart, A:W, In: Synthetic Lubricants, Reinhold, New York,1992, pp. 388-399.

28.Moura, D. and Legeron, J.-P., J. Synt. Lubr.8(1) (1990) 15-27.29.0di-Owei, S.J., Soc.TriboI.Lubr.Eng. 45 )1989) 685-690.30.Ethyl, U.S.Patent 1 301 003, 1970.31.Berens, G. and Milton, L.H., U.S.Patent 4 263 159 (1981).32.Cooley, S.D. and Siovinsky, M., U.S.Patent 2991 297 (1961).33.Metro, S.J., Hoffman, E.M. and Matuszak, AH., Canadian Patent 859771 (1970).34.Schick, J.W and Kaminski, J.M., U.S.Patent 4 175047 (1979).35.Leleu, G., Bedague, P. and Sillion, B., U.S.Patent 4061 581 (1977).36.Macrae, AR. and Hammond, R.C., BiotechnoI.Gen.Eng.Rew. 3 (1985) 193-217.37.Macrae, AR., In: Microbial Enzymes and Biotechology Applied Science

Publishers, London, 1983, pp.225-250.38.Macrae, AR., J.Am.Oil Chem.Soc. 60 (1983) 291-294.39.Kilara, A, Process Biochem. 20(2) (1985) 235-245.40.Stryer, L., Biochemistry, 2nd edition,WH.Freeman and Company, San Fransisco,

1981 ,pp.32,131.41.Bj6rkling, F., Godtfredsen,S.E. and Kirk,O.,Tibtech. 9 (1991) 360-363.42.Yamane, T., J.Am.Oil Chem.Soc. 64 (1987) 1657-1662.43.Buhler, M. and Wendrey, C., Fat.ScLTechnol. 94 (1992) 82-94.44.Malcata, F.x., Reyes, HR, Garcia, H.S., Hill Jr., C.G. and Amundson, C.H.,

Enzyme Microb.Technol.14 (1992) 426-446.45.0sterberg, E., Blomstrom, A-C. and Holmberg, K., J.Am.Oil Chem.Soc. 66

(1989) 1330-1333.46.Jacobsen, T., Olsen, J. and Allerman, K., Biotechnol.Lett.12 (1990)121-126.47.Sonnet, P.E., J.Am.Oil Chem.Soc. 65 (1988) 900-904.48.Tahoun, M.K., Fat.ScLTechnol. 89 (1987) 318-320.49.Zaks, A and Russel, AJ., Biotechnol. 8 (1988) 259-270.50.West, S., Food Technol. 42 (1988) 98-102.51.Takaharu, M., Sawamura, N., Hashimoto, Y. and Hashida, W, U.S. Patent 2 035

359 ( 1979).52.Hills, J.M. and Mukherjee, K.D., Appl.Biochem.Biotechnol. 26 (1990) 1-10.53.Posorske, L.H., J.Am.Oil Chem.Soc. 61 (1984) 1758-1760.54.Mittelbach, M. , J.Am.Oil Chem.Soc. 67 (1990) 168-170.55.Ergan, F.,Trani, M. and Andre, G., J.Am.Oil Chem.Soc. 68 (1991) 412-417.56.Shaw, J.-F., Wang, D.-L. and Wang, Y.J., Enzyme Microb.Technol.13 (1991) 544-

546.57.Miller, C., Austin, H., Posorske, L. and Gonzlez, J., J.Am.Oil Chem.Soc. 65(6)

1988927-931.58.Malcata, F.x, Reyes, H.R., Garcia, H.S., Hill Jr., C,G, and Amundson, C.H.,

J.Am.Oil Chem.Soc. 671990890-910.59.Halling, P.J., Enzyme Microb.Technol., 16, 1994, 178-206.60.Svensson, I., Wehtje, E., Adlercreutz, P.and Mattiasson, B., Biotechnol.Bioeng.

44 (1994) 549-556.

85

61.Zaks, A and Klibanov, AM., Science 224 (1984) 1249-1251.62. Khan, SA, Halling, P.J. and Bell, G., Enzyme Microb.Technol.12 (1990) 453-458.63.Linko, Y.-Y. and Linko, P., In:Biocatalysts in Organic Synthesis, (J.Tramper,H.C.

van der PIas and P.Linko,eds) Elsevier, Amsterdam 1985 pp.159-178.. 64.Linko, Yu-Yen, Koivisto, U.-M., and Kautola, H., Ann. New York Acad. Sci. (1990)

613-691.65.Sharma, SK, Vasudevan, P. and Tewari, V.S., Tribology 16(4) (1983) 213-219.66.Lee, Y.-K. and Choo, C.-L., Biotechnol. Bioeng. 33 (1989) 775-778.67.Goldberg, M., Thomas, D. and Legoy, M.-D., Enzyme Microb.Technol.12 (1990)

976-981.68.Chowdhury, J. and Foudy, K., Chem.Eng.2 (1993), 35-39.69.Kobitz, w., News from Bio-Diesel- world, December, 1993, Vienna, Austria, pp.3.70.Mikkonen, S., Neste, research raport 183/1991, Augus~ 1991, pp.19.71.Report of Biodiesel-project in Finland, Ministry of Agriculture and Forestry, 1992.72.Tollefson, C., Chemical Marketing Reperter, August 9th, 193, 11.73.Dessau, R.M., EPO 01 35385 (1984).74.Germani, M.M.G., German Patent DE 43 24 875 (1993).75.vellguth, G., SAE Tech.Pap. No 831 358 1983 12p.76.voltz, M., Schmierungstechnic 22 (1991) 177-181.77.Stenmark, G. and Vainio, B., Test Method for Detergents,R&D Laboratory., Raisio

Oil Milling Div.,Raisio,1986.78.Lamsa, M., Proceedings of the 2nd Tampere International Conference on Fluid

Power,19-21 March, 1991, Tampere, Finland,pp. 73-89.79.Klein, E.W., Wire World International 27 (1985) 130-132.80.Tedrow, L.E., Tappi Proc.,Atlanta,1985 535-542.81.Plagge, A, Tribologie+Schmierungstechnik 341987148-156.82.Lamsa, M., Proceedings of Technische Akademie Esslingen, Biologisch

Abbaubare Schmierstoffe und Arbeitsflussigkeiten,12-14 February, 1992,Esslingen, Germany, no 13, p.1-14.

83.Lamsa, M., Proceedings of 9th International Colloquium in Tribology 1994, 9-11January, 1994, Esslingen, Germany, voI.12.8.

84.0perating Instructions Practical Hints for Making Viscosity Measurements withCapillary Viscometers, Schott Gerate Gmbh.,1985, pp.20.

85.Cosmancini, E. and Virtuani, M.L., Proceedings of Colloqium TechnischeAcademie Esslingen, January, 1986,pp.7.

86.Raport of Filterability- project, Royal Institute of Techn.,dep. of fluid techn.,Stockholm, Sweden, 1988-1994, pp.80.

87.ISO- standard, SFS, Fed. of Stand. in Finland,P.O.Box 116, Helsinki,Finland.88.Annual Book of ASTM Standards, vol 05.01.,Petr.Prod. and Lubr.,

ASTM,Philadelphia,USA89.1hrig, H., Tribologie+Schmierungstechnik 39 1992 121-125.90.Newingham, T.D., Hydraul.Pneum. 198754-56.91.Standards of the Institute of Petroleum, United Kingdom, 1982.92.DIN-Taschenbuch 192, Schmierstfoffe,Beuth Verlag Gmbh,Berlin,

1983.

86

93.The Lubr.Eng.Manual, 1st ed.,United Steel Corp., USA.94.CEC-L-33-T-82, Biodegrab. of two-:stroke cycle outboard engine oils in water,

London.

Appendix Iwith the permission of the publisher,

Lipase-Catalyzed Transesterification of Rapeseed Oiland 2-Ethyl-1-HexanolY.·Y. LinkoB,*, M. Uimsab, A. Huhtalaa and P. LinkoBaHelsinki University of Technology, Laboratory of Biotechnology and Food Engineering, FIN-02150 Espoo, Finland and bRaisioGroup Oil Milling Industry, FIN-21200, Raisio, Finland

14111//

Lipase-catalyzed transesterification (alcoholysis) of low­erucic acid rapeseed oil and 2-ethyl·l·hexanol without anadditional organic solvent was studied in stirred batchreactors. Of a number of commercially available enzymesinvestigated, the best results were obtained with a Can­dida rugosa lipase. The optimal transesterification condi­tions were an oilIalcohol molar ratio of 1:2.8, a minimumof 1.0% (wlw) added water, and with a temperature of37-55°C. Under the optimal conditions, a nearly completeconversion was obtained in one hour with 14.6% (wlw)lipase, whereas 0.3% (wlw) lipase required 10 h for similarresults. The enzyme was inactivated at 60°C.

KEY WORDS: Alcoholysis, biocatalysis, enzyme, 2-ethyl·l-hexanol,2-ethyl-l-hexyl ester, lipase, rapeseed oil, transesterification.

Interest in Iipase-catalyzed biosynthesis is rapidly increas­ing (1,2). Lipases have great potential in food-related lipidmodifications (3,4), in the production of esters (5,6), biode­gradable polyesters (7,8) and fatty acids (9,10). Lipases(triacylglycerol acylhydrolase, EC 3.1.1.3) are esterases thatcatalyze hydrolysis and synthesis of glycerol esters. In trans­esterification, the acyl moiety is exchanged either betweenan ester and acid (acidolysis), ester and alcohol (alcoholysis)or two esters (acyl exchange) (10,11). Acyl exchange betweentwo molecules is also called interesterification, and betweentwo acyl groups within a molecule it is called intraesterifica­tion. Ester synthesis is favored under restricted wateravailability (low water activity) (12), although a minimumquantity of water is necessary for enzyme catalysis to takeplace (131.

Lipase-catalyzed alcoholysis in the absence of solvent isimportant in industrial applications, especially for food uses.Complete transesterification between one mole of triacyl­glycerol and three moles of alcohol yields three moles of esterand one mole of glycerol (14). Although it has been claimedthat the presence of additional organic solvent may beuseful for example, in controlling water activity andmicrobial contaminations (15,16), the absence of solventallows higher substrate and product concentrations (17),simplifies downstream processing (18), and improves safety(19). Zaks and Klibanov (20) were the first to study thetransesterification of tributyrin with a number of primaryand secondary alcohols, catalyzed by porcine pancreaticlipase, to yield an ester of butyric acid and dibutyrin. Theydemonstrated the importance of the quantity of water pres­ent, both to the activity and stability of the biocatalyst.Macrae (12) and Halling (21), among others, have furtheremphasized the importance of water relationships in lipase­catalyzed synthetic reactions. Hirata et aL (22) later demon­strated that water requirements in the alcoholysis oftributyrin by different lipases may vary widely.

In spite of the great technical interest of fatty acid esters(23,24), lipase-catalyzed transesterification involving high­molecular weight fatty acids has only recently been in­vestigated.. Mittelbach (25) has studied the alcoholysis of

*To whom all correspondence should be addressed.

Copyright © 1994 by AOCS Press

sunflower oil both in petroleum ether as solvent and withoutadditional solvent, to synthesize methyl and ethyl esters asdiesel oil substitute. Shaw et aL (26) used Celite-inunobilizedPseuOOmonas fluorescence lipase in the alcoholysis of oliveoil An excess of alcohol has been claimed to be beneficialin alcoholysis by suppressing the hydrolytic side reaction(17,18). Nevertheless, in such cases, disadvantages such asthe removal of excess alcohol from the product should alsobe considered. Erueyl erucate, the main component of jo­joba oil, has been produced by transesterification of high­erucic acid rapeseed oil and erucyl alcohol (18). The aim ofthe present work was to investigate Iipase-catalyzed trans­esterification (alcoholysis) of rapeseed oil and 2-ethyl·1­hexano!, currently used in the chemical synthesis of anumber of important compounds (27), without the use ofan additional organic solvent and as a possible alternativeto acid or base catalysis.

MATERIALS AND METHODS

Materials. Refined, low-erucic acid rapeseed oil and syn­thetic rapeseed oil2-ethyl-1-hexylester were obtained fromthe Raisio Group (RaisiO, Finland). The approximate fattyacid composition of the oil was 57% oleic acid, 22% linoleicacid. 12% linolenic acid, 4% palmitic acid, 1% stearic acid,2% eicosanoic acid <1% erucic acid and 1% others.2-Ethyl-1-hexanol was obtained from Fluka Chemie AG(Buchs, Switzerland). MonO', di- and triolein standardswere from Sigma (St. Louis, MO), and glycerol from May& Baker (Dagenham, United Kingdom).

Enzymes. The following powdered microbial lipaseswere obtained from Biocatalysts Ltd. (Pontypridd, UnitedKingdom): Candida rugosa (ex. cylindracea) (42,500 Ulg;water 5.0% w/w), Chromobacterium viscosum (13,300 DIg;water 5.9% w/w), Mucor miehei (7,200 Ulg; water 7.4% wlw)and P. fluoreseens (11,900 Ulg; water 3.1% w/w).

Transesterifieation. A preliminary study with those fourIipases (10 mg; 3.3% wlw) was carried out with 0.277 mmol(ea. 0.2 g) rapeseed oil and 0.680 mmol (107 ilL) of2-ethyl-1-hexanol (molar ratio of 1:3) in capped 13-mL testtubes under magnetic stirring at 200 rpm with 3.0%added water. Transesterification was allowed to continuefor 72 h, after which lipase was separated by centrifuga­tion for 5 min at 2000 rpm (Martin Christ '1YPe UJ3;Osterode, Germany), and the supernatant was pipettedinto Eppendorf tubes for storage at -20°C and lateranalysis. Further transesterification reactions were car­ried out for up to 72 h with varying substrate molar ratios(rapeseed oillethyl hexanol from 1:1 to 1:10), C rugosalipase (from 0.3 to 14.6% w/w) and added water (from 0.25to 50% w/w) quantities and temperatures (from 37-60°C).

Lipase activity. Lipase activity was determined accord­ing to the Biocatalysts Ltd. assay method "Lipase AssaY,'which is based Q.tl the hydrolysis of 50% (vol/vol) olive oilemulsion (Product No. 800-1; Sigma) as shbstrate at pH7.7, and 37°Cin one hour. The quantity of free fatty acidsformed was titrated with O.lM sodium hydroxide. Oneunit of lipase activity was defined as the quantity of

JAOCS, Vol. 71, no. 12 (December 1994)

1412

Y.·Y. LINKO ET AL.

LII

FIG. 1. Thin-layer chromatograms with different Iipases (rapeseedoilJ2-ethyl-l-hexanol substrate molar ratio 1:3; 3.5%, wlw, lipase; 3.0%,w/w, added water; reaction time 24 h). 1, Rapeseed oil; 2,2-ethyl·l-hexanol; 3, triolein; 4, diolein; 5, monoolein; 6, Candidarugosa lipase; 7, Mucor mkhei lipase; 8, Pseudomo1l48 floorescencelipase; 9, Chromobacterium viscosum lipase; 10, blank.

Consequently, the use of alcohol excess was not furtherinvestigated.

About 50% rapeseed oil conversion was reached in onehour, with a nearly complete conversion in 10 h when thesubstrate molar ratio was between 1:2.8 to 1:3.0, lipasequantity was 3.3% (w/w) and the added water 3.0%. Itcould be concluded from several replicate transesterifica­tions that the highest ester yield with the least residualalcohol was obtained with the substrate molar ratio of1:2.8, although the differences with different substratemolar ratios, down to 1:2.5, were small. Consequently, themolar ratio of 1:2.8 was used in most of the subsequenttrials.

Lipase quantity. As could be expected, an increase inlipase quantity markedly increased the rapeseed oil con·version during the first few hours, but after seven hoursthe differences had almost leveled off. Figure 2 illustrates,as an example, the rapeseed oil conversion as the functionof time with 0.3, 2.3 and 14.6% (w/w) lipase, substratemolar ratio of 1:2.8 and 3.0% (w/w) added water, The reac­tion was nearly complete (the maximum theoretical con·version under the conditions used is 93.3%) in one hourwith the highest lipase quantity used, whereas with only0.3% (w/w) lipase, the conversion in one hour was onlyabout 20%. Nevertheless, a nearly complete conversionwas obtained in 10 h, even with the lowest lipase quantityused which, in addition to the stability of the enzyme, isimportant in considering costs. Interestingly, Goldberget aL (29) reported that an increase in the quantity ofpowdered C rugosa lipase results in a decrease in the ap­parent enzyme activity in the production of heptyl oleate,owing to an increase in diffusion limitation, a problemwhich may be minimized in large-scale experiments by op­timal biocatalyst and reactor design.

enzyme that catalyzes the release of one ,.,mole of freefatty acid from olive oil in one minute under thoseconditions.

Analytical methods. Qualitative analyses were carriedout by thin-layer chromatography (TLC). Samples werediluted 1:10 (vol/vol) with ethanol, and 0.01 mL of thediluted samples were used for TLC analysis. Hexane/di­ethyl ether/acetic acid (80:20:1) was used as solvent onKieselguhr 60 F 254 plates (E. Merck. Darmstadt.Germany) with one hour running time. Slightly driedplates were sprayed with 0.1% 2',7'-dichlorofluorescein(Aldrich-Chemie, Steinheim, Germany) in 99.5% ethanol(Alko Ltd., Rajamaki. Finland) for detecting the spots at254 and 360 nm.

Rapeseed oil conversion (% rapeseed oil used) and esteryield (% of theoretical) were determined by reversed-phasehigh-performance liquid chromatography (HPLC), asmodified from EI-Hamdy and Perkins (28) and Forssell etaL (3), with a Perkin-Elmer (Norwalk, CT) 4 pump module,ISS-lOO sampler, and 101 oven, Novapack C18 3.9 X 150mm column with 4,.,m silica particles. HP 1047A refrac­tive index detector (Hewlett-Packard, Palo Alto, CAl, PE316 integrator and PE 7500 professional computer.Samples were diluted with acetone to 10-20 mg/mL,filtered through a Millex-LCR4 disposable filter with 0.5,.,m porosity (Millipore, Bedford, United Kingdom), and0.02 mL of the filtrate was used for the analysis. The run­ning solvent was acetone/acetonitrile (1:1) at 1.0 mL/min,37°C, 30 min. Residual 2-ethyl-1-hexanol could not bedetermined by the HPLC method because the alcoholoverlapped with the acetone used as the diluent. Conse­quently, any excess 2-ethyl+hexanol was determined byTLC as described above. Moisture content of the enzymepreparations was determined by drying about 4-g samplesovernight at 105°C.

RESULTS AND DISCUSSION

Lipase. Th identify the most suitable enzyme for subse­quent transesterification trials, preliminary experimentswere carried out with the most promising commerciallipases of a total of 25 previously screened for n-butyloleate biosynthesis (5). A substrate molar ratio of 1:3, 3.3%(w/w) of lipase and 3.0% (w/w) of added water were used.Figure 1 shows that in 24 h the use of C rugosa lipaseas biocatalyst resulted in the highest ester production,with no detectable residual rapeseed oil and little by­product. A 98% conversion of rapeseed oil was obtainedin 24 h. Also, a superior cost/benefit ratio has been re­ported previously for this lipase in the direct biocatalyticsynthesis of n-butyl oleate (4). The use of P. fluorescensand C viscosum lipases also resulted in relatively highester production, with 96% conversion in 24 h and 97%conversion or higher in 48 h, although clearly more resid­ual alcohol and by-products could also be seen. The poor­est results were obtained with M miehei lipase. Conse­quently, C rugosa lipase was chosen for further studies.

Substrate molar ratio. One of the aims was to obtaina maximum rapeseed oil conversion with no or littleresidual 2-ethyl+hexanol. When an alcohol excess wasused, rapeseed oil conversion was always low, and the pro­duct mixture contained large quantities of residual alcoholand, in some cases, residual oil. The relative ester yielddecreased with an increase in the alcohol molar excess.

JAOCS, Vol. 71, no. 12 (December 1994)

1 2 3 5 6 7 8 9 10

3/11413

LIPASE-CATALYZED TRANSESTERIFICATION

FIG. 2. Effect of Candida cylindracea lipase (.,0.3%; 1:0., 2.3%; v,14.6%; w/wl quantity on transesterification (substrate molar ratio1:2.8; 3.0%, wlw added water).

jJII1

3 4 5 6

Time (h)

FIG. 4. Effect of temperature (., 37°C; 0, 45°C; ~,50°C; 1:0., 55°C;., 60°C) on tr&nsesterification (substrate molar ratio 1:2.8; 3.3%, wIw,lipase; 3.0%, w/w added water).

100

I

1 80

I ""J :::

I c 60

iII "

I;> 40:::cU

I 20I

I.---.J

6 24

Time (h)

o

Added water. The importance of the control of watercontent (and of water activity) in lipase-catalyzed esterifi­cations has been frequently emphasized. Figure 3 showsthe effect of added water on transesterification when thesubstrate molar ratio was 1;2.8 and the lipase quantitywas 3.3% (wlw). No phase separation took place with upto 5% (wfw) of added water. At higher water quantities,the organic phase was used for HPLC analyses. The water(ca. 5% wfw of lipase) present in the lipase preparation wasinsufficient, and only about 25% conversion was reachedin seven hours without added water. With 0.25% (wfw) ofadded water, about 60% conversion was reached in onehour, but an increase in time did not bring about a fur­ther increase in rapeseed oil conversion. With a minimumof about 1.0% added water, about 50% conversion wasreached in one hour, and a nearly complete conversion infive hours. Little difference was observed between 1.0 and

3.0% added water. Additional increases in added water didnot result in further improvements. Interestingly, the reac­tion proceeded nearly identically with 50% (wlw) addedwater as with 3.0%. Although direct lipase-catalyzed estersynthesis may not be directly compared with transesteri­fication, it is of interest to note that 90% or higher butyloleate yields have been reported in the presence of excesswater (6,30).

Temperature. Figure 4 illustrates the effect of tempera­ture on the time course of rapeseed oil conversion. Therewas little difference within the temperature range of 37­55°C. with about 90% conversion reached in 2-3 h andnearly a complete conversion in 7 h. However, at 60°Clipase was clearly inactivated under the experimental con­ditions. The results agreed well with those of Mittelbach(25), according to whom the optimal temperature for Can­dida sp. lipase-catalyzed sunflower oil alcoholysis is45-50°C. Hirata et aL (22) also reported 50°C as the op­timal temperature for the transesterification of tributyrinand l-octanol with C rugosa lipase.

80 -

::: 60c

'" 40 -

c

U ':.V.-----.-----.J067

Time (h)

FIG. 3. Effect of added water (.,0%; 0, 0.25%; .,3.0%; 1:0., 50%)on transesterification (substrate molar ratio 1:2.8; 3.3%; wlw, Iipasel.

REFERENCES

1. Macrae. A.R. and RC Hammond, BiotechnoL Genetit: Eng. &v.3:193 (1985).

2. Bjorkling, F.. S.E. Godfredsen and O. Kirk, TI-ends BiotechnoL9:360 (1991).

3. Forssell, P., R Kervinen, M. Lappi, P. Linko, T. Suortti and K.Poutanen, J. Am. Oil Chem. Soc. 69:126 (1992l.

4. Forssell, P., P. Parowon, P. Linko and K. Poutanen, Ibid. 70:1105(1993).

5. Linko, Y.){.. U.·M. Koivisto and H. Kautola,Ann. New YorkAcad.Sci. 613:691 (l990l.

6. Linko, Y.::'l., O. Rantanen, H.-C Yu and P. Linko, in Biocatalysisin Non-Conventional MedUl, edited by J. 'Iramper, M.H. Venniie.RH. Beefink and U. von Stockar, Elsevier, Amsterdam, 1992,pp. 601-608. .

7. Linko, Y.::'l., Z. Wang and J; Seppala, Biocatalysis 8:1 (1993).8. Linko, Y.){., z..L. WangandJ. Seppala. in Proceedings of3rdIn­

temational Workshop on Bictkgradable Plastit:s and Polymers,Osaka, November 9-11, 1993, 570-576.

9. Linko, Y.::'l., and H.-C. Yu, Ann. New York Acad. Sci. 672:492(1992).

JAOCS, Vol. 71, no. 12 (December 1994)

1414

YN. LINKO ET AL.

4/1

10. Yamane, T., J. Am. Oil Chern. Soc. 64:1657 (1987).11. Malcata. F.X., H.R. Reyes, H.S. Garcia, C.G. Hill, Jr. and C.H.

Amundson, Enzyme Microb. TechnoL 14:426 (1992).12. Macrae, A.R., in Biccatalysis in Organic Syntheses, edited by

J. 'framper, RC. van der Plas and P. Linko, Elsevier SciencePublishers, Amsterdam, 1985, pp. 195-208.

13. Zaks, A.. and A.M. Klibanov, Proc. NatL Acad. Sci. USA 82:3192(1985).

14. Freedman, B., R.O. Butterfield and E.H. Pryde, J. Am. Oil Chem.Soc. 63:1375 (1986).

15. Halling, P.J., BictechnoL Biceng. 35:691 (1990).16. Khmelnitsky, Y., A. Levashov, N. Klyach.ko and K Martinek, En­

zyme Microb. TechnoL 10:710 (1988).17. Ergan, F.• M. Trani and G. Andre, J. Am. Oil Chem. Soc. 68:412

(1991).18. Trani. M., F. Ergan and G. Andre, Ibid. 68:20 (1991).19. Ison. A.P., P. Dunnill and M.D. Lilly. Enzyme Microb. TechnoL

10:47 (1988).20. Zaks, A., and A.M. Klibanov, Science 224:1249 (1984).21. Halling, P.J., in Biccatalysis in Organic Media., edited by C. Laane,

J. 'framper and M.D. Lilly, Elsevier Science Publishers. Amster­dam, 1987, pp. 125-132.

JAOCS, Vol. 71, no. 12 (December 1994)

22. Hirata, H.. K. Higuchi and T. Yamashina,J. BictechnoL 14:157(1990).

23. Meffert, A., J. Am. Oil Chern. Soc. 61:255 (1984).24. Godfredsen, SE., inEnzymes in Food Processing, 3rd edn., edited

by T. Nagodawithana, and G. Reed. Academic Press, San Diego,1993, pp. 205-219.

25. Mittelbach, M., J. Am. OiL Chem. Soc. 67:168 (1990).26. Shaw, J.-F., D..L. Wang and Y.J. Wang, Enzyme Microb. TechnoL

13:544 (1991).27. Chemical Marketing &porter 243:5 (1994).28. El-Hamdy, A.H., and E.G. Perkins,J. Am. Oil Chem. Soc. 58.867

(1981).29. Goldberg, M.. D. Thomas and M.-D. Legoy, Enzyme Microb.

TechnoL 12:976 (1990).30. Nishio, T., T. Chikano and M. Kamimura, Agric. BicL Chern.

52:1203 (1988).

[Received May 14, 1994; accepted September 13, 1994]

9'1Appendix U .with the permission of the publisher

BIOTECHNOLOGY TECHNIQUESVolume 8 No.6 (June 1994) pp.451-456Received 5th May

2-ETHYL-I-HEXANOL FATTY ACID ESTERS FROMRAPESEED OIL BY TRANSESTERIFICATION

Merja Uimsa*, Anne Huhtala. Yu-Yen Linko*" and Pekka Linko

Laboratory ofBiotechnology and Food Engineering,Helsinki Universiry ofTechnology, FIN-02l50 Espoo. Finland

SUMMARY

FaIry acid eSlers of2-elhyl-I-hexanol were produced in a small pilol scale from rapeseed oil byCandida cylindracea lipase catalyzed traJlseslerijication (alcoholysis) wilhow added solven!. Up1090% conversion of rapeseed oil (97% of theoretical) was obtained in 8 h in 2 kg scale QI 37'C wilh 3.4% (w/w) lipase imnwbilized on an anion exchange resin Amberlite XAD-7. rapeseedoil:2-ethyl-I-hexanol substrale molar ratio of2.8, alld 3% (wlw) ofadded warer.

INTRODUCTION

There is an increasing interest in using vegetable oils to obtain biodegradable estersvaluable as lubricants, biodiesel, surface active agents, solvents, etc. (Meffen, 1984;Harrington and D'Arcy-Evans, 1985; Chopineau et al., 1988; van der Waa1, 1989).Special attention has been recently paid to lipase catalyzed ester biosynthesis andtransesterification (Macrae and Hammond, 1985; Yamane, 1987: Mukerhjee, 1990;Bjorkling et af., 1991; Linko et al., 1992). Nevertheless, enzymic alcoholysis of\'egetable oils without additional organic solvent has been little investigated. Lazar et af.(1985) have described the methanolysis of tallow in a bufferlhexane 2-phase system, andChoo et al. (1987) used Candida cylindracea lipase in the methanolysis of vegetable oilsin hexane with an ester yield of 98%. Miuelbach (1990) employed Pseudomonas,r7uorescens lipase both in petroleum ether and without an additional solvent for thebiosynthesis of methyl and ethly esters as diesel oil substitutes. Shaw et aJ. (1991) usedP. fluorescens lipase immobilized on Celite as a biocatalyst for the alcoholysis of oliveoil. Trani et af. (1991) used Lipozyme IM-20 (Novo Nordisk, Bagsvaerd, Denmark), aMucor miehei lipase immobilized on a weak anion exchange resin, to produce stearyloleate by transesterification of triolein and stearylalcohol, and erucylcrucate, the maincomponent of jojoba oil, by transesterification of rapeseed oil and erucyl alcohol. Theaim of the present work was to produce a blend of 2-elhyl-l-hexanol fatty acid esters byenzymic alcoholysis of rapeseed oil.

Z Present address: Raisio Group Oil Milling Industry, r.O.Box 101, FIN-21201, Raisio, Finland

z* To whom all correspondence should be addressed

451

1/11

MATERIALS AND METHODS

Materials. Refined low erucic acid rapeseed oil and synthetic rapeseed oil 2-ethyl-l­hexylester were obtained from Rasio Group (Raisio. Finalnd). The approximate fatty acidcomposition of the oil was 57% oleic acid, 22o/c linoleic acid, 12% linolenic acid, 4%palmitic acid, 1% strearic acid, 2% eicosaenic acid. < I% erucic acid, I % others. 2­Ethyl-l-hexanol (water solubility at 25°C ca. 2.5%) was obtained from Fluka ChemieAG (Buchs, Switzerland). Mono-, di- and triolein standards were from Sigma (St. Louis,MO), glycerol from May & Baker (Dagemham, UK), and powdered Candida rugosalipase (42,500 Ulg; water 5.0% w/w) was obtained from Biocatalysts Ltd (Pontypridd,UK).

Transesterijication. The reaction mixture contained 50 g to 2000 g of rapeseed oil. 25cm3 (20 g) to 1.0 dm3 (829 g) 2-ethyl-l-hexanol (suh~tratemolar ratio of 1:2.8), 3.4%(w/w) lipase, and I % to 5% (w/w) added water. Tne reaction was carried out for up 10 5hours at 37°C in a stirred vessel of varying shape either with free powdered lipase orwith enzyme immobilized on various carriers. Glass beads (20 g per 100 g rapeseed oiland 5 g lipase), polyurethane foam (0.8 g), and adsorption resins Amberlite XAD-2 (2.5g) and Amberlite XAD-7 (2.5 to 15 g) were used a~ carriers.

Lipase activity. Lipase activity was determined according 10 the Biocatalysts Ltd. assaymethod Lipase Assay, which is based on the hydrolysis oi 50% (v/v) olive oil (ProductNo. 800-1, Sigma Chemical Company, St. Louis, Missouri) as substrate at pH 7.7, 37°Cin one hour. The quantity of free fatty acids formed was titrated with O.IM sopdiumhydroxyde. One unit of lipase activity was defined as the quantity of enzyme whichcatalyses the release of one 11 mole of free fauy acid from olive oil in one minute at theabove conditions.

Analytical methods. Qualitative analyses were carried OUl by thin layer chromatography(TLC). Samples were diluted 1: 10 (v/v) with ethanol, and 0.01 ml of lhe diluted sampleswere used for TLC analysis. Hexane:diethylether:acelic acid (80:20: 1) was used assolvent on Kieselguhr 60 F254 plates (E. l\'1erck, Darm~l<ldt, Germany) wilh one hourrunning time. Slightly dried plates were sprayed Wilh 0.1 % 2',7'-dichlorofluorescein(Aldrich-Chemie, Steinhein1, Germany) in 99.5'* ethanol (Alko Ltd, Finland) fordetecting of the spots at 254 and 360 nm.

Rape seed oil conversion (% rapeseed oil used) and ester yield (% of theoretical) weredetermined using reversed phase high performance liquid chromatography (HPLC) asmodified from El-Hamdy and Perkins (22) and Forsell et a1. (1), employing a Perkin­Elmer (Norwalk, Connecticut) 4 pump module, ISS-lOa sampler, and 101 oven,Novapack C18 3.9 x 150 mm column with 4 11m silica pJ.nicles, HP 1047A refractiveindex detector (WaterslMillipore, Milford. UK) PE 316 integratOr and PE 7500professional computer. Samples were diluted Wilh acetone to 10 to 20 mgiml, filteredthrough a Millex-LCR4 disposable iilter 0.5 l-lm porosity (Millipore, Bedford, UK), and0.02 ml of the filtrate was used for the analysis. The running solvent wasacetone:acetonitrile (1:1) at 1.0 ml/min, 37°C, 30 min. Residual 2-ethyl-l-hexanOl couldnot be determined by the HPLC method inasmuch as the alcohol overlapped withacetone used as the diluent. Consequently, any excess 2-ethyl-I-hexanol was determinedby TLC as described above.

Moisture content of the enzyme preparations was determined by drying about 4 gsamples over night at 105°C.

452

2/11

3/11

RESULTS AND DISCUSSION

Figure I shows the effect of different type of mixing on rapeseed oil conversion in a

series of experiments with substrate molar ratio of 1:2.8, 1.0% (w/w) added water and

3.4% (w/w) (25 g) lipase. With 100 g of oil and 50 cm3 (41 g) of 2-ethyl-I-hexanol in

500 em3 flat bottomed flask under 300 rev min-I magnetic stirring conversion increased

in a few minutes up to about 65% and in 1.5 h to 80%, with a final conversion in 5 h of

87%(94% of theoretical maximum). Within 30 min the solid lipase was 'immobilized' on

the walls, adsorbing the glycerol formed in the reaction. Ester production was 21.7 g per

g lipase and 108.4 g per 100 g rapeseed oil. In another experiment with 500 g of oil and

700 rev min- l mechanical stirring for the first 5 h, followed by 300 rev min- l for a further

5 h the final conversion was only 61 % (66% of theoretical). During the frrst 5 h lipase

remained evenly suspended, and the conversion was only 46% (50% of theoretical). The

decrease of mixing rate resulted in the aggregation of lipase at the bottom, with an

increase in conversion. This agrees well with the observation of Goldberg et af. (1990)

during studies on C. rugosa lipase catalyzed heptyl octanoate synthesis, that when the

mixing rate was increased sufficiently for complete lipase suspension, the enzyme was

inactivated. Also Lee and Chao (1989) have reponed that C. rugosa lipase is sensitive to

shear forces in a 0.1 % solution with the activity decreasong with an increase both in

mixing rate and time. The degree of lipase inactivation could he markedly reduced by

propylen glycol addition.

!..!

10642

100 r IJ ....------....._ 1

80 1/--....----.... i

r

./~60 - Q--~ I

//?,---."./,/" I40 - c./~/ 1

I /

20 1- /! /i I1/,/o • -L -'- '--__---'- ..J

o

>

Reaction lime (11)

Fig. I. Effect of mixing on transesterification {. I(}() g rapeseed oil,300 rev minot magnetic stirring; 0 500 g rapeseed oil, 700 rev min-!

mechanical stirring for 5 h, followed by 300 rev min- l for 5

453

Figure 2 shows as a result of an experiment with 200 g rapeseed oil under otherwise

similar conditions that with mechanical stirring at a 300 rev min-I from the beginning,

conversion increased to about 73% (78% of theoretical) in 30 min, without a further .

increase. As much as 89% of the product could be recover by simple decanting. Similar

results were obtained when the batch quantity of rapeseed oil was increased to 550 g.

Only a slight increase in conversion was observed when the flask was frlled to minimized

the air/liquid boundary as suggested by Lee and Choo (1989). When the lipase was

reused afer decanting the product, conversion was only about 3%.

100 r-----,r-----,r---,---.,.----,

tf2. 80 !

~--~~ 60V ~'" r !

'" 40 r~ ~~ io w J

u 0 .=====r=::=:-t~--lo J 2 3 5

Reaction time (h)

Fig. 2. Effect of reusing the biocatalyst (. 200 g rapeseed oil,300 rev min-l mechanical stiffing, 1st use: II 2nd usc).

Because the preliminary experiments were suggested [hat lipase immobilization might be

of advantage in increasing rapeseed oil conversion. a numher of carriers were tested. No

improvement was obtained with glass beads, polyurethane foam or Amberlite XAD-2

resin immobilized lipase (Figure 3). Figure 4 shows the effeCt of increa<;ing the relative

quantity the Amberlite XAD-7 resin as the carrier on conversion, with the highest

rapeseed oil conversion (50 g of oil) of about 95'X (100<;1 of theoretical) obtained in 5 h

with 7.5 g of resin per 2.5 g of ]ipa<;e. Figure 5 ckarly illustrates the imponance of the

quantity of added water on rapeseed oil conversion using Amberlite XAD-7 immobilized

lipase, with best result obtained with 3% added water. Finally, the optimal conditions

found were tested with a 2 kg batch of rapeseed oil (I dm3, 829 g, of 2-cthyl-I-hexanol,

3% added water, 100 g lipase with 300 g XAD-7 resin. 170 rev min'], 37°C)(Fig. 6).

The result was an about 90% conversion (97% of theoretical) in 8 h, equal to 22.4 g of

ester per g lipase or 112 g ester per 100 g rapeseed oil. Although the total reaction time

increased with the immobilized biocatalyst, final conversion could be increased at least by

about 20% to nearly theoretical.

454

4/11.

95

. ] ,':r '~3"J ~ ",; 1

0" O. Io 1 2 0 1 2

Reaction time (lJ) Reaction time (lJ)

5/11

80

o 60

Fig. 3. Transesteriiication with (e) glassbeads, (~) polyurethane. and (0)

Amberlite XAD-2 resin as camer.

100 r--~---""----'---"---"

~.~ /f

I /~II ~~ i, ~. 1

~401/ ~~ .;~ 20ly~~

It.........o~---~,--~--~-~

o J 2

Reaction time (Ii)

Fig. 5. Effect oi added w~ler Ie 1'i<:. 0

2%, ~ 3% w/w) on lranSeslerillcalionwith Arnberlite XAD-7 resin as carrier.

455

Fig. 4. Transesteriiication withAmber-lite XAD-7 immobilized lipase(e 3.75 g, /). 5.0 g and 0 7.5 g lipase)

Reaction time (ll)

Fig. 6. Time course of lransesteriiictionof 2 kg rapeseed oil with AmberlileXAD-7 resin immobilized lipase (insen:BPLC analysis of product as the iunc­li,m of lime; l, 5 min; 2, 2 h; 2h: 3, 11:

4.611; 5, R h)

CONCLUSION

It has been demonstaned that fatty acid esters of 2-ethyl-l-hexanol can be produced in a

small pilot scale from rapeseed oil in near 100% conversion by Canditk1 cylindracea

lipase catalyzed transesterification (alcoholysis) without added so]venL Best results were

obtained with about 3% (w/w) added water and a slight oil excess under relatively mild

mixing with lipase immobilized on hydrophobic anion exchange resin Amberlite XAD-7.

REFERENCES

BjorkIing, E, Godfredsen, S. E and Kirk, O. (1991). Trends Biotechnol. 3, 360-363.

Choo, Y. M., Ong, S. H., Goh, S. H. and Khor, H. T. (1987) Brit. UK Pat. Appl. GB 2.188,057.

Chopineau, J., McCaffeny, F. D., Therisod, M. and KJibanov, A. M. (1988). Biolechnol.Bioeng. 31, 208-214.

Goldberg" M., Thomas, D. and Legoy, M.-D. (1990) Enzyme Microb. Teelmol. 12,976-98 I.

Harrington, K. J. and D'Arey-Evans, C. (1985). J. Am. Oil Chem. Soc. 62, 1009- 1013.

Lazar, G. (1985). Fel/e, Seifen, Anstrichm. 87: 394-400.

Lee, Y.-K. and Choo, c.-L. (1989) Biotedmo!. Bioeng. 33,183-190.

Unko, Y.-Y., Rantanen, 0., Yu, H.-C. and Unko, P. (1992). Factors affecting lipase catalyzedn-butyl oleate synthesis. In Biocataiysis in Non-Conventionai Media, J. Tramper, M. H.Vermiie, H. H. Beefink and U. von Stockar, eds. Elsevier, Amsterdam

Macrae, A. R. and Hammond, R. C. (1985). BiOlechnoL. Genetic eng. Rev. 3, 193-217.

Meffert, A. (1984).1. Am. Oil Chem. Soc. 61,255-258.

Mittelbaeh, m. (1990). J. Am. Oil Chem. Soc. 67,168-170.

Mukerhjee, K. D. (1990). Biocatalysis 3, 277-293.

Trani, M., Ergan, E and Andre, G. (1991) J. Am. Oil Chem. Soc. 68,20-22.

Yamane, T. (1987). J. AnL Oil Chem. Soc. (A, 1567-1662.

van der Waal, B. (1989). NLGI Spokesman 53(8), 359-368.

456

6/11

Appendix III

with the permission of the publisher

Process for preparing a synthetic ester from a vegetable oil

97

95395

1/111

The objects of the present invention are a process for preparing a synthetic ester

from a vegetable oil and lubricants which contain a synthetic ester prepared by

said process.

Natural fats and oils have been used as lubricants already for thousands of years.

With industrialization mineral based lubricants came also to the market. The app­

lications of lubricants and thus also the requirements set for them have changed,

and developed with the advance of technology. Various types of synthetic esters

and lubricants containing the same have been developed to meet the new require­

ments.

The purpose of a lubricant is to minimize friction and wearing of metals. Lubri­

cants are developed according to the use and they consist of a base fluid and

additives improving the lubricative properties. With the development of technolo­

gy, lubricants are used under more and more severe conditions, such as at very

low or very high temperatures (e.g. the turbine engines of aeroplanes). At the

same time biodegradability, unburden to the environment, non-toxicity and the

use of renewable raw materials have emerged as new requirements. The use of

biodegradable lubricants is of particular importance in the machines and devices

used in the fields of agriculture, forestry and building, as the oil used may be left

in the environment.

By the synthetic esters developed as lubricants are meant esters prepared from

mono-, di- or trialcohols and mono- or dicarboxylic acids by known esterification

and transesterification methods. Usually the process comprises combining all the

reactants and letting the reaction happen in one stage. The reaction may be car­

ried out in the presence of catalysts, such as acids, bases or metal oxides.

The structure of the synthetic ester used has a profound effect on the stability of

the lubricant. Esters decompose by the effect of heat and/or oxygen. It is known

to increase the thermal stability of synthetic esters by using in the preparation no

953952

beta hydrogen alcohols. Oxidative properties on the other hand can be improved

by deuteration of esters.

Synthetic esters intended for a lubricative use are classified by structure as mono­

carboxylic acid, dicarboxylic acid, polyol and complex esters. Due to their low

viscosity and high volatility monoesters are poorly suitable as lubricants. Polyol

esters are chemically more stable than for example diesters due to the structure of

the polyols used in the preparation of said esters, wherein no hydrogen atom is at­

tached to the {3 carbon atom. Complex esters have promising lubricative proper­

ties but the manufacture thereof on an industrial scale is difficult because of the

severe conditions required by the reaction, especially if said esters are prepared ­

from purified fatty acids and alcohols.

If polyol esters are prepared by using no alfa hydrogen acids, the stability proper­

ties of the esters can be further improved. Metro et al. (CA 859 771) have shown

that the no alfa hydrogen carboxylic acids increase the thermal and oxidative

stability of esters prepared from no beta hydrogen alcohols, as well as slow down

the hydrolysis of the esters.

As the low viscosity polyol esters are not suitable for traditional uses wherein

high viscosity is required, it has been aimed at preparing polyol esters of higher

viscosity from for example trimethylol propane (TMP). However, it has been

found that it is difficult to obtain simple TMP esters with both high viscosity and

a low pour point (cf. for example US 4,061,581).

Products based on vegetable oils are nowadays used more and more as lubricants

because of their safety to the environment. Natural vegetable and animal oils are

glyceride diesters, i.e. tri-, di- or monoesters of glycerol and straight chain satu­

rated and unsaturated fatty acids. The lubricant industry uses for instance ra­

peseed, rape, soybean, castor, olive, coconut, palm and pine oils.

21111

9f

953953

The advantageous properties of vegetable oils include user friendliness and non­

toxicity. In addition to this they degrade in the environment, do not accumulate in

the food chain of nature and are renewable raw materials. However, the use of

vegetable oils as lubricants has been limited by their poor stability properties. The

poor thermal and oxidative stability is due to unsaturated and polyunsaturated

fatty acids. On the other hand, the unsatisfactory behaviour of vegetable oils at

low temperatures is due to the saturated fraction of fatty acids. By using suitable

additives and by favouring in cultivation such varieties which do not have a too

high degree of saturation, it has been possible to somewhat improve the stability

properties. Also the purification of the oil for technical use is helpful.

Furthermore, attempts have been made to modify natural glyceride esters in order·

to improve their stability properties. Known processes include catalytic hydroge­

nation, alcoholysis, geometrical isomerization and sulfurization. For example by

hydrogenation a certain amount of double bonds can be removed, from the unsa­

turated part of vegetable oils and by isomerization the amount of undesired iso­

mers can be decreased.

Van der Waal and Kenbeek have presented a process for the preparation of synt­

hetic esters from vegetable oils or animal fats (proceedings of the Tribology

2000, 8th International Colloqium, Technische Akademie Esslingen, Germany,

14-16 June 1992, Vol II, pp 13.3-1 - 13.3-8). The process comprises first decom­

posing the glyceride esters of the starting material into fatty acids and glycerol

and subsequently separating the fatty acid fraction into liquid and solid phases.

The fatty acids of the liquid phase are separated by distillation into single fatty

acids which can be further modified e.g. by hydrogenation or cracking to obtain

the desired raw material. Fractions containing a single fatty acid are esterified

with no beta hydrogen polyols for preparing a synthetic ester.

The fatty acids of the ester prepared according to the above described process

usually contain less unsaturated double bonds than the fatty acids of the starting

material, which improves the oxidative stability. However, the costs of the pro-

3nll

4 95395

4/111

cess are extremely high, due to the multistage separation and purification reac­

tions and the most severe conditions (high pressure and temperature) required by

the reaction. Moreover, it has been found that when fractions containing only a

single fatty acid are reacted with polyols, plenty of mono- and diglycerides are

formed, i.e. all the OH groups of the polyols do not react. This decreases the

triglyceride yield and the raw material has to be recycled several times if the

yield is to be improved. Furthermore, the reaction of a fatty acid and an alcohol

creates water which has to be removed during the reaction.

According to the invention it has now been found that it is possible to prepare

synthetic esters with good lubricative properties from vegetable oils by a process

which avoids the multistage reaction with several separations and recyclings and

by which good yields are obtained.

In the process according to the invention a vegetable oil is first transesterified by

reacting the vegetable oil with a lower alkanol to obtain a mixture of fatty acid

lower alkyl esters. The process is characterised in that the obtained mixture of

esters is further transesterified by reacting the said mixture with a no beta hydro­

gen polyol of the formula

CH20H/

R - C - CH20H\CHzOH

wherein R is a C1-C6 alkyl group, particularly a C1-C4 alkyl group, or a -CH20H

group, and the synthetic ester obtained is recovered.

Vegetable oils suitable as a starting material of the process are for example ra­

peseed, rape, soybean, castor, olive, coconut, palm, pine, maize, walnut, flax­

seed, cotton, sunflower, sesame and almond oils, especially rapeseed oil, rape oil,

pine oil and soybean oil, particularly rapeseed oil.

5 95395

snll.

The fIrst transesterifIcation reaction of the process according to the invention is

carried out by a process known per se, by reacting a refIned or alkalirefIned

vegetable oil with a lower alkanol to obtain a mixture of fatty acid lower alkyl

esters.

The lower alkanol used in the fIrst transesterifIcation reaction is preferably a C I ­

C4 alkanol, especially methanol or ethanol. The obtained mixture of lower alkyl

esters of the vegetable oil is thus preferably a mixture of CI-C4 alkyl esters,

especially a mixture of methyl or ethyl esters. If desiIed, usual esterifIcation

catalysts may be used in the reaction, and the amounts of the reactants and the

reaction conditions (pressure, temperature, reaction time) are either commonly

known or easily chosen by a person skilled in the art.

The fIrst transesterification reaction may be illustrated by the following general

reaction scheme I:

H2C-O-C( =O)-RII

Hr-O-C(=O)-R2 + 3 ~-OH <=± 3 ~-O-C(=O)-Rx +H2C-O-C(=O)-R3

(I)

wherein RI , R2 and R3 are fatty acid residues, ~ is an alkyl residue, especially a

CI-C4 alkyl residue, and Rx is RI , R2 or R3• Glycerol is formed as a by-product.

The fatty acid lower alkyl ester obtained from the fIrst transesterification reaction

is thus a mixture comprising various fatty acids of the vegetable oil used as the

starting material. It is typical of the invention that this mixture of fatty acid lower

alkyl esters may be used directly as the starting material of the second transesteri­

fication reaction without separation or purification of fatty acids.

In the second transesterifIcation reaction according to the invention the mixture of

fatty acid lower alkyl esters obtained from the first transesterification reaction is

reacted with a no beta hydrogen polyol, such as for example trimethylol ethane,

6 95395

6/111

trimethylol propane, trimethylol butane or pentaerythritol, especially with penta­

erythritol or trimethylolpropane. The conditions required by the reaction are not

so severe than those required by the process according to the prior art, and the

by-products formed may be present in the reaction.

The second transesterification reaction may be illustrated with the following gene­

ral reaction scheme II:

/CH20H

3 ~-O-C(=O)-R + R-C - CH20H --+

x " CHpH

+ 3 ~-OH

wherein RI , R2, R3, ~ and Rx have the same meanings as in the reaction scheme

I and R is a CI-C6 alkyl group, especially a CC C4 alkyl group, or a -CH20H

group.

Consequently, it is the question of a totally different chemical reaction than in the

process of the prior art wherein a free fatty acid is esterified with an alcohol. In

the process according to the invention, an ester is reacted with an alcohol, and

thus it is the question of a transesterification reaction which reaction, as well as

the reaction conditions required by it and the by-products formed therein, is

totally different from the reaction used in the process of the prior art.

The synthetic ester obtained from the second transesterification reaction is reco­

vered and, if desired, purified by conventional methods. for example by neutrali­

zation and washing with an aqueous acid. No distillation or any other special

treatment is needed as the obtained ester is ready to use as such as a raw material

of lubricants.

When a polyol is reacted with a mixture of fatty acid lower alkyl esters, almost

all OH-groups of the potyol react into trigtycerides. Over 90 % of the theoretical

yield of the triglyceride is obtained, the proportion of mono- and diglycerides

7

/03

95395

7/111

being in total about 10 %. The product obtained does not contain any free fatty

acids which makes it an especially advantageous raw material for lubricants whe­

rein the oxygenation of free fatty acids would cause problems (corrosion, cliange

of viscosity). The process is well adapted for industrial scale and the synthetic

ester obtained has better stability properties than the vegetable oil used as the raw

material, while at the same time the advantageous properties of a vegetable oil

(biodegradability, non-toxicity, user friendliness) are maintained. By the process

it is thus possible to prepare synthetic esters from vegetable oils, for example

from rapeseed oil, in a yield of even over 90 % of the theoretical.

According to the invention the second transesterification reaction may preferably

be carried out in two stages, the reaction temperature of the first stage being from

about 50 to about 110 °e and of the second from about 110 to about 160 °e.Preferably the reaction temperature of the first stage is from 85 to 100 °C and of

the second stage from 110 to 140 °e. Reaction time may vary for example from

two to twelve hours. Preferably the reaction time of the first stage is about 1 to 7

hours and of the second stage about 1 to 10 hours.

The no beta hydrogen polyol and the mixture of esters are preferably reacted with

each other in a molar ratio of about 1:2 to 1:5, especially in the molar ratio of

about 1:3,5.

The second transesterification reaction is preferably carried out under reduced

pressure, for example under negative pressure of 1.3 to 13 kPa, and optionally in

the presence of a catalyst. As catalysts known esterification catalysts, such as acid

and base catalysts, for example p-toluene sulfonic acid, phosphoric acid, sodium

hydroxide. sodium ethoxide and sodium methoxide, of which sodium hydroxide

and sodium methoxide are especially advantageous may be mentioned.

The synthetic ester prepared by the process according to the invention is an excel­

lent raw material for the preparation of lubricants. Lubricants, especially hyd­

raulic oils, which contain a synthetic ester prepared by the process of the inventi-

8 95395

8/111

on, optionally with one more additives, are also included in the scope of the in­

vention. As additives for example oxidation inhibitors, antiwear agents, antifoam

agents, corrosion inhibitors, dispersants, viscosity index improvers and/or pour

point depressers which are generally known in the art, may be used.

Oxidation inhibitors include for example amines and phenols. As antiwear agents

and corrosion inhibitors for example phosphates or sulfonates and as antifoam

agents for example metal sulfonates, metal phenates, polyesters or silicones may

be used. Viscosity index improvers include for example polyisobutenes, styrene­

butadiene and ethene-propene-eopolymers ~hich all are thus suitable also as pour

point depressers.

In the following the invention is further described by means of examples, the pur­

pose of which is to illustrate but not to limit the invention.

Example 1. Preparation of a methyl ester of rapeseed oil

Rapeseed oil (0.3 moles) was weighed into a three-necked flask provided with a

thermometer, cooler and a stirring device. Stirring was started and methanol (2.0

moles) was added. The reaction mixture was heated to 60°C and the alkali cata­

lyst used was added (0.5 % by weight). Stirring was continued for three hours.

The progress of the reaction was followed by thin layer chromatography. The

reaction mixture was washed with an aqueous acid. The glycerol created in the

reaction mixture which as a heavier component settles on the bottom of the vessel

was separated and the product mixture was analyzed. Rapeseed oil ester content

was 97 %.

Example 2. Preparation of a synthetic ester from rapeseed methyl ester

The methyl ester of rapeseed oil (0.65 moles) was weighed into a three-necked

flask provided with a thermometer, a cooler, a stirring device and a reduced pres­

sure generator. The weighed rapeseed oil ester was heated to 50-110 °e, after

9

/a;;-

95395

9/11'

which trimethylol propane (TMP, 0.19 moles) was added in small proportions

with proper stirring. After the alcohol was well mixed, sodium hydroxide used as

a catalyst was added (0.1-1.0 % by weight of the reaction mixture). Then the

reaction mixture was heated under reduced pressure (about 8 kPa) until it started

boiling. The reduced pressure was maintained during the whole reaction. The

mixture was allowed to boil at the lower temperature (50-110°C) for 1 to 7 hours

and at the higher temperature (110-160 0c) for 1 to 10 hours. The progress of the

reaction was followed by thin layer chromatography.and quantitative IR spectrum.

At the end of the reaction the product mixture was neutralized and· washed with

an aqueous acid, filtrated and washed with water. Drying was performed with

anhydrous sodium sulfate. A liquid chromatogram and an IR spectrum were run

of the final product. The yield was 90.5 % of the theoretical.

Example 3. Preparation of an ethyl ester of soybean oil

Soybean oil (0.2 moles) was weighed into a three-necked flask provided with a

thermometer, a cooler and a stirring device. Stirring was started and ethanol (1.5

moles) was added. The reaction mixture was heated to 80°C and the alkali cata­

lyst used (0.4 % by weight) was added. Stirring was continued for two hours.

The progress of the reaction was followed by thin layer chromatography. The

reaction mixture was washed with an aqueous acid. Glycerol was separated from

the reaction mixture and the product mixture was analyzed by liquid chromato­

graphy. Soybean oil ester content was 96 %.

Example 4. Preparation of a synthetic ester from soybean oil ethyl ester

Soybean oil ethyl ester (0.7 moles) was weighed into a three-necked flask provi­

ded with a thermometer, a cooler, a stirring device and a reduced pressure gene­

rator. After the weighed ester was heated to 50-110°C, trimethyloI ethane (TME,

0.2 moles) was added in small proportions with proper stirring. When the alcohol

was well mixed, the catalyst used (sodium hydroxide, 0.1-1.0 % by weight of the

reaction mixture) was added. Then the reaction mixture was heated under reduced

10 95395

10/1/1.

pressure (about 8 kPa) until it started boiling. The reduced pressure was main­

tained during the whole reaction. The mixture was allowed to boil at the lower

temperature (50-110°C for I to 7 hours and at the higher temperature (110-160

0c) for 1 to 10 hours. The progress of the reaction was followed by thin layer

chromatography and quantitative IR spectrophotometry. At the end of the reaction

the product mixture was neutralized and washed with an aqueous acid, fI1trated

and washed with water. Drying was performed with sodium sulphate. A liquid

chromatogram and an IR spectrum were run from the final product. The yield

was 92 % of the theoretical yield.

Example 5. Preparation of a methyl ester of pine oil

Pine oil (0.3 moles) was weighed into a three-necked flask provided with a ther­

mometer, water separator and a cooler and a stirring device. Stirring was started

and methanol (2.0 moles) was added. The reaction mixture was heated to 60°C

and the acid catalyst used (0.3 % by weight) was added. Stirring was continued

for six hours. The progress of the reaction was followed by thin layer chromato­

graphyand by the amount of water created. The reaction mixture was washed

with alkaline water and dried with sodium sulphate. The mixture was fI1trated and

analyzed by liquid chromatography. Pine oil ester content was 97 %.

Example 6. Preparation of a hydraulic oil from a synthetic rapeseed oil ester and

comparison of hydraulic oils

The raw material used was the synthetic rapeseed oil ester obtained in Example 2.

Said ester was mixed at a certain temperature with additives to obtain a hydraulic

oil having the following composition:

The synthetic ester from Example 2

Oxidation inhibitor

Pour point depresser

Antiwear agent

90 - 98 % by weight

0.1 - 2.5 % by weight

o-5.0 % by weight

0.1 - 2.0 % by weight

Antifoam agent

11

/07

95395o- 0.5 % by weight

11/111

The technical properties studied of this ester containing additives were wearing,

friction, oxidation, low temperature properties and corrosion.

Wearing and friction were studied with a four ball test (ASTMD 2783, IF 239)

wherein wearing with respect to loading or the extreme loading where the lubrica­

tion still works, are measured. Oxidative properties were studied with an oxygen

bomb test (ASTMD 925) and with the oxidation test DIN 51586 where the change

of viscosity at 40°C was monitored. In a corrosion test (Cincinnati-Milacron test)

the aging of the oil as well as copper and steel corrosion were studied. In said

test, the change of the total acid number (TAN) and viscosity, the weight change

of the copper and steel rods used as oxidation catalysts in the test procedure and

the formation of a precipitate under the test conditions are measured. Furthermo­

re, the pour point which illustrates the low temperature properties of an oil was

analyzed, i.e. the temperature where the oil is still fluid.

The corresponding properties were examined also from hydraulic oils based on

rapeseed oil and hydraulic oils based on commercial synthetic esters. All the hyd­

raulic oils were supplemented with the same additives as the hydraulic oil based

on the ester prepared by the process of the invention. The results are shown in

Table 1.

12

95395-Table 1. Comparison of the properties of hydraulic oils. A = hydraulic oil withthe ester prepared by the process of the invention as raw material, viscosity grade32; B1 and B2 = hydraulic oils with commercial synthetic ester as raw materials,viscosity grades 46 and 68; C = commercial hydraulic oil based on rapeseed oil,viscosity grade 32.

A B1 B2 C

Four ball test- extreme loading, N 2000 3000 2500 2000- wearing, mm 0,4 0,46 0,41 0,64

Oxygen bomb test 42 39 29 30ASTDM D445, psi

Oxidation inhibition testDIN 51586, viscosity change, 12,4 20,3 24,1 28,8%

Cincinnati-Milacron test- TAN mg KOH/g

before 1,39 1,39 1,40 1,72after 1,56 3,71 2,41 0,61TAN 0,17 2,32 1,01 I,ll

- viscosity change, % 19,1 16,9 6,2 8,2- total precipitate, mg/100 ml 1,1 17,0 28,8 4,4- weight change of Cu rod, mg 1,7 -16,9 ° -0,5- weight change of steel rod, -0,3 0,4 1,2 -0,5mg

Pour point, •C -41 -36 -39 -39

From the results it can be seen that as regards oxidative and low temperature

properties, the hydraulic oil based on the ester prepared by the process according

to the invention is better than the commercial hydraulic oil based on rapeseed oil.

The ester prepared by the process according to the invention and the correspon­

ding commercial esters are equal with respect to oxidative and low temperature

properties. From the Cincinnati-Milacron test it can be seen that the change of

total acid number (TAN) is clearly lowest with the ester of the invention. The

increase in viscosity at 40°C is almost of the same order with all, as well as the

12/11\

13

13/111

weight change of copper and steel rods. Thus no corrosion is observed with any

of the tested hydraulic oils under the test conditions used. The results of the

oxygen bomb test are equal, as well as the results of the test according to DIN

51586 and the four ball test. Thus the wearing and friction properties are equally

good.

14

Claims

953951. A process for preparing a synthetic ester from a vegetable oil, comprising

- transesterification of said vegetable oil by reacting it with a

lower alkanol to form a mixture of lower alkyl esters of fatty acids,

- a second transesterification reaction wherein the obtained

mixture of esters is reacted with a no beta hydrogen polyol of the formula

CHzOH,/

R - C - CHzOH

"CHzOH

wherein R is a CCC6 alkyl group, particularly a C1-C4 alkyl group, or a -CHzOH

group, and

- recovering the synthetic ester obtained.

2. The process according to claim 1, wherein the vegetable oil is selected from

the group consisting of rapeseed oil, rape oil, pine oil and soybean oil.

3. The process according to claim 1, wherein the lower alkanol is a C1-C4 al­

kanol, especially methanol or ethanol.

4. The process according to claim 1, wherein the fatty acid lower alkyl ester is a

methyl ester of a fatty acid.

5. The process according to claim I, wherein the no beta hydrogen alcohol is

selected from the group consisting of trimethylol ethane, trimethylol propane,

trimethylol butane and pentaerythritol.

6. The process according to claim I, wherein the second transesterification reac­

tion is carried out under reduced pressure in the presence of a catalyst.

14nll

15

III

953957. The process according to claim 1, wherein the second transesterification is

carried out in two stages, with a reaction temperature of from 50 to 110 °c in the

first and with a reaction temperature of from 110 to 160 °C in the second stage.

8. The process according to claim 1, wherein the no beta hydrogen polyo1 and the

mixture of esters are reacted with each other in a molar ratio of from about 1:2 to

about 1:5, and especially in the molar ratio of about 1:3,5.

9. The use of a synthetic ester obtained according to claim 1 for the preparation

of lubricants, especially for the preparation of a hydraUlic oil.

10. A lubricant which comprises a synthetic ester obtained according to anyone

of the claims 1 to 8, optionally with one or more additives.

11. The lubricant according to claim 10 which comprises about 90 to 98 % of a

synthetic ester and about 2 to 10 % of additives.

12. The lubricant according to claim 10 wherein the additive(s) is (are) an oxida­

tion inhibitor, an antiwear agent, an antifoam agent, a corrosion inhibitor, a

dispersant, a viscosity index improver and/or pour point depresser.

16

Abstract

95395The object of the invention is a process for preparing a synthetic ester from a

vegetable oil by a two-stage transesterification process. Further objects of the in­

vention are lubricants containing a synthetic ester prepared by the process accor­

ding to the invention.

161111·

1/3Appendix IV

with the permission of the publisher

95367An enzymatic process for preparing a synthetic ester from a vegetable oil

The objects of the present invention are a process for preparing a synthetic ester

from a vegetable oil by means of lipase enzymes, and lubricants which contain a

synthetic ester prepared by said process.

Natural fats and oils have been used as lubricants already for thousands of years.

With industrialization mineral based lubricants carne also to the market. The app­

lications of lubricants and thus also the requirements set for them have changed

and developed with the advance of technology. Various types of synthetic esters

and lubricants containing the same have been developed to meet the new require­

ments.

The purpose of a lubricant is to minimize friction and wearing of metals. Lubri­

cants are developed according to the use and they consist of a base fluid and

additives improving the lubricative properties. With the development of technolo­

gy, lubricants are used under more and more severe conditions, such as at very

low or very high temperatures (e.g. the turbine engines of aeroplanes). At the

same time biodegradability, unburden to the environment, non-toxicity and the

use of renewable raw materials have emerged as new requirements. The use of

biodegradable lubricants is of particular importance in the machines and devices

used in the fields of agriculture, forestry and building, as the oil used may be left

in the environment.

By the synthetic esters developed as lubricants are meant esters prepared from

mono-, di- or trialcohols and mono- or dicarboxylic acids by known esterification

and transesterification methods. The conventional chemical process comprises

combining all the reactants and letting them react in one stage. The reaction may

be carried out in the presence of catalysts, such as acids, bases or metal oxides.

In addition to chemical agents, also lipase enzymes can act as catalysts of transes­

terification reactions.

111V

2 95367Lipases (triacylglycerol acylhydrolase; EC 3.1.1.3) belong to the esterase enzyme

group, and fats and oils are their natural substrates. Several microbes (yeasts,

molds, bacteria) secrete in their growth media lipases by means of which lipids

decompose into nutrients of the microbe. Lipases catalyze the hydrolysis reactions

of oils and fats but under suitable conditions they also catalyze the synthesis and

transesterification of tri-, di- and monoglyceride esters (Yamane et al., J. Am.

Oil Chern. Soc. 64, 1987, 1657-1662).

On the basis of their specificity, lipases are divided into three groups, nonspeci­

fic, 1,3-specific and fatty acid specific lipases. Nonspecific 1ipases are produced

by for instance the yeast Candida rugosa (ex. cylindracae) and the bacteria Cory­

nebacterium acnes and Staphylococcus aureus. Nonspecific lipases release fatty

acids from all three positions of a triglyceride. According to their name, 1,3-li­

pases release fatty acids from positions 1 and 3 of trig1ycerides. These lipases are

produced by for instance the molds Aspergillus niger, Mucor javanicus, Mucor

miehei and Rhizopus arrhizus as well as by the yeast Candida lipolytica. The fatty

acid specific lipases release only certain fatty acids from triglycerides. Mucor

miehei, for example, produces also a lipase which in addition to 1,3-specificity is

also specific to fatty acids with 12 carbon atoms. However, the specificity is not

absolute.

The structure of the synthetic ester used has a profound effect on the stability of

the lubricant. Esters decompose by the effect of heat and/or oxygen. It is known

to increase the thermal stability of synthetic esters by using in the preparation no

beta hydrogen alcohols. Oxidative properties on the other hand can be improved

by deuteration of esters.

Synthetic esters intended for a lubricative use are classified by structure as mono­

carboxylic acid, dicarboxylic acid, polyol and complex esters. Due to their low

viscosity and high volatility monoesters are poorly suitable as lubricants. Polyo1

esters are chemically more stable than for example diesters, due to the structure

of the polyols used in the preparation of said esters wherein no hydrogen atom is

211V

3

Il5'

95367

311V

attached to the fJ carbon atom. Complex esters have promising lubricative proper­

ties but the manufacture thereof on an industrial scale is difficult because of the

severe conditions required by the reaction, especially if said esters are prepared ­

from (purified) fatty acids and alcohols.

If polyol esters are prepared by using no alfa hydrogen acids, the stability proper­

ties of the esters can be further improved. Metro et ai. (CA 859 771) have shown

that the no alfa hydrogen carboxylic acids increase the thermal and oxidative

stability of esters prepared from no beta hydrogen alcohols, as well as slow down

the hydrolysis of the esters.

As the low viscosity polyol esters are not suitable for traditional uses wherein

high viscosity is required, it has been aimed at preparing polyol esters of higher

viscosity from for example trimethylol propane (TMP). However, it has been

found that it is difficult to obtain simple TMP esters with both high viscosity and

a low pour point (cf. for example US 4,061,581).

Products based on vegetable oils are nowadays used more and more as lubri­

cants because of their safety to the environment. Natural vegetable and animal

oils are glyceride diesters, i.e. tri-, di- or monoesters of glycerol and straight

chain saturated and unsaturated fatty acids. The lubricant industry uses for instan­

ce rapeseed, rape, soybean, castor, olive, coconut, palm and pine oils.

The advantageous properties of vegetable oils include user friendliness and non­

toxicity. In addition, vegetable oils are renewable raw materials and degrade in

the environment without accumulating in the food chain of nature. However, the

use of vegetable oils as lubricants has been limited by their poor stability proper­

ties. The poor thermal and oxidative stability is due to unsaturated and polyun­

saturated fatty acids. On the other hand, the unsatisfactory behaviour of vegetable

oils at low temperatures is due to the saturated fraction of fatty acids. By using

suitable additives and by favouring in cultivation such varieties which do not have

4 95367

4/1V

a too high degree of saturation, it has been possible to somewhat improve the

stability properties. Also the purification of the oil for technical use is helpful.

Furthermore, attempts have been made to modify natural glyceride esters in order

to improve their stability properties. Known processes include catalytic hydro­

genation, alcoholysis, geometrical isomerization and sulfurization. For example

by hydrogenation a certain amount of double bonds from the unsaturated part of

vegetable oils can be removed, and by isomerization the amount of undesired iso­

mers can be decreased.

Van der Waal and Kenbeek have presented a process for the preparation of synt­

hetic esters from vegetable oils or animal fats (proceedings of the Tribology

2000, 8th International Colloqium, Technische Akademie Esslingen, Germany,

14-16 June 1992, Vol II, pp 13.3-1 - 13.3-8). The process comprises first decom­

posing the glyceride esters of the starting material into fatty acids and glycerol

and subsequently separating the fatty acid fraction into liquid and solid phases.

The fatty acids of the liquid phase are separated by distillation into single fatty

acids which can be further modified e.g. by hydrogenation or cracking to obtain

the desired raw material. Fractions containing a single fatty acid are esterified

with no beta hydrogen polyols for preparing a synthetic ester.

The .fatty acids of the ester prepared according to the above described process

usually contain less unsaturated double bonds than the fatty acids of the starting

material, which improves the oxidative stability. However, the costs of the pro­

cess are extremely high, due to the multistage separation and purification reac­

tions and the most severe conditions (high pressure and temperature) required by

the reaction. Moreover, it has been found that when fractions containing only a

single fatty acid are reacted with polyols, plenty of mono- and diglycerides are

formed, i.e. all the OH groups of the polyols do not react. This decreases the

triglyceride yield and the raw material has to be recycled several times if the

yield is to be improved. Furthermore, the reaction of a fatty acid and an alcohol

creates water which has to be removed during the reaction.

5

1/7

95367

5JlV'4

Transesterification of fats by means of lipases is known as such. The literature in

the field discloses especially various systems for the immobilization of the lipases

used (cf. for example EP patent application 579 928 and US patents 4,798,793

and 4,818,695). The immobilization of lipases facilitates their application both in

continuous and batch processes. Patent publication GB 1 577 933 discloses a pro­

cess for modifying trig1ycerides with a lipase, especially with an immobilized

lipase. However, the literature in the art does not describe the use of lipases as a

catalyst in the process according to the present invention.

According to the invention it has now been found that it is possible to prepare

synthetic esters with good lubricative properties from vegetable oils by an enzy­

matic process which avoids the multistage reaction with several separations and

recyclings and by which good yields are obtained.

In the process according to the invention a vegetable oil is first transesterified by

reacting the vegetable oil with a lower alkanol to obtain a mixture of fatty acid

lower alkyl esters. The process is characterised in that the mixture of esters obta­

ined from the first reaction is further transesterified by reacting said mixture with

a no beta hydrogen polyol of the formula

wherein R is a C1-C6 alkyl group, particularly a C1-C4 alkyl group, or a -CHzOH

group, in the presence of a lipase enzyme, and the synthetic ester obtained is

recovered.

Vegetable oils suitable as a starting material in the process are for example ra­

peseed, rape, soybean, castor, olive, coconut, palm, pine, maize, walnut, flax­

seed, cotton, sunflower, sesame and almond oils, especially rapeseed oil, rape oil,

pine oil and soybean oil, particularly rapeseed oil or rape oil.

6 95367

611V

The first transesterification reaction of the process according to the invention is

carried out by a process known per se, by reacting a refined or alkalirefined

vegetable oil with a lower alkanol to obtain a mixture of fatty acid lower alkyl

esters.

The lower alkanol used in the first transesterification reaction is preferably a C1­

C4 alkanol, especially methanol or ethanol. The obtained mixture of lower alkyl

esters of the vegetable oil is thus preferably a mixture of C1-C4 alkyl esters,

especially a mixture of methyl or ethyl esters. If desired, usual esterification

catalysts may be used in the reaction, and the amounts of the reactants and the

reaction conditions (pressure, temperature, reaction time) are either commonly

known or easily chosen by.a person skilled in the art. The reaction may also be

carried out by using a suitable enzyme as a catalyst.

The first transesterification reaction may be illustrated by the following general

reaction scheme I:

H2C-O-C( =O)-R1I

HC-O-C( =O)-R2I

H2C-O-C( =O)-R3

+ 3 ~-OH ~ 3 ~-O-C(=O)-Rx

H2C-OH1

+ HC-OHI

H2C-OH

(I)

wherein R1, R2 and R3 are fatty acid residues, ~ is an alkyl residue, especially a

CC C4 alkyl residue, and Rx is Rb Rz or R3- Glycerol is formed as a by-product.

The fatty acid lower alkyl ester obtained from the first transesterification reaction

is thus a mixture comprising various fatty acids of the vegetable oil used as the

starting material. It is typical of the invention that this mixture of fatty acid lower

alkyl esters may be used directly as the starting material of the second transesteri­

fication reaction without separation or purification of fatty acids.

In the second transesterification reaction according to the invention, the mixture

of fatty acid lower alkyl esters obtained from the first transesterification reaction

7

Ilfj

95367

711V

is reacted with a no beta hydrogen polyol, such as for example trimethylol etha­

ne, trimethylol propane, trimethylol butane or pentaerythritol, especially with

pentaerythritol or trimethylolpropane, in the presence of a lipase.

The second transesterification reaction may be illustrated with the following gene­

ral reaction scheme II:

/CH20H

3 ~-O-C(=O)-Rx + R-C - CH20H ~

"'CH20H

+ 3 ~-OH

wherein Rj , R2, R3, ~ and Rxhave the same meanings as in the reaction scheme

I and R is a C\-C6 alkyl group, especially a C j -C4 alkyl group, or a -CH20H

group.

Consequently, it is the question of a totally different chemical reaction than in the

process of the prior art wherein a free fatty acid is esterified with an alcohol. In

the process according to the invention, an ester is reacted with an alcohol, and

thus it is the question of a transesterification reaction which reaction, as well as

the reaction conditions required by it and the by-products formed therein, is

totally different from the reaction used in the process of the prior art.

The synthetic ester obtained from the second transesterification reaction is reco­

vered and, if desired, purified by conventional methods, for example by neutrali­

zation and washing with an aqueous acid. No distillation or any other special

treatment is needed as the obtained ester is ready to use as such as a raw material

of lubricants.

When a polyol is reacted with a mixture of fatty acid lower alkyl esters in the

presence of a suitable lipase, almost all OH-groups of the polyol react into trigly­

cerides. From 75 to 98 % of the theoretical yield of the triglyceride is obtained,

the proportion of mono- and diglycerides being in total from about 2 to 25 %.

8 95367

allv

The product obtained does not contain any free fatty acids which makes it an

especially advantageous raw material for lubricants wherein the oxygenation of

free fatty acids would cause problems (corrosion, change of viscosity). The pro­

cess is well adapted for industrial scale and the synthetic ester obtained has better

stability properties than the vegetable oil used as the raw material, while at the

same time the advantageous properties of a vegetable oil (biodegradability, non­

toxicity, user friendliness) are maintained.

By the process according to the invention it is thus possible to prepare synthetic

esters from vegetable oils, for example from rapeseed oil, in a yield of even over

95 % of the theoretical. In this case, the di- and monoglycerides of the product

are also calculated in the yield. During the tests carried out it has been observed

that the advantageous properties of the product are maintained in spite of the

moderate (up to 30 %) di- and monoglyceride content.

The no beta hydrogen polyol and the mixture of esters are preferably reacted with

each other in a molar ratio of about 1:2 to 1:6, especially in the molar ratio of

about 1:3 to 1:3,5.

The second transesterification reaction, being characteristic of the invention, is

preferably carried out in a reduced pressure generator provided with reflux, for

example under negative pressure of 2.0 to 12 MFa, preferably under negative

pressure of 5.3 MPa. The reaction is carried out at a temperature wherein the

lipase used is active, for example at a temperature between 37°C and 69 °c,

preferably at a temperature between 42°C and 47 °C. A suitable reaction time is

from 24 hours up to 72 hours, depending on the other conditions and the enzyme

used. It is preferred to add water to the reaction mixture, for example about 0.1 ­

29 %, preferably 8 - 15 %, or to carry out at a higher temperature without ad­

ding water.

The amount of the enzyme is preferably from about 2 % up to about 50 % calcu­

lated (w/w) on the substrates. With a 68 hour reaction, a methyl ester of rapeseed

9

/2/

9536/

19/IV

oil is completely made to react into products only with an enzyme amount of 10

%. The amount of the enzyme needed may be decreased by immobilizing the

enzyme. In the process according to the invention, a lipase obtained for example

from Candida rugosa (ex. cylindraceae), Mucor miehei or Pseudomonas fluores­

cellS may be used. The lipase may also be produced by transforming a gene co­

ding for the desired enzyme into another host organism, by cultivating the host

thus obtained and by isolating the lipase produced by it. Commercially obtainable

immobilized lipases may be used, or the free lipase may be immobilized before

use for example on an ion exchange resin, adsorption resin, celites, diatomaceous

earth or silica gel according to the conventional immobilization methods.

The synthetic ester prepared by the process according to the invention is an excel­

lent raw material for the preparation of lubricants. Lubricants, especially hyd­

raulic oils, which contain a synthetic ester prepared by the process of the inventi­

on, optionally with one more additives, are also included in the scope of the in­

vention. As additives for example oxidation inhibitors, antiwear agents, antifoam

agents, corrosion inhibitors, dispersants, viscosity index improvers and/or pour

point depressers which are generally known in the art, may be used.

Oxidation inhibitors include for example amines and phenols. As antiwear agents

and corrosion inhibitors for example phosphates or sulfonates and as antifoam

agents for example metal sulfonates, metal phenates, polyesters or silicones may

be used. Viscosity index improvers include for example polyisobutenes, styrene­

butadiene and ethene-propene-eopolymers which all are thus suitable also as pour

point depressers.

In the following the invention is further described by means of examples, the pur­

pose of which is to illustrate but not to limit the invention.

10

Example 1

95367A methyl ester of rapeseed oil was prepared as follows: Rapeseed oil (0.3 moles)

was weighed into a three-necked flask provided with a thermometer, cooler and a

stirring device. Stirring was started and methanol (2.0 moles) was added. The

reaction mixture was heated to 60°C and the alkali catalyst used was added (0.5

%, w/w). Stirring was continued for three hours. The progress of the reaction

was followed by thin layer chromatography. The reaction mixture was washed

with an aqueous acid. The glycerol created in the reaction mixture was separated

and the product mixture was analyzed. Rapeseed oil ester content was 97 %.

Example 2

In a 25 cm3 round pottom flask attached to a Liebig-refluxer of 20 cm with a cold

(about +6°C) tap water circulating in the cooling jacket, was weighed 0.607 g

(4.52 mmoles) of solid trimethylol propane (Merck, Darmstadt, Germany), and

0.7 ml of destilled water was added. After dissolution, 4.00 g (13.56 mmoles) of

methylated rapeseed oil (Raision Yhtyma, Finland) was added and finally 1.79 g

of Candida rugosa lipase (Biocatalysts Ltd., Pontypridd, Great Britain) in powder

form. A negative pressure of 5.3 MPa was sucked into the device. For stirring a

magnetic stirrer was attached to the device. The reaction mixture was stirred with

the magnetic stirrer at a speed of 200 rpm. The starting point of the reaction was

counted from the moment the suction for the reduced pressure was connected to

the device. The reaction temperature was 42°C and the total reaction time 72

hours. The amount of substituted TMP esters in the final product was over 98 %

in total.

Example 3

Example 2 was repeated with 1.84 g of a Mucor miehei lipase Lipozyme 1M

(Novo Nordisk A/S, Bagsvaerd, Denmark) bound to a solid support. Water was

not added to the reaction mixture. The reaction temperature was 58°C. The

10llV

11

(23

95367

1111V

TMPE content of the final product was 75.0 % after 24 hours and 92.5 % after

66 hours. There were no starting materials left after 66 hours.

Example 4

Example 2 was repeated with 1.84 g of a Candida rugosa lipase bound to a solid

support. 0.7 ml water was added to the reaction mixture. The reaction temperatu­

re was 47°C. The TMPE content of the fmal product was 62.7 % after 48 hours

and 72.9 % after 78 hours.

The enzyme bound to the solid support was prepared as follows: 3.33 g of lipase

was dissolved in 100 ml of 0.05 M sodium phosphate buffer, stirred for 2 hours

and filtrated. To an erlenmeyer flask of 250 ml 40 g of a buffered support (e.g.

MWA-1, Mitsubishi Chemical Company, Japan; 43.4 % dry matter) and 60 m1 of

enzyme solution (2 g lipase) was added, shaken for 3 hours at a speed of 130

rpm, filtrated and lyophilized for 30 hours to a dry solids content of 98.9 %.

Example 5 Preparation of a hydraulic oil from a rapeseed oil ester and compa­

rison of hydraulic oils

The raw material used was the synthetic rapeseed oil ester obtained in Example 2.

said ester was mixed at a certain temperature with additives to obtain a hydraulic

oil having the following composition:

The synthetic ester from Example 2

Oxidation inhibitor

Pour point depresser

Antiwear agent

Antifoam agent

90 - 98 % by weight

0.1 - 2.5 % by weight

o- 5.0 % by weight

0.1 - 2.0 % by weight

o - 0.5 % by weight

The technical properties studied of this ester containing additives were wearing,

friction, oxidation, low temperature properties and corrosion.

12 95367

12/1V

Wearing and friction were examined with a four ball test (ASTMD 2783, IP 239)

wherein wearing with respect to loading or the extreme loading where the lubrica­

tion still works, are measured. Oxidative properties were studied with an oxygen

bomb test (ASTMD 925) and with the oxidation test DIN 51586 where the change

of viscosity at 40°C was monitored. In a corrosion test (Cincinnati-Milacron test)

the aging of the oil as well as copper and steel corrosion were studied. In said

test, the change of the total acid number (TAN) and viscosity, the weight change

of the copper and steel rods used as oxidation catalysts in the test procedure and

the formation .of a precipitate under the test conditions are measured. Furthermo­

re, the pour point which illustrates the low temperature properties of an oil was

analyzed, i.e. the temperature where the oil is still fluid.

The corresponding properties were examined also from hydraulic oils of the state

of the art containing the same additives and from hydraulic oils based directly on

rapeseed oil containing also the same additives. The results are shown in Table 1.

/2?

13

95367-

Table 1. Comparison of the properties of hydraulic oils. A = hydraulic oil withthe ester prepared by the process of the invention as raw material, viscosity· grade32; B1 and B2 = hydraulic oils with commercial synthetic esters as raw mate­rials, viscosity grades 46 and 68; C = commercial hydraulic oil based on rape­seed oil, viscosity grade 32.

A B1 B2 C

Four ball test- extreme loading, N 2500 3000 2500 2000- wearing, mm 0.42 0.46 0.41 0.64

Oxygen bomb test 40 39 29 30ASTDM D445, psi

Oxidation inhibition testDIN 51586, viscosity change, 11.5 20.3 24.1 28.8%

Cincinnati-Milacron test- TAN mg KOH/g

before 1.38 1.39 1.40 1.72after 1.58 3.71 2.41 0.61TAN 0.20 2.32 1.01 1.11

- viscosity change, % 19.0 16.9 6.2 8.2- total precipitate, mg/l00 m1 1.0 17.0 28.8 4.4- weight change of Cu rod, mg 1.5 -16.9 0 -0.5- weight change of steel rod, 0.2 0.4 1.2 -0.5mg

Pour point, ·C -41 -36 -39 -39

From the results it can be seen that as regards low temperature properties, the

ester prepared by the process according to the invention is equal to the commer­

cial raw materials on the market and better than the commercial product based on

rapeseed oil. From the Cincinnati-Milacron test it can be seen that the change of

the total acid number (TAN) is clearly the lowest with the ester of the invention.

The increase in viscosity at 40°C is of the same order with all, as well as the

weight change of copper and steel rods. The results of the oxygen bomb test are

equal, as well as the results of the test according to DIN 51586 and the four ball

test.

13/1V

14

Claims

953671. An enzymatic process for preparing a synthetic ester from a vegetable oil,

comprising

- transesterification of said vegetable oil by reacting it with a

lower alkanol to form a mixture of lower alkyl esters of fatty acids,

- a second transesterification reaction wherein the obtained

mixture of esters is reacted in the presence of a lipase (triacylglycerol acylhydro­

lase; EC 3.1.1.3) with a no beta hydrogen polyol of the formula

wherein R is a CC C6 alkyl group, particularly a C\-C4 alkyl group, or a -CH20H

group, and

- recovering the synthetic ester obtained.

2. The process according to claim I, wherein the vegetable oil is rapeseed oil.

3. The process according to claim 1, wherein the lower alkanol is a C\-C4 al­

kanol, especially methanol or ethanol.

4. The process according to claim l, wherein the fatty acid lower alkyl ester is a

methyl ester of a fatty acid.

5. The process according to claim I, wherein the no beta hydrogen alcohol is

selected from the group consisting of trimethylol ethane, trimethylol propane,

trimethylo1 butane and pentaerythritol.

14/1V

15

/21

95367

1511V

6. The process according to anyone of the claims 1 to 5, wherein the second

transesterification reaction is carried out in the presence of an immobilized lipase.

7. The process according to anyone of the claims 1 to 6, wherein the second

transesterification reaction is carried out in the presence of a Candida rugosa

lipase.

8. The process according to anyone of the claims 1 to 6, wherein the second

transesterification reaction is carried out in the presence of a Mucor miehei lipase.

9. The process according to anyone of the claims 1 to 8, wherein the enzyme is

separated after the reaction and recycled.

10. The process according to anyone of the claims 1 to 9, wherein the second

transesterification reaction is carried out with a lipase obtained by transforming a

gene coding for said enzyme into another host organism for producing the lipase.

11. The process according to anyone of the claims 1 to 10, wherein the reaction

mixture contains about 0.1 to 29 % water.

12. The process according to anyone of the claims 1 to 11, wherein the reaction

temperature in the second transesterification is between 37°C and 69 °c.

13. The process according to claim 1, wherein the no beta hydrogen polyol and

the mixture of esters are reacted with each other in a molar ratio of from about

1:2 to 1:6, especially in a molar ratio of about 1:3 to 1:3,5.

14. The use of a synthetic ester obtained according to claim 1 for the preparation

of lubricants, especially for the preparation of a hydraulic oil.

15. A lubricant which comprises a synthetic ester obtained according to anyone

of the claims 1 to 13, optionally with one or more additives.

16 95367

( 1611\

16. The lubricant according to claim 15 which comprises about 90 to 98 % of a

synthetic ester and about 2 to 10 % of additives.

17. The lubricant according to claim 15 wherein the additive(s) is (are) an oxida­

tion inhibitor, an antiwear agent, an antifoam agent, a corrosion inhibitor, a

dispersant, a viscosity index improver and/or a pour point depresser.

17

Abstract

95367The object of the invention is a process for preparing a synthetic ester from a

vegetable oil by means of lipase enzymes. Further objects of the invention are

lubricants containing a synthetic ester prepared by the process according to the

invention.

17/1\

Appendix V 1Nwith the permission of the publisher

Merja Lamsa

R&D Manager

Research and Development LaboratoryOil Milling DivisionRaisio GroupP.O.Box 101

21201 Raisio

Tel: 921- 434 2746Fax: 921- 434 2911

ECOLOGICALLY ACCEPTABLE SYNTHETIC HYDRAULIC FLUIDS BASED

ON VEGETABLE OILS

1. Introduction

There is an increasing interest of veqetan~e oils to beused as raw material for synthetic escer basedecologically acceptable fluid (1,2,3).Products based on natural oils and fats are used widelyas lubricants, mainly chainsaw oils and hydraulicfluids. vegetable oil based hydraulic fluids entered tothe market at the middle of 80s (4,5).These products fullfill their function very well, butwhen requirements are higher for example fortemperature, synthetic esters corne on picture.First synthetic esters were made for gas turbine engines(6). In the late sixties and seventies the advantages ofsynthetic esters were recoqnized also for otherapplications like gear oils, hydraulic fluids andcompressor lubricants (7,8).The-main raw materials for esters are alcohols and fattyacids. Fatty acids are originally from natural oils andfats. Alcohols may vary from mono- to oolyalcohols. Thecareful choice of-fatty acid and alcohol gives thedemanded properties :or the final product. In hydraulicfluids it is important to have right viscosity, goodoxidation and hydrolytic stability, low pourpoint andstability of viscosity at cold temperatures andnaturally good antiwear properties.In Raisio Finnish raoeseed oil and its esters have beenused together with polyols to synthetize polyolesters,starting materials for synthetic hydraulic fluids(9,10) .Laboratorytest results are demonstrated from raw

materials and final products, hydraulic fluids.Fieldtest results are shown from different hydraulicapplications.From these testresults can easily be seen thatvegetableoil based synthetic esters work very well asraw material for hydraulic fluids.

814

/J/

2.Methods

Synthetic esters of vegetable oil can be prepared eitherchemically or enzymatically (9,10). Raw material can berapeseed-, tallo, soyabean-, coconut- or tallow oilfatty acids. Alcohols used can be short, straightchainor longer, sidechain alcohol, dialcohols or polyols(1,5)" Examples of different esterification reactionsare shown in Figure 1.

Figure 1: Examples of molecular schedule of syntheticester reaction (11,12,13).

-------------------------------------------------------

2N

CH. -OR""

oCH -O-C-RI a 0

II

CH-O-C-R"

I ';I 0

C~ -O-C-R"

+ 3 R" OR.D. ,catalyst 0.-====-~' 3 R'''-O-C-R

o1

' + CH-OR"

B. Shortchain alcohol and diacids

A, catalyst Q 02ROR .,. ROOC- (CH,2.)~ -COOR, - J RO-C- (CR.:l,)" -C-OR + R" 0

C.Polyols and triglycerides

aCH... -a-CoRI a

3 CH-O-C-R'

I ~ ,j

C~ -a-c-R

CH"OHI

R-C-CR~ ORI

CH;:.OR

6, catalyst- 'R..,

aCR~ -O-C-RI 9

-C-CR.-O-C-R +I "0CH.l -O-C-R

CH.t -ORI

3 CR-ORI

CR.-OR...-------------------------------------------------------

815

Figure 2: The chainlenght and sidechains of acid effectonto the physical properties of the polyolesters (14) .

3N

--------------------------------------------------------

Influence of chain length

Vjsc.l~

Influence of side Chains

VISC'!~

cnatn ••noU1

I~cour IPOint!

V.l.

pour IPOint I

side cnalna--------------------------------------------------------

Figure 3: Effect of polyol functionality and chainlenght on the viscosity of polyol ester(15J.

40

3!l

30 ""1

., I..~ 2~9

o Iff6!--~7--.-::8:----:9:--~fO---:':I,---:'IZ·

_-ACro CHAIN LENGTH

--------------------------------------------------------

As an example of how the properties of the final estercan be affected by raw materials are shown in Figure 2and 3. The following features of the starting compoundsaffect to the properties of the resulting ester:molecular weight, the size of the acyl group, thefunctionality of polyols and the method of preparationof the ester or mixture of esters (15,16).

816

Bes~ qualities of synthetic esters for hydraulic fluidsuse one can get from polyolesters or combination ofpolyolesters with diesters. It is also very importantnot to forget the environmental side of syntheticesters. There exist for example diesters that areto~ally biodegradable, but do not work technically asthey should or vice versa. For environmental sidebiodegradability is not enough, also toxicity andbioaccumulation properties should be tested.Different type of additives were also tested. They have

changed enourmously during last two- three years.Information of their environmental side is now availableand additive industry make more custom- made additives.

4N

Figure 4 Examples of environmentally acceptbaleadditives for hydraulic fluids.

--------------------------------------------------------

Add:"::ive Purpose Biodegradability Toxicity Bioaccumulation

Phenolic Antioxidant > 80%'(OECD 301B) EC >100 mg/l

Amineohos- Antiwear andpor:Jussalt antifirction > 70%'(OECD 301B) LC >10 rng/l log P""" »6

Ester of Pouroointoleic acid depresser > 70%' (OECD 301D) LC >50 rng/l log Po"" >6

--------------------------------------------------------

These new additives have been tested in both natural andsynthetic esters of veqetable oil based hYdraulicfiuids. All tests are performed in independent, outsidelaboratories with standardized methods.

3.Results

PhYsical and chemical properties of natural andsynthetic esters of veget~le oil based hydraulic fluidshave been measured. The hydraulic fluids in test werevegetable oil based hydraulic fluid (RES 32L old and newversion), commercial synthetic ester based hydraulicfluids ( RBS 465 and RB5 685), the latestdeveploment work ( SE) and comparable mineral oil basedhydraulic fluid (MO).

817

Table 1: Physical and chemical properties of somehydraulic fluids (13).

--------------------------------------------------------

5N

HF SE RBS32L RBS46S RBS 68S MO

Viscosi-ty (mm2/s) (19)

40" C 32,9 33,S 32,61000 C 8,0 7,8 6,4VI 220 220 187

Filtera-blity (%) (20) 92 96 72

Cleanclass (21) 13/8 B/S 14/10

Colour (22) 4- 3+ 4 4+--------------------------------------------------------

One of the most important basic properties is the coldstability behavior~ Pourpoint is-one way to measure it,more important is the stability of viscosity at coldtemperatures. Pourpoint reflects the temperature wherethe-fluid is still-fluidy. Viscosity measurement at coldtemperature show the stabolity of the fluid in cold(17) .

Table 2: pourpoint ( ASTM D97) (17) and viscosity atcold- temperatures ( ASTM D 445) (17) of- somehydraulic fluids.

--------------------------------------------------------

HF SE RBS 32L RBS 468 RBS68S MO

Pour-point ('C) -41

Viscosity(mm2/s)Temp. (~C)

o-10-20-25-30

-39

3728291628

2940

-36

57314302350

-39

21603670

4961221

2038

Time ( days)1 30503 30507 3050--------------------------------------------------------

818

~3S- 6N

Oxidation stability has been claimed to be one of theweaknesses of vegetable oil based hydraulic fluids. withthe right choice of raw material and additives this isno longer a problem. Oxidation stability can be measuredby many different standardized test like oxidation bombtest ( ASTM D 525) , Baader test ( DIN 51 554 teil 3), viscosity test (DIN 51586) (19,21). problematic isthat there does not exist norms, that everyone would usethe same method.

Table 3: Examples of oxidation stability tests ofenvironmentally acceptable hydraulic fluids.

--------------------------------------------------------

SE

Test­methodASTM D445(psi) 42

DIN 51586(1:) %) 12,4

DIN 51554TAN

(mgKOH/g)vise.(~%)

RBS 32L

30

28,8

0,05

6,2

RBS 46S

39

20,3

RBS 68S

29

24,1

MO

39

--------------------------------------------------------

All vegetable oil based products have good propertiesfor wear and friction reduction. This is due to theirchemical structure. They form an unimolecular, strenghtfilm on the metal surface, which lowers friction andreduces wear. The molecular structure of vegetable oil,vegetable oil based ester and mineral oil is showed inFigure 5 to compare their chemical structures. As canbe seen, vegetable oil and its esters contain oxygenatoms at their molecular structure. This explains thementioned properties. -

819

Figure 5: Molecular structure of vegetable oil (A),vegetable oil based ester (B) and mineral oil(C) for hydraulic fluid use (13,22).

--------------------------------------------------------

7N

oCH.. -O-C-RI 0

CH-O-C-R'o.\

CH~ -O-C-R"

R"' - 0 - ~ - R"" Et MeEt-CH" -CH-CH.t, -CH-Me

(A) (B) (C)

--------------------------------------------------------

Friction and wear nronerties can be measured atlaboratory scale for example by FZG-test ( DIN 51 354teil 2) (21), Vickers- pumntest ( DIN 51 389 teil 2)(21), Mobil-test (23) and fourball test (ASTM D 2783, IP239) (17).

Table 4: Fourball test- eauinment and some testresultsfrom hydraulic fluids ( ASTM D 2783, IP 239) .

--------------------------------------------------------

HF SE RBS32L RBS46S RBS8S MO limit

maxload (N) 2000 2000

wear (rom) 0,4 0,48

2000

0,46

2500

0,41

2200

0,54 < 0,5--------------------------------------------------------

Corrosion does not cause problems with vegetable oilbased products, but of coarse resistance can be improvedby additives. One laboratorytest to measure it, isCincinnati- Milacron test (24). In the testresults canclearly be seen the affect of additives, how much betterare the custom-made additives.

820

/37

Table 5: Cincinnati- Milacron test from vegetable oil,synthetic ester of it based hydraulic fluids.

--------------------------------------------------------

8N

HF SE RES 32Lo RBS 32Ln RBS 46S RBS 68S

TAN(mgKOH/g) 0,17

vise.(%)/40 C 19,1

totalsludge 1,1(mg/l00ml)

weight change (mg)

Cu- plates 1,7Fe-plates -0,3

1,74

10,2

45,4

0,30,5

1,11

6,2

20,5

- 0,8o

1,01

16,9

28,2

o1,2

2,32

12,9

17,0

-1,60,4

--------------------------------------------------------

4.Environmentally ~riendlv properties

Biodegradability can be tested by many differentmethods. They are all standardlized methods and none ofthem can be said to be clearly better than others.OECD-301- serie (A-F) (25) is the newest and mostreliable for todays knowledge. Others are like CEC-L-33­A-93(26), DIN 38412 (21).

Table 6: Biodegradability of vegetable oil and syntheticesters of it based hydraulic fluids.

--------------------------------------------------------

HF SE RBS 32L RBS46S RBS68S limit

TestmethodCEC >90% >90% >90% >90% 70%

DIN 38412

OECD 301F

5 days

>85%

14 days

60%--------------------------------------------------------

As important as biodegradability are bioaccumulation andtoxicity properties of lubricants. Bioaccumulation meansif product or parts of it accumulates in nature inmicro- organisms, plants, animals etc. It is usuallymeasured by OECD- standards ( 3U5 A-E) (27). Toxicityproperties are also measured by OECD- standards, OECD201- 203 (28). As toxicity- meter are used fishes,water-fleas or algaes.

821

Table 7: Examnle of bioaccumulation ( OECD SOlD) andtoxicity test ( OECD 201- 203) of vegetable oil

based hydraulic fluid.

--------------------------------------------------------

9N

HF

Eioaccumulation ( 305D)

Toxicityfishes ( 201)waterfleas (202)algaes ( 203)

RES 32L

not bioaccumulating

LQo > 1000 mg/lLC:., > 100% WSEETCs.. > 50% WSE

--------------------------------------------------------

5. Fieldtests

synthetic esters of vegetable oil based hydraulic fluidshave been in fieldtests over two years. In fieldtest areallife machinery is filled with tested oil andoilsamnles are taken after certain usaqehours. Thesesamples are examined at laboratory. -

Table 8: Fieldtest results from Wille- multifunctionaltractor (30) and Timberjack- forestmachine (type 1210) with RES 46S and RES 68S hydraulicfluids.

--------------------------------------------------------

Machinery Hvdraulic Hours Viscosity VI Purity Colour TANfiuid in use 40~C 1.00"C

(h) (rnm2/s) (mgKOH/g)

Wille 845 RES 68S 0 59,0 11,1 1.83 10/8 4 2,45290 54,9 10,8 193 15/13 5 2,05872 54,9 10,5 181. 15/18 6 2,00

1140 55,1 10,7 190 14/12 6 1,951515 55,4 10,7 184 15/10 8 1,80

Timberjack121.0 RES 465 0 43,3 8,7 186 13/11 1,67

629 42,7 8,7 187 13/11 10 1,681213 42,7 8,6 184 15/12 11 1,631577 43,0 8,8 188 14/11 11

--------------------------------------------------------

From fieldtest results can be seen ( Table 8) thatviscosity and viscosity index are very stable with time.Also oxidation and hydrolytic stability are very goodand friction and wear additives work perfectly.

822

After these fieldcests the manufacturer of machinery cangrant approval for the cested hydraulic fluid. Alsoresearch and develooment acauires a lot of informationto develope even betcer lubricants.

6. Conclusions

Both natural and synthetic esters based on vegetable oilhydraulic fluids seem to fullfill their job very well.Based on laboraccry tests, oxidation stabilityespecially at higner temperatures of vegetableoil basedsynthtetic esters is better than in natural esters. Thisdifference could not be seen in fieldtests.Coldstability, especially viscosity at low temperatures,is clearly becter in synthetic escers at temperaturearea -10 to -20 C, but then at colder temperatures itis vice versa, natural esters are better in temoeracurearea -25 to -40 C. At fieldtests these differences couldnot be noticed. In friction and wear by laboratory testsno differences could be seen. In additives one couldvery clearly see that with careful chose of raw materialand custom- made additive- package, at labscale,testresults were much better than older packages.From fieldtest results could be noticed that syntheticester based lubricants maintain thewir chemical andphysical properties very well.Results from laboratory and fieldtests have been sopromising, that synthetic ester of vegetable oil basedhydraulic fluids are coming into the market.

References:

1. Linko, Y.-Y., Larosa, M.,Huhtala, A and Rantanen, 0.,J.Am.Oil Chern. Soc. soecial issue on Biocatalysisbased on the speech at 86th AOCS Anual Meeting, SanAntonio, Texas, USA 6.-15.5.1995.

2. Metro, E.M. and Matuszak, A.H., Canadian Patent 859771 ( 1970).

3. Berens, G. and Milton, L.H., U.S. Patent 4 263 159(1981) .

4. Vilenius,M., Proceedings of the Nordtrib 1984, NordicSymposium on Tribology, Tampere, Finland.

5. Carr,D.D. andDeGeorge, N. U.S. Patent 4 826 633(1989) .

6. Flowerday, P. and Robson, R., G.B.Patent 1 307 7271(1970) .

7. Leleu,G., Bedague,p. and Sillion, B., u.S.Patent 4061 581 (1977).

8. van der Waal,G. and Kenbeek,D., proceedings ofTribology 2000, 8. International Colloqium,Esslingen,Germany, Band I,vol IV, 1992, 13.3.

9. Larosa,M., A method for the preparation of syntheticester from vegetable oil, Finnish patent apply number944 118, 1994.

823

1Q.(.\{

10.Larosa, M., Linko Y,-Y., Linko P. and Uosukainen E.,An enzymatic method for preparing synthetic esterfrom vegetable oil, Finnish patent number 944 119,1994.

11.Raport of Biodiesel project in Finland, Ministry ofAgriculture and Forestry, Finland ,1992.

12.Wildersohn,M., Tribologie+Schmierrungstechnik 32(1985) 70-75.

13.Larosa,M.,Finnish Tribol., 14(1), 1995 39-45.14.Niedzielski,E.L., Ind.Eng.Chem.prod.Res. Dev. 15(1)

1976 54-58.15.van der Waal, G., Proceedings of NLGI 55th Annual

Meeting 53(1989), Florida, USA, 359-378.16.Szydywar,J., J.Synth.Lubr. 1 (1984) 153-169.17.Annual Book of ASTM Standards, volume 05.02.,

Petroleum products an Lubricants, ASTM, Philadelphia,PA 19103, USA.

18.Filterability- project, Royal Institute of Tecnology,Stockholm, Sweden, 1988-1994.

19.ISO 4406-standard, International Orqanization forStandardization, 1980, lOp. -

20.Annual Book of ASTM Standards,vol.05.01,PetroleumProducts and Lubricants,ASTM, Philadelphia, USA.

21.DIN-Taschenbuch 192, Schmierstoffe, Beuth VerlagGMbH, Berlin, Germany,1983.

22.Larosa,M., proceedings of the 2nd Tampere Intern.Conf. on Fluid POwer 1991, Tampere, Finland, 73-89.

23.The Lubrication Engineers Manual, 1st edition, UnitedStates Steel, combiled and edited by Bailey,C.A. andAarons,J.S., 1971, 146-147.

24.The Lubrication Engineers Manual, 1st edition, UnitedStates Steel, conbiled and edited by Bailey,C.A. andAarons. J.S., 1971, 152-153.

25.0ECD-Guidelines for testing of Chemicals, advice 301,Paris, France, 1981.

26.CEC-L-33-A-93, Biodegradability of Two Stroke cycleOutboard Engine Oils in Water, London, UnitedKinadom.

27.0EcD- Guidelines for Testing of Chemicals, advice305, Paris, France, 1981.

28.0ECD-Guidelines for Testing of Chemicals, advice 201­203, Paris, France, 1981.

29.Virtanen, T., Research and Tests of New SyntheticEsters as Hvdraulic Fluids, Master of Science

work(chem.eng.) ,Turku, Finland, 1993.

824

11N

/tf/Appendix 'VI

with the permission of the publisher

Merja LiimsiiRaisio GroupOil Milling Division. R&D LaboratoryP.O.Box 10121201 RaisioTel. 921- 434 2746Fax. 921- 434 2911

VEGETABLE OIL BASED LUBRICANTS

INTRODUCTION

Vegetable oils have been used as lubricants since 1000 B.C. 0,2). Revolution of

mineral oils started early 1900 and is still continuing (3,4). So called green values have

disiurbed this hegemonia. The green values demand in lubricant section products, that

would biodegradme. not bioaccumulate and not to be toxic in nature (5). These demands

are totally fullfilled by natural esters or synthetic esters based on vegetable oils. These

products operate also technically well or even better than conventional mineral oil based

products (6.7.8). New development in vegetable oil based lubricant section are totally

environmental friendly additives and lubricants based on synthetic esters of vegetable

oils. Latest development results and test results of new products are discussed.

METHODS

Synthetic esters of vegetable oils have been prepared either chemically or enzymatically

(9.10). Raw material has been mainly Finnish rapeseed oil. although tall-, soyabean-.

coconut- and tallow oils have also been in tests. Esters are mainly polyolester types.

These new products have been in laboratory tests and some of them already in field

tests. Detailed test methods and results are shown later.

Additive industry has started to produce environmentally more friendly additives during

last three years. These new generation additives have been tested in both natural esters

and synthetic esters of vegetable oil based lubricants. Also detailed information of them

is shown later. All tests are performed in external. independent laboratories with

standardized methods.

39

1N1

RESULTS

Physical and chemical properties of natural and synthetic esters of vegetable oil based

lubricants

Viscosity is determined according to ASTM D445-standard (11). It is measured in 40

0C and 100 0C and from them the viscosity index can be calculated. Viscosity in cold

and the stability of it in time are also very important quantities. They are measured by

the same standard.

Table 1: Viscosity of natural and synthetic ester based lubricants from vegetable oils.

ASTM D 445- standard.

2NI

Temp. Viscosity

(0C) (mm2ls)

CSE RBS 32Lo RBS 32Ln RBS46S SE

40 29,3 33,8 34,0 50,4 32,9

100 7,5 7,9 7,8 9,8 8,0

VI 242 220 220 191 220

MO

32,6

6,4

187

CSE= comparable,commercial synthetic ester based hydraulic fluid

RBS 32Ln= Raisio Bio Safe 32, new, vegetable oil based hydraulic fluid

RBS 32Lo= Raisio Bio Safe 32, old

RBS 46S= Raisio Bio Safe 46, synthetic ester based hydraulic fluid

SE= synthetic ester of vegetable oil based hydraulic fluid. 32

MO= commercial, comparable mineral oil based hydraulic fluid

Filterability is important to be good enough not to cause troubles in hydraulic systems.

If hydraulic fluid contains many particles in it, wear rate increases and finally filters

block- up. Filterability is always measured from the raw material (= natural and

synthetic esters) and from the final lubricant.

Method used is under standardizing and is developed in Royal University of Stockholm

together with users (12). It is called %- method, because results are given in pro cents.

The biggerthe value is (max 100%) the better.

40

1t.f33M

Table 2: Filterability of vegetable oil, synthetic ester based and comparable mineral oil

and synthetic lubricants.

Filterablity (%)

SE CSE

92 63

RES 32Ln

96

RES 32Lo

98

RES 46S

80

MO

72

Small particles can cause serious problems in fine hydraulic systems, for example high

wear. These particles can exist already in basic fluids, in additives or in combining these

two. They are calculated from new and used oil by ISO 4406- standard.

Table 3: ISO 4406- cleanclass of vegetable oil, synthetic ester and comparable synthetic

and mineral oils.

SE

13/8

CSE

15/10

RBS 32Ln

13/8

RES 32Lo

13/8

RBS46S

13/8

MO

14/10

Cold stability of hydraulic t1uids is very important for machines working outside during

wintertime, such as forest machines. Pourpoint is one way to measure it. It tells at what

temperature the t1uid is still fluidy. Standard is ASTM D 97 (11). Cold stability can also

be expressed by long-term coldstability (11). How many days fluid stays fluidy in

certain temperature. Viscosity should also be and stay at certain level in cold

temperature.

Table 4: Pourpoint and viscosity at cold temperatures of some hydraulic fluids.

Pour­

point ( C)

SE

-41

CSE

-48

41

RBS 32Ln

-39

RBS46S

-39

MO

-40

Table 5: Example of viscosity stability of some hydraulic fluids at cold temperatures.

Temp. Visco

(C) (mm2/s)

RES 32Ln MO

0 322

-10 829 496

-20 1628 1221

-25 2038

-30 2940

I day 3050

3 days 3050

5 days 3050

All these base fluids examined here have good properties for reduce wear and lower

friction. This is due to their chemical structure. that is shown in figure I. All these

components contain oxygen, which fonn an unimolecular, strenght film over metal

surface. This film lowers friction and reduces wear.

4NI

o[H ·of.RI 2 0 1

CH -(){.RI 2

[H-~R3rapeseed 0 i I

minerai 011

~H3~~-CHi"SH-{~~- CH3

SHs

Figure I: Molecular structure of vegetable oil, synthetic ester of it and mineral oil (6).

42

To measure friction and wear at laboratory level, quite many different standardized

methods exist, such as FZG- test, Vickerspump- test, Mobil- test. One of them is so­

called fourballtest (standards ASTM D 2783, IP 239)( II), where wear is measured

under loading and maximum loading until lubrication fails. Limits for hydraulic fluids

are <0.5 mm (damage is small) and> 1.0 mm (damage is big).

Table 6: Fourballtest- equipment and some test results from hydraulic fluids.

5NI

SE

max 10ad(N) 2000

wear<mm) 0,4

RBS 32Ln

2000

0,48

RES 32Lo

3000

0,45

RBS46S

3000

0,46

MO

2200

0.54

Corrosion does not cause problems with vegetable oil based products. Of course.

corrosion resistance can also be improved by additives. Example of how these new

generation products are better than lubricants earlier is shown in table 7. This test is

called Cincinnati- Milacron test and it is done for 200 ml of tested oil, kept in 135 0C

for 168 hours with steel- and copper plates in it (12).

43

Table 7: Cincinnati - Milacron test from vegetable oil, synthetic ester based hydraulic

fluids.

TAN mgKOHIg SE RBS 32Ln RBS 32Lo RBS46S

before 1,39 1,72 0,83 1,40

after 1,56 0,61 2.57 2,41

TAN 0,17 1,11 1,74 1,01

Total sludge 1,1 20,S 45,4 28,2

(mgllOOml)

change in vise. 19,1 6,2 10,2 16,9

40 C(mm2/s)

weight change 1,7 -0,8 0,3 0

in Cu-plates (mg)

weight change -0,3 0 0.5 1,2

in Fe-plates (mg)

ENVIRONMENTALLY FRIENDLY PROPERTIES

Biodegradability can be measured by many different methods. They are all standardized

and no one of them can be said to be better than other. Examples of methods: CEC-L­

33-T-82(14),DlN 38412(15), Sturm- test, MITI- test. In table 8 some examples are

presented. As important as biodegradability are bioaccumulation and toxicity properties

of lubricants. Bioaccumulation happens if product or part of it accumulates in nature:

micro-organisms, animals, plants etc.

It can be measured for example by OECD-standard tests (16). Toxicity must be

measured together with bioaccumulation. because product can be, when disposed, toxic

in nature. This value can also be measured by OECD- standards.

Table 8: Biodegradability.

6NI

Testmethod

CEC-L-33-T-82 (%)

DIN 38412

Sturm -test

RBS 32Ln

>90%

5days

>90%

44

limit

70%

14 days

60%

/£/7

CONCLUSIONS

Both natural and synthetic ester of vegetable oil based lubricants seem to fullfill their

function very well. Based on laboratory test results. it can be said that oxidation and

t>specially thermal stability of synthetic esters are better than those of natural esters.

Cold stability. especially viscosity at low temperatures. is clearly better in synthetic

esters at temperature area -10 0C to - 20 oC, but then the role change: Natural esters are

better in colder temperatures ( -25 0C - -40 0C). In friction and wear properties no

difference can be found. Results are. however, so promising, that field test are already

going on.

REFERENCES

I. Ihrig, H.. Mineraloltechnik 35 1990 1-9,12-19.

2:-Boylan. J.B .. Nat. Lubr. Grease Inst. Spokesman 1987 185-195.

3. Odi- Owei. S..1. Soc. Trib. Lubr. Eng. 45 1989685-690.

4. Van der Waal. G.B., Nat. Lubr. Grease Inst. 55th Annual Meeting 53 1989359-368.

5. Stempiel. E.M. and Schmid. L.A.. Proceedings of the Tribology 2000 8th Intern.

Colloquium 1992 voll.

6. Uimsa. M.. Proceedings of the 2Nd Tampere Intern. Conf. on Fluid Power 1991 73­

89.

7. Lamsa. M.. Proceedings of Tech. Acad. Esslingen, Biologisch abbaubare

Schmierstoffe und Arbeitstlussigkeiten 1992 13.7.

8. Lamsa. M.. Proceedings of the 9th Intern. colloquium in Tribology 1994 vol I 2.8.

9. Lamsa. M.. Menetelma synteettisen esterin valmistamiseksi kasvioljysta. Finnish

patent apply number 944118 1994.9. 9.

10. Lamsa. :vl.. Linko. Y.-Y. and Linko. P.. Entsymaattinen menetelma synteettisen

esterin valmistamiseksi kasvioljysta. Finnish patent apply number 944119 1994.

II. Annual Book of ASTM- standards. vol 05.01.. Petro Prod. and Lubr.• ASTM,

Philadelphia. USA.

12. Filterability- project. Royal Institute of Technology, Stockholm. Sweden 1988­

1994.

13. The Lubr. Eng. Manual 1st edition. United Steel Corporation USA.

14. CEC-L-33-T-82. Biodegr.of two-stroke cycle outboard engine oils in water. London.

United Kingdom.

15. DIN-Taschenbuch 192. Schmierstoffe. Beuth Verlag GMbH, Berlin, Germany 1983.

16. OECD Guidelines for testing of Chemicals. Advice 301. Paris, France 1981.

45

7NI