Supplementary Materials for - Science Advances...all other metathesis catalysts adversely affect the...
Transcript of Supplementary Materials for - Science Advances...all other metathesis catalysts adversely affect the...
advances.sciencemag.org/cgi/content/full/3/6/e1602624/DC1
Supplementary Materials for
Biofuel by isomerizing metathesis of rapeseed oil esters with
(bio)ethylene for use in contemporary diesel engines
Kai F. Pfister, Sabrina Baader, Mathias Baader, Silvia Berndt, Lukas J. Goossen
Published 16 June 2017, Sci. Adv. 3, e1602624 (2017)
DOI: 10.1126/sciadv.1602624
The PDF file includes:
Supplementary Materials and Methods
Supplementary Text
fig. S1. Ru-based metathesis catalysts tested in the isomerizing hexenolysis,
including second-generation indenylidene-ruthenium complexes Umicore M41
(Ru-2) and M31 (Ru-3) and Hoveyda-type catalysts Umicore M51 (Ru-4), M72
SIMes (Ru-5), and M74 SIMes (Ru-6).
fig. S2. Olefin blends obtained by isomerizing hexenolysis with different Ru
catalysts.
fig. S3. State-of-the-art isomerization catalysts tested in the isomerizing
hexenolysis.
fig. S4. Mass-corrected GC with IC-1.
fig. S5. Mass-corrected GC with IC-2.
fig. S6. Gas chromatogram with IC-3.
fig. S7. Gas chromatogram with IC-4.
fig. S8. Mass-corrected GC with IC-5.
fig. S9. Mass-corrected GC with IC-6.
fig. S10. Mass-corrected GC with IC-7.
fig. S11. Mass-corrected GC with IC-8.
fig. S12. Mass-corrected gas chromatogram with 0 equiv 1-hexene.
fig. S13. Mass-corrected gas chromatogram with 0.3 equiv 1-hexene.
fig. S14. Mass-corrected gas chromatogram with 1 equiv 1-hexene.
fig. S15. Mass-corrected gas chromatogram with 1.5 equiv 1-hexene.
fig. S16. Calculated boiling point curves of RME product blends after isomerizing
cross-metathesis with different amounts 1-hexene, along with pure RME and
petrodiesel.
fig. S17. Boiling point curves of commercial diesel and biodiesel (RME) before
and after isomerizing hexenolysis.
fig. S18. Experimental chain length distributions; MCL: 12.9; 14.4; 17.5.
fig. S19. Simulated distributions; turnover number (TON) M = 30,000; TON I =
7500; MCL: 10.3; 13.8; 17.3.
fig. S20. Simulated distributions; TON M = 30,000; TON I = 15,000; MCL: 10.4;
13.7; 16.8.
fig. S21. Simulated distributions; TON M = 30,000; TON I = 30,000; MCL: 10.5;
13.5; 16.4.
fig. S22. Simulated distributions; TON M = 20,000; TON I = 5,000; MCL: 10.2;
13.8; 17.5.
fig. S23. Simulated distributions; TON M = 40,000; TON I = 10,000; MCL: 10.3;
11.7; 15.1.
fig. S24. Mass-corrected gas chromatogram of the mixture obtained by sequential
isomerizing ethenolysis.
fig. S25. Additional Ru-based metathesis catalysts tested in the isomerizing
ethenolysis.
fig. S26. Raw gas chromatograms of the product mixture obtained by single-step
isomerizing ethenolysis before and after hydrogenation.
fig. S27. Mass-corrected gas chromatogram of the product mixture obtained by
single-step isomerizing ethenolysis.
fig. S28. Boiling point curves of commercial diesel and biodiesel (RME) before
and after isomerizing ethenolysis.
fig. S29. Experimental chain length distributions; MCL: 12.3; 13.2; 15.7.
fig. S30. Simulated distributions; TON M = 30,000; TON I = 15,000; MCL: 8.9;
12.3; 15.7.
fig. S31. Experimental chain length distributions; MCL: 12.3; 11.8; 13.9.
fig. S32. Simulated distributions; TON M = 15,000; TON I = 3000; MCL: 7.6;
10.6; 13.6.
table S1. Product distributions obtained experimentally by isomerizing
hexenolysis of RME.
table S2. Equilibrium product distributions calculated for the isomerizing
hexenolysis of RME.
table S3. Comparison of product distributions obtained from isomerizing
hexenolysis of RME.
table S4. Optimization of the one-step isomerizing ethenolysis of RME.
table S5. EN ISO 3405 distillation data of isomerizing metathesis reactions with
RME.
Legend for movie S1
Legend for data file S1
References (37–49)
Other Supplementary Material for this manuscript includes the following:
(available at advances.sciencemag.org/cgi/content/full/3/6/e1602624/DC1)
movie S1 (.mp4 format). Webra “Winner” 2.5-cm3 self-igniting model diesel
engine operated with the fuel obtained via isomerizing ethenolysis of rapeseed
methyl ester.
data file S1 (.m format). MatLab simulation.
Supplementary Materials and Methods
General Methods and Chemicals
All reactions were performed in oven-dried glassware containing a Teflon-coated stirring
bar. Gas chromatography (GC) analyses were carried out using a HP 6890 with an HP-5
capillary column (Phenyl Methyl Siloxane 30 m x 320 x 0.25, 100/2.3-30-300/3) and a
time program beginning with 4 min at 60°C, followed by a 2°C/min ramp to 300°C, then
10 min at this temperature. Retention times of the substances were determined using
reference substances. The Envantage Dragon SimDist© program was used for the
simulation of boiling point curves from gas chromatograms (37). Commercial substrates
were used as received unless otherwise stated. 1-Hexene was used in 99% purity without
further purification. Ethylene was used in N4.5 (99.995%) and N3.5 (99.95%) purity. The
metathesis catalysts used herein are available commercially, for example, from Sigma-
Aldrich or Umicore. Ru-1, 1,3-bis(mesityl)-2-imidazolidinylidene]-[2-[[(2-
methylphenyl)imino]-methyl]-phenolyl]-[3-phenyl-indenyliden]-ruthenium-(II)chloride,
Umicore M42, CAS-no. 934538-12-2; Ru-2, 1,3-bis(2,4,6-trimethylphenyl)-2-
imidazolidinylidene-[2-[[(4-methylphenyl)imino]-ethyl]-4-nitrophenolyl]-[3-phenyl-1H-
inden-1-ylid-ene]ruthenium(II)chloride, Umicore M41, CAS-no. 934538-04-2; Ru-3,
1,3-bis(2,4,6-trimethylphenyl)-2-imidazolidinylidene]dichloro-(3-phenyl-1H-inden-1-
ylidene)(pyridyl)-ruthenium(II), Umicore M31, CAS-no. 1031262-76-6; Ru-4, [1,3-
Bis(2,4,6-trimethylphenyl)-2-imidazolidinylidene]dichloro[[2-(1-methyl-2-
oxopropoxy)phenyl]methylene]-ruthenium(II), Umicore M51, CAS-no. 1031262-71-1,
Ru-5, [1,3-Bis(2,4,6-trimethylphenyl)-2-imidazolidinylidene] dichloro[(2-isopropoxy)(5-
pentafluorobenzoylamido) benzylidene]ruthenium(II), Umicore M72 SIMes, CAS-no.
1030618-02-0; Ru-6 [1,3-Bis(2,4,6-trimethylphenyl)-2-imidazolidinylidene]dichloro[5-
(2-ethoxy-2-oxoethanamido)-2-isopropoxy-benzylidene]ruthenium(II), Umicore M74
SIMes, CAS-no. 1030618-11-1; Ru-7 [1,3-Bis(2,4,6-trimethylphenyl)-2-
imidazolidinylidene]dichloro(2-iodophenylmethylene)ruthenium(II), Umicore M91,
CAS-no. 1415725-62-0; Ru-8 [1,3-Bis(2,4,6-trimethylphenyl)-2-
imidazolidinylidene]dichloro[(8-iodo-1-naphtalene)methylene]ruthenium(II), Umicore
M92, CAS-no 1415725-73-3; Ru-9 [1,3-Bis(2,4,6-trimethylphenyl)-2-
imidazolidinylidene]dichloro[(2-bromo-5-
dimethylamino)phenylmethylene]ruthenium(II), Umicore M93, CAS-no. 1415725-68-6;
Ru-10 cis-[1,3-Bis(2,4,6-trimethylphenyl)-2-imidazolidinylidene]dichloro(3-phenyl-1H-
inden-1-ylidene)(triisopropylphosphite)ruthenium(II), Umicore M22, CAS-no. 1255536-
61-8; Ru-11 [1,3-Bis(2,6-diisopropylphenyl)-2-imidazolidinylidene]dichloro[5-
(isobutoxycarbonylamido)-2-isopropoxybenzylidene]ruthenium(II), Umicore M73 SIPr,
CAS-No. 1212009-05-6; Ru-12 [1,3-Bis(2,4,6-trimethylphenyl)-2-
imidazolidinylidene]dichloro[5-(isobutoxycarbonylamido)-2-
isopropoxybenzylidene]ruthenium(II), Umicore M73 SIMes, CAS-no. 1025728-57-7;
Ru-13 [1,3-Bis(2,4,6-trimethylphenyl)-2-imidazolidinylidene]dichloro[(2-isopropoxy)(5-
trifluoroacetamido)benzylidene]ruthenium(II), Umicore M71 SIMes, CAS-no. 1025728-
56-6; Ru-14 [1,3-Bis(2,4,6-trimethylphenyl)-2-imidazolidinylidene]dichloro(3-phenyl-
1H-inden-1-ylidene)(4,5-dichloro-1,3-diethyl-1,3-dihydro-2H-imidazol-2-
ylidene)ruthenium(II), Umicore M81 SIMes, CAS-no. 1228169-92-3; Ru-CAAC [[1-
[2,6-bis(1-methylethyl)phenyl]-3,3,5,5-tetramethyl-2- pyrrolidinylidene]dichloro[[2-(1-
methylethoxy-κO)phenyl]methylene-κC]ruthenium(II), CAS-no. 959712-80-2 was
prepared according to (36); Hoveyda Grubbs I catalyst: Dichloro(o-
isopropoxyphenylmethylene)(tricyclohexylphosphine)ruthenium(II), CAS-no. 203714-
71-0. The isomerization catalyst IC-1 can be obtained, for example, from Sigma-Aldrich:
Bromo(tri-tert-butylphosphine)-palladium(I)dimer, CAS-no. 185812-86-6. IC-6
Acetonitrile(cyclopentadienyl)[2-(di-i-propylphosphino)-4-(t-butyl)-1-methyl-1H-
imidazole]ruthenium(II) hexafluorophosphate, CAS-no. 930601-66-4; IC-7
bis[(1,2,3,4,5-η)-2,4-cyclooctadien-1-yl]hydro-ruthenium tetrafluoroborate, CAS-no.
104390-14-9; IC-8 Chloro(1-phenylindenyl)bis(triphenylphosphine)ruthenium(II), CAS-
no. 1360949-97-8. Hoveyda Grubbs I catalyst Dichloro(o-
isopropoxyphenylmethylene)(tricyclohexylphosphine)ruthenium(II), CAS-no. 203714-
71-0.
Supplementary Text
Preparation of Rapeseed Methyl Ester (RME)
Transesterification was performed by adding commercial rapeseed oil (500 mL) to a
stirred solution of NaOH (2.00 g, 50.0 mmol) in methanol (500 mL). The solution was
vigorously stirred and heated to reflux for 3 h. After cooling to 25°C, the methanolic
glycerin phase was separated and residual methanol was removed from the crude product
by distillation. After washing with cold water (2 x 100 mL), pure RME was obtained by
distillation (1x10-3 mbar, 160°C) and stored under an argon atmosphere.
Identification of a catalyst system for the model reaction of RME and 1-hexene
Objective. In this series of experiments, we investigated catalyst systems for the
conversion of RME into a distribution of olefins, mono-, and diesters by isomerizing
metathesis. To identify a catalyst system effective for the isomerizing cross-metathesis of
RME and 1-hexene, catalysts and conditions were systematically tested. 1-Hexene was
added in order to reduce the mean chain lengths from 18 carbon atoms (excluding the
methyl ester carbon).
General procedure for the isomerizing hexenolysis of RME
In a glovebox under nitrogen atmosphere, an oven-dried 20 mL headspace vial with a
Teflon-coated stirring bar was charged with IC-cat., Ru-cat., RME and 1-hexene. The
vial was closed with a Teflon-coated crimp cap, removed from the glovebox, and the
mixture was stirred at 50°C for 20 h. The reaction mixture was cooled to 25°C, opened
and diluted with EtOAc (3 mL). GC analysis of the raw mixture gave the required
chromatogram for the SimDist boiling point curve simulation (see chapter Simulated
boiling curves of the isomerizing hexenolysis product blends).
To obtain simplified GC chromatograms, the mixture was then hydrogenated, converting
all double-bond isomers of a given olefinic product into a single saturated compound.
Hydrogenation of the diluted product mixture was carried out in the presence of Pd/C
(0.1 mmol) and hydrogen (10 bar) at 50°C for 12 h. After releasing the hydrogen
pressure, the sample was filtered over celite and MgSO4 and analyzed by gas
chromatography. The signals of the chromatograms were corrected for molecular weight
(38) and assigned by GC-MS.
Screening of Ru-based metathesis catalysts
The isomerizing hexenolysis of RME (1.65 g, 5.00 mmol, 1.88 mL) with 1-hexene (420
mg, 5.00 mmol, 558 µL) was carried out following the general procedure using IC-1
(3.89 mg, 5.00 µmol) and Ru-cat. (5.00 µmol, fig. S1). The product mixtures were
hydrogenated and analyzed by GC. For each Ru-cat., the signals of the olefin fraction
were selected from the chromatogram and corrected for molecular weight. They are
visualized in the superimposed histograms of the olefin fractions in fig. S2.
fig. S1. Ru-based metathesis catalysts tested in the isomerizing hexenolysis,
including second-generation indenylidene-ruthenium complexes Umicore M41 (Ru-
2) and M31 (Ru-3) and Hoveyda-type catalysts Umicore M51 (Ru-4), M72 SIMes
(Ru-5), and M74 SIMes (Ru-6).
fig. S2. Olefin blends obtained by isomerizing hexenolysis with different Ru
catalysts. Mono- and diester fractions displayed an analogous distribution.
In the olefin fractions shown in fig. S2, the non-isomerized metathesis products decene,
tetradecene and octadecene are overrepresented for all catalyst systems except for the IC-
1 / Ru-1 combination (black bars with superimposed black trend line). This indicates that
all other metathesis catalysts adversely affect the isomerization activity of IC-1.
Screening of isomerization catalysts
The isomerizing hexenolysis of RME (1.65 g, 5.00 mmol, 1.88 mL) with 1-hexene (434
mg, 5.00 mmol, 640 µL) was carried out following the general procedure using IC-cat.
(5.00 µmol, fig. S3) and Ru-1 (4.22 mg, 5.00 µmol). The hydrogenated product fractions
are shown in figs. S4-S11.
fig. S3. State-of-the-art isomerization catalysts tested in the isomerizing hexenolysis.
For a detailed description of the catalysts see:
IC-1: (19, 39-42).
IC-2, IC-3: [Rh]/BiPhePhos = 1:10, (43-45).
IC-4: (31, 46, 47).
IC-5: (33).
IC-6: (26, 32).
IC-7: (30).
IC-8: (48, 49).
fig. S4. Mass-corrected GC with IC-1.
fig. S5. Mass-corrected GC with IC-2.
fig. S6. Gas chromatogram with IC-3. Conditions: 2.5 mmol methyl oleate, 0.50 mol%
IC-3, 0.20 mol% Ru-1, neat, 20 h, 45°C.
fig. S7. Gas chromatogram with IC-4. Conditions: 2.5 mmol methyl oleate, 0.50 mol%
IC-4, 0.20 mol% Ru-1, neat, 20 h, 70°C.
fig. S8. Mass-corrected GC with IC-5. (Sample hydrogenated prior to GC analysis).
fig. S9. Mass-corrected GC with IC-6. (Sample hydrogenated prior to GC analysis).
fig. S10. Mass-corrected GC with IC-7. (Sample hydrogenated prior to GC analysis).
fig. S11. Mass-corrected GC with IC-8. (Sample hydrogenated prior to GC analysis).
Figures S4 to S11 show that IC-1 was the only isomerization catalyst to be compatible
with the ruthenium metathesis catalyst.
Result. A combination of isomerization catalyst IC-1 (0.05 mol%) and metathesis
catalyst Ru-1 (0.05 mol%) effectively mediates the isomerizing cross-metathesis of RME
and 1-hexene at 50°C with a reaction time of 20 h.
Optimization of the boiling behavior with different 1-hexene/RME ratios
Objective. The effect of different 1-hexene / RME ratios on the boiling behavior of the
resulting olefin blends was investigated, with the goal of optimizing the mean boiling
temperature, initial and terminal boiling points of the curve.
Synthesis of hexenolysis products
Several olefin blends were synthesized using the optimal catalyst system IC-1 and Ru-1
following the general procedure of the isomerizing hexenolysis using different amounts
of 1-hexene (0, 0.3, 1.0, 1.5 equivalents).
The gas chromatograms obtained for the resulting product fractions after hydrogenation
show a shift of the distributions towards lower chain lengths (figs. S12 to S15).
fig. S12. Mass-corrected gas chromatogram with 0 equiv 1-hexene.
fig. S13. Mass-corrected gas chromatogram with 0.3 equiv 1-hexene.
fig. S14. Mass-corrected gas chromatogram with 1 equiv 1-hexene.
fig. S15. Mass-corrected gas chromatogram with 1.5 equiv 1-hexene.
The olefin / monoester / diester ratio increases with increasing amounts of 1-hexene,
while the carbon-chain lengths of the largest peaks and the mean chain lengths of the
distributions decrease (table S1).
table S1. Product distributions obtained experimentally by isomerizing hexenolysis
of RME.
[1-hexene] / [RME]
Mean chain lengths found Olefins Monoesters Diesters
0.0 <16.8a 17.5 18.8 0.3 <14.8a 16.1 18.3 1.0 <12.9a 14.4 17.5 1.5 <12.4a 14.0 17.6
a) Volatile olefins < C8 not integrated.
Calculated boiling curves of the isomerizing hexenolysis product blends
Boiling point curves were calculated from GC data of the raw product mixtures using the
Envantage Dragon SimDist© software (37). The software specifies a GC method for
sample analysis. It compares the GC signal areas at given retention times with the boiling
points of reference substances with known retention times and from this, calculates the
boiling point curve. This allows predicting the overall boiling behavior of the product
mixture (ca. 5 mmol/1.6 mL reaction volume) without the need to synthesize large
quantities of material.
The ability of the calculations to predict the atmospheric distillation results is limited by
the absence of reference material at high recovery, as well as the low intensity and low
response factor of GC signals at high temperatures. Values >90% recovery are
extrapolated by SimDist from the signal area of the GC sample. As a result, the intrinsic
drawback of this method are the unrealistic steps within the curves, and the steep rise at
>95% recovery.
To account for this limitation, the boiling point curves were also calculated for pure RME
and petrodiesel GC samples for comparison (fig. S16).
fig. S16. Calculated boiling point curves of RME product blends after isomerizing
cross-metathesis with different amounts 1-hexene, along with pure RME and
petrodiesel.
All curves displayed the desired evenly rising shape (fig. S16). With increasing
1-hexene / RME ratio, the mean boiling points gradually shift towards lower
temperatures and the slope becomes more even. At low 1-hexene / RME ratios, the
predicted boiling points were clearly too high at >80% recovery. However, the calculated
boiling point curve for the reaction of RME with 1.0 eq. of 1-hexene approximated that
of the diesel fuel reference even at higher recoveries. Increasing the amount of 1-hexene
further did not result in a significant shift of the boiling point curve.
Large-scale procedure for the isomerizing hexenolysis of RME
In order to measure the atmospheric boiling behavior of the product mixture, the reaction
was performed on large scale.
In a glovebox under nitrogen atmosphere, a 100 mL Büchi bmd 075 miniclave drive
autoclave was charged with RME (26.4 g, 30.1 mL, 80.0 mmol), 1-hexene (6.73 g, 9.93
mL, 80.0 mmol), IC-1 (62.2 mg, 80.0 µmol) and Ru-1 (67.5 mg, 80.0 µmol), then
removed from the glovebox. The resulting reaction mixture was stirred for 18 h at 50°C.
The reactor was cooled to ambient temperature and a 30% solution of H2O2 (16.3 mL,
160 mmol) was slowly added at 0°C under vigorous stirring to remove the catalyst (1000
rpm, overhead stirrer). The organic phase was separated, dried over 3 Å molecular sieves
and filtered over a short column of celite and MgSO4. Because RME is a mixture of
different compounds, it is only possible to give a volume-based yield. Starting from 30.1
mL RME and 9.93 mL 1-hexene, 32 mL of the isomerizing metathesis blend was isolated
(80%).
Atmospheric distillation of the product provided the boiling point curve shown in fig.
S17.
fig. S17. Boiling point curves of commercial diesel and biodiesel (RME) before and
after isomerizing hexenolysis. The hashed areas represent the limits specified in EN
590.
* = smoke formation.
Result. A 1-hexene / RME ratio of 1:1 was determined to provide the optimal boiling
behavior of the product blend. The resulting boiling point curve closely matches that of
petrodiesel in terms of initial and mean boiling point. It still has deficits in the final
recovery percentages, because it crosses the specification limits and progressively
decomposes with smoke formation above 360°C.
Simulation of the isomerizing hexenolysis reaction mixtures
Objective. A simulation program was developed with the aim of understanding the
factors related to stoichiometry and catalyst performance that affect product distributions
in reaction mixtures at any stage before reaching equilibrium.
General methods for the simulations
All simulations were generated using MathWorks MATLAB R2014b. The simulation
code is available as a separate .m file in the supplementary material or directly from the
authors. In the simulation program, a given number of randomly chosen molecules from a
mixture of methyl oleate/hexene at a given ratio undergo a single shift of their double
bond. Then, another given number of randomly chosen molecules undergo metathesis.
These two steps are iterated a given number of times, so that different overall and relative
turnover numbers for both catalyst systems can be modelled.
The simulated histograms were then compared with the experimental data, and the mean
chain lengths (MCL) of the simulated distributions were calculated by weighted
arithmetic mean.
The simulated boiling point curves (Fig. 3 in the main manuscript) were calculated from
the simulated product distributions. The boiling point curves could not be calculated with
SimDist, because this requires GC input data. Instead, the boiling points of all simulated
product molecules were either taken from the literature or extrapolated from literature
values, and were used as input to calculate the boiling point curves using a Visual Basic
macro. The macro file is available from the authors upon request.
The simulation output obtained with this macro was calibrated with experimental results
obtained in a series of model reactions and SimDist results for the respective sample.
Simulations for a 1:1 mixture of methyl oleate and 1-hexene
The experimentally observed product fractions, i.e. olefins (blue bars), mono- (red bars),
and diesters (green bars), were plotted as separate histograms. These were overlaid with
normalized integrals for the simulated product fractions (solid black lines) that were set
to 1 for each fraction (figs. S18 - S23).
Since olefins (blue) <C8 were not detected via GC, the normalized integrals were too
high compared to the simulated solid black curve, which also covered the olefins <C8.
The simulated curves were scaled to account for this issue (dotted black lines).
The simulation was run with different turnover numbers for the isomerization (TON I)
and metathesis (TON M). For the isomerizing hexenolysis, the amount of olefin
molecules was set to 2,000 per catalyst molecule. This equals the experimental catalyst
loading of 0.05 mol%.
For the isomerizing hexenolysis, the best fit was achieved for turnover numbers of 30,000
for the metathesis and 7,500 for the isomerization catalyst (fig. S19). This translates to 15
metathesis and 3.75 isomerization steps per molecule. The simulated histogram in fig.
S19 demonstrates that the reaction had not quite reached equilibrium but was close to
becoming homogeneous.
fig. S18. Experimental chain length distributions; MCL: 12.9; 14.4; 17.5.
fig. S19. Simulated distributions; turnover number (TON) M = 30,000; TON I =
7500; MCL: 10.3; 13.8; 17.3.
fig. S20. Simulated distributions; TON M = 30,000; TON I = 15,000; MCL: 10.4;
13.7; 16.8.
fig. S21. Simulated distributions; TON M = 30,000; TON I = 30,000; MCL: 10.5;
13.5; 16.4.
fig. S22. Simulated distributions; TON M = 20,000; TON I = 5,000; MCL: 10.2;
13.8; 17.5.
fig. S23. Simulated distributions; TON M = 40,000; TON I = 10,000; MCL: 10.3;
11.7; 15.1.
Calculation of the carbon-chain length distribution
The carboxylate carbons do not take part in the transformation and is therefore omitted in
all calculations. For example, methyl oleate has 19 carbon atoms but is calculated as C18.
Close to equilibrium, the product distribution is determined by the relative abundance of
C–C double bonds, functionalized and unfunctionalized chain termini. Thus, for the
isomerizing hexenolysis of pure C18 monounsaturated fatty acid esters, the calculated
ratio of olefins, mono- and diesters is (¾ * ¾) : (¾ * ¼ * 2) : (¼ * ¼) = 9:6:1.
The mean carbon-chain lengths of the olefin, mono- and diester fractions at equilibrium
were calculated from the ratio of the unfunctionalized and ester termini and the ratio of
carbon atoms per double bond. The following formula was used, with a, b, c, X and Y
according to:
Ø (b+c) = (Y*15 + X*a) / (X + Y)
For example, in the isomerizing hexenolysis with 1.0 eq. 1-hexene X = Y = 1 and
a = 4. Therefore Ø (b+c) = 9.5. This still excludes the two carbons of the double bond, so
that the mean chain length of the olefin fraction is 9.5 + 2 = 11.5. For the full chain
lengths of the mono- and diester fractions, 1 and 2 carbons, respectively, need to be
added to this value to include the carboxylate termini (table S2).
table S2. Equilibrium product distributions calculated for the isomerizing
hexenolysis of RME.
[1-hexene] / [RME]
Mean chain lengths equilibrium Olefins Monoesters Diesters
0.0 17.0 18.0 19.0 0.3 14.5 15.5 16.5 1.0 11.5 12.5 13.5 1.5 10.4 11.4 12.4
Simulation of the isomerizing metathesis at high turnover numbers for both isomerization
and metathesis (150 isomerization and metathesis steps each per olefin molecule)
provided MCL values that match the ones obtained by the formula of the weighted
arithmetic mean above (table S2). This indicates that 150 isomerization and metathesis
steps each per olefin molecule are enough to ensure equilibrium distributions.
Compared with the experimentally obtained results of <12.9 carbons for the olefins
(excluding volatiles not detectable by GC analysis), 14.4 for the monoesters and 17.5 for
the diesters (see table S1), the calculated values at equilibrium are smaller for all
fractions (MCL: 11.5; 12.5; 13.5.). This shows how the experimental conditions, that are
not fully accounted for by the assumptions made in the simulation, affect the distributions
in practice:
RME contains saturated methyl palmitate (4%) and stearate (1%), which cannot
undergo isomerizing metathesis. This explains the protruding signals of C16 and
C18 esters and increases the proportion and mean chain length of the monoester
fraction.
The different stability and reaction rates of certain olefinic compounds, e.g. terminal
or conjugated systems, are not accounted for in the calculations.
The partial evaporation of short-chain olefins from the mixture leads to larger mean
chain lengths for all fractions.
The simulated curve that best fits the experimental distribution for the isomerizing
hexenolysis of methyl oleate was obtained with 15 metathesis and 3.75 isomerization
steps per molecule. These are substantially fewer steps than required to reach
equilibrium. This best-fit simulated curve translates to mean chain lengths of 10.3, 13.8
and 17.3 for olefins, mono- and diesters respectively (table S3, fig. S19).
table S3. Comparison of product distributions obtained from isomerizing
hexenolysis of RME.
[1-hexene] / [RME]
Mean chain lengths equilibrium / fitted curvea / found
Olefins Monoesters Diesters
0.0 17.0 / 18.0 / <16.8b 18.0 / 18.0 / 17.5 19.0 / 18.0 / 18.8 0.3 14.5 / 13.8 / <14.8b 15.5 / 15.7 / 16.1 16.5 / 17.7 / 18.3 1.0 11.5 / 10.3 / <12.9b 12.5 / 13.8 / 14.4 13.5 / 17.3 / 17.5 1.5 10.4 / 9.2 / <12.4b 11.4 / 13.1 / 14.0 12.4 / 17.1 / 17.6
a) Simulated values for 2,000 olefin molecules per catalyst after 30,000 metathesis and
7,500 isomerization steps. b) Excluding volatile olefins < C8.
In all cases, simulated mean chain lengths depart only slightly from the experimentally
observed values. The remaining deviation may be a result of the factors discussed above.
The simulations for higher isomerization rates show that although the mean chain length
of the diester fraction is shifted towards lower values, a broadening of the mono- and
diester fractions results in an undesirable overall increase of the fraction boiling at high
temperatures >380°C.
Result. The simulation program provides curves that are adjustable with regard to
substrate stoichiometry and turnover of both the isomerization and metathesis catalyst.
Development of the isomerizing ethenolysis of RME
Objective. To explore the possibility of employing gaseous ethylene in isomerizing
metathesis reactions with the goal of producing an EN 590-compatible fuel.
Procedure for the sequential isomerizing ethenolysis
A comparison of the isomerizing cross-metathesis using ethylene to the results previously
obtained with 1-hexene required a process in which an ethylene / RME stoichiometry of
close to 1:1 was reproducibly ensured. This also allowed calibrating the subsequent
simulations. It was achieved by non-isomerizing ethenolysis of RME. The product
mixture was then subjected to isomerizing metathesis for further reaction development.
Ethenolysis of RME. Under an atmosphere of argon, a 1 L stirred Parr autoclave was
charged with Hoveyda Grubbs I catalyst (3.60 g, 6.00 mmol) and rapeseed oil methyl
ester (178 g, 200 mL, 600 mmol). The vessel was pressurized with 10 bar ethylene and
stirred for 18 h at ambient temperature. The reactor was cooled to -20°C and the ethylene
pressure was slowly released. After warming up to ambient temperature, the reaction
mixture was filtered over silica and distilled under vacuum (1x10-3 mbar, up to 250°C),
yielding a mixture of 1-decene and methyl decenoate : methyl oleate : dimethyl octadec-
9-enedionate with the ratio 82.7 : 10.0 : 5.00, along with small amounts of additional
olefins and saturated components of RME.
Isomerizing metathesis. In a glovebox under nitrogen atmosphere, a 100 mL Büchi bmd
075 miniclave drive autoclave was charged with the previously prepared ethenolysis
mixture (54.0 g, 60 mL), IC-1 (303 mg, 390 µmol) and Ru-1 (329 mg, 390 µmol). The
resulting reaction mixture was stirred for 18 h at 50°C. The reactor was cooled to ambient
temperature and a 30% solution of H2O2 (27.6 mL, 270 mmol) was slowly added at 0°C
under vigorous stirring (1000 rpm, overhead stirrer). The organic phase was separated,
dried over 3 Å molecular sieves and filtered over a short column of celite and MgSO4.
Because RME and therefore the ethenolysis product are mixtures of different compounds,
it is only possible to give a volume-based yield. Starting from 60 mL of the ethenolysis
mixture, 50 mL of the isomerizing metathesis blend was isolated (83%). The resulting
product fraction is shown in fig. S24.
fig. S24. Mass-corrected gas chromatogram of the mixture obtained by sequential
isomerizing ethenolysis. (sample hydrogenated for GC analysis).
Optimization of the one-step isomerizing ethenolysis of RME
Table S4 shows the optimization of an isomerizing ethenolysis of RME as a one-step
procedure. Various volumes of ethylene were added by pressurizing a closed vessel, by
attaching an ethylene balloon, or by passing a stream of ethylene through a rapidly stirred
reaction mixture containing 2.50 mmol RME (based on methyl oleate) and a catalyst
system in the absence of solvent at 60°C for 16 h. The product mixtures were
hydrogenated and analyzed by GC.
table S4. Optimization of the one-step isomerizing ethenolysis of RME.
table S4 (continued).
Entry Ru-cat.
Ru-cat. (mol%)
Ru-CAAC (mol%)
IC-1 (mol%)
Av. MCLa
Homogeneous distribution
Gas Volume (mL)
1b Hexenolysis (see above) 15.0 Yes - 2b Seq. ethenolysis (see above) 13.7 Yes -
3b,c Ru-1 0.05 - 0.05 - No conversion 300 4 Ru-1 0.10 - 0.10 18.0 No 20 5 Ru-3 0.10 - 0.10 16.1 No 20 6 Ru-4 0.10 - 0.10 15.7 Almost 20 7 Ru-5 0.10 - 0.10 15.3 Almost 20 8 Ru-6 0.10 - 0.10 15.7 Almost 20 9 Ru-7 0.10 - 0.10 15.2 Yes 20
10 Ru-8 0.10 - 0.10 16.3 Almost 20 11 Ru-9 0.10 - 0.10 15.8 No 20 12 Ru-10 0.10 - 0.10 17.9 No 20 13 Ru-11 0.10 - 0.10 15.3 Almost 20 14 Ru-12 0.10 - 0.10 15.3 No 20 15 Ru-13 0.10 - 0.10 15.9 Almost 20 16 Ru-14 0.10 - 0.10 16.8 No 20 17 Ru-5 0.10 - 0.40 15.8 Almost 20 18 Ru-7 0.10 - 0.40 15.5 Yes 20 19 Ru-11 0.10 - 0.40 15.4 Yes 20 21 Ru-5 0.10 0.10 0.40 15.2 Yes 20 22 Ru-7 0.10 0.10 0.40 16.0 Yes 20 23 Ru-11 0.10 0.10 0.40 15.3 Yes 20 24 Ru-5 0.10 0.10 0.40 13.2 Yes 300 25 Ru-7 0.10 0.10 0.40 12.7 Yes 300 26 Ru-11 0.10 0.10 0.40 12.5 Yes 300 27c Ru-11 0.10 0.10 0.40 14.4 Yes 300 28d Ru-11 0.10 0.10 0.40 12.9 Yes Stream 29d Ru-11 0.10 - 0.40 15.1 Yes Stream 30 - - 0.10 0.40 14.1 No 300
a) Av. MCL = The average of the mean chain length of all three product classes (olefins,
mono-, diester) indicates how far the overall product distribution has shifted towards
lower chain lengths. The average MCL for the isomerizing self-metathesis of RME
without additional olefins is 18; b) Reaction at 50°C; c) 6 bar ethylene pressure; d) 45
mmol-scale under constant stream of ethylene at atmospheric pressure.
The average mean chain length (av. MCL) of all three product fractions is a measure of
the ethylene intake and the shift of the product distributions towards lower boiling points
(compare figs. S12-S15 for the effect with 1-hexene).
Without ethylene, the average MCL is 18 (isomerizing self-metathesis of RME). Adding
short-chain olefins, for example in the isomerizing hexenolysis and the sequential
isomerizing ethenolysis, led to lower MCL values (entries 1, 2). In addition, we
compared the shape of the product distributions obtained. The system Ru-1/IC-1 was
inactive under ethylene pressure, possibly due to decomposition of the ruthenium
complex (entry 3). Conversion was first observed after switching to atmospheric ethylene
pressure (entry 4). The catalyst loading was increased to 0.10 mol% to compensate for
lower activity and stability in the presence of ethylene.
fig. S25. Additional Ru-based metathesis catalysts tested in the isomerizing
ethenolysis.
Screening of different ruthenium-NHC complexes (entries 4-16, figs. S1 and S25)
identified three catalysts (Ru-5, Ru-7, Ru-11) that mediate the isomerizing metathesis of
RME in the presence of ethylene with homogeneous product distributions and decreased
average MCLs (entries 7, 9, 13). At an IC-1 loading increased to 0.4 mol%, the
homogeneity of the distributions increased, and protruding cross-metathesis products
were no longer observed (entries 17-19).
Addition of the specialized ethenolysis catalyst Ru-CAAC as well as increasing the
ethylene volume to 300 mL resulted in a greater ethylene incorporation into the products
and shorter av. MCL down to 12.5 (entries 21-26). An increase of the ethylene pressure
to 6 bar had an adverse effect (entry 27, 300 mL gas volume at 6 bar). The best results
were obtained when passing a constant stream of ethylene through the reaction vessel
(entry 28). Control experiments confirmed the necessity of all catalyst components
(entries 29, 30).
All reactions in table S4 were performed with N4.5 (99.995%) ethylene. When
employing N3.5 (99.95%) ethylene in the optimized reaction (entry 26), the distributions
and mean chain lengths were identical.
Screening of isomerization catalysts in the isomerizing ethenolysis
In a glovebox under nitrogen atmosphere, an oven-dried 10 mL headspace vial with a
Teflon-coated stirring bar was charged with IC-cat. (16.0 µmol, fig. S3), Ru-11. (3.31
mg, 4.0 µmol) and RME (1.19 g, 4.00 mmol, 1.35 mL). The vial was closed with a
Teflon-coated crimp cap with a gas inlet and placed in an autoclave. The autoclave was
removed from the glovebox, evacuated (10-3 mbar), pressurized with ethylene (6 bar) and
the mixture was stirred at 60°C for 16 h. The autoclave was cooled to 0°C, opened and
the reaction mixture was diluted with EtOAc (3 mL).
To obtain simplified GC chromatograms, the mixture was then hydrogenated and
analyzed by GC as described in the standard procedure.
Isomerization catalysts IC-1, IC-3, IC-4, IC-6, IC-7 and IC-8 (fig. S3) were tested under
the conditions mentioned above. IC-1 was the only isomerization catalyst to give
homogeneous product distributions. With all other catalysts, predominantly cross
metathesis products were observed.
Optimized procedure for the single-stage isomerizing ethenolysis
In a glovebox under nitrogen atmosphere, a 30 mL glass reactor was charged with Ru-
CAAC (30.3 mg, 50.0 µmol), IC-1 (155 mg, 200 µmol), Ru-11 (41.3 mg, 50.0 µmol)
and RME (16.9 mL, 50 mmol based on methyl oleate). The resulting reaction mixture
was stirred under a stream of ethylene at atmospheric pressure for 16 h at 60°C. Two
such batches were combined, cooled to ambient temperature and a 30% solution of H2O2
(5.11 mL, 50 mmol) was slowly added at 0°C under vigorous stirring. The organic phase
was separated, dried over MgSO4 and filtered over a short column of celite and MgSO4,
yielding 25 mL of a brown oil (74% based on volume). After high-temperature vacuum
distillation (10-3 mbar, >350°C), 24 mL of the product mixture were obtained as a light
yellow liquid (96% recovery after distillation).
Reaction scale-up for the single-stage isomerizing ethenolysis
In a glovebox under nitrogen atmosphere, a 1 L Parr autoclave was charged with Ru-
CAAC (243 mg, 0.40 mmol), IC-1 (1.24 g, 1.60 mmol), Ru-11 (330 mg, 0.40 mmol) and
RME (135 mL, 400 mmol based on methyl oleate). The resulting reaction mixture was
stirred under a stream of ethylene at atmospheric pressure for 16 h at 60°C. The reactor
was cooled to ambient temperature and a 30% solution of H2O2 (40.9 mL, 400 mmol)
was slowly added at 0°C under vigorous stirring. The organic phase was separated, dried
over MgSO4 and filtered over a short column of celite and MgSO4, yielding 75 mL of a
brown oil (55% based on volume). After high-temperature vacuum distillation (1x10-3
mbar, >350°C), the product mixture was obtained as a light yellow liquid (73 mL, >98
wt-% recovery after distillation). A sample was analyzed by GC following hydrogenation
(fig. S26). The peaks were assigned by GC-MS and corrected for their mass to generate
the histogram in fig. S27.
fig. S26. Raw gas chromatograms of the product mixture obtained by single-step
isomerizing ethenolysis before and after hydrogenation.
fig. S27. Mass-corrected gas chromatogram of the product mixture obtained by
single-step isomerizing ethenolysis. (Sample hydrogenated for GC analysis).
Atmospheric distillation of the product provided the boiling point curve shown in fig.
S28. All experimental recovery temperatures are given in table S5.
fig. S28. Boiling point curves of commercial diesel and biodiesel (RME) before and
after isomerizing ethenolysis. The hashed areas represent the limits specified in EN 590.
Result. An effective ternary catalyst system was found that mediates the isomerizing
ethenolysis of RME, composed of the ethenolysis catalyst Ru-CAAC (0.1 mol%), the
isomerization catalyst IC-1 (0.4 mol%) and the metathesis catalyst Ru-11 (0.1 mol%),
which operates at atmospheric ethylene pressure. The product mixture obtained fulfils the
standard EN 590.
Simulation of the isomerizing ethenolysis mixtures
Objective. The isomerizing ethenolysis was simulated in order to determine the optimum
stoichiometry and catalyst turnovers.
The general methods for simulations described above were also applicable in this section.
Simulations for a 1:0.83 mixture of methyl oleate and ethylene
This simulation models the sequential isomerizing ethenolysis. It was parameterized for
an isomerizing ethenolysis starting with methyl oleate and ethylene. The ratio of methyl
oleate to ethylene was corrected for the result obtained in the non-isomerizing ethenolysis
of RME (83% ethenolysis products), which served as the starting material in the
isomerizing metathesis experiment. The simulated distributions were then compared to
the experimental data (figs. S29 and S30).
fig. S29. Experimental chain length distributions; MCL: 12.3; 13.2; 15.7.
fig. S30. Simulated distributions; TON M = 30,000; TON I = 15,000; MCL: 8.9;
12.3; 15.7.
Simulations for a 1:1.33 mixture of methyl oleate and ethylene
This simulation models the one-step isomerizing ethenolysis. The starting parameters
were set to a 1:1.33 mixture of methyl oleate and ethylene. Since the exact amount of
ethylene in the reaction mixture and the ethylene uptake could not be determined
experimentally under flow conditions, it was assumed that each double bond in the
starting material reacts with one molecule of ethylene. Based on the FAME composition
of RME (65% methyl oleate [18:1], 22% methyl linoleate [18:2], 8% methyl linolenate
[18:3], 1% methyl stearate [18:0], and 4% methyl palmitate [16:0]) it was calculated that
100 molecules of RME contain an average of 133 olefinic double bonds and react with
133 molecules of ethylene. Therefore, a ratio of 1:1.33 was used.
The simulation was set to 1000 olefinic molecules per catalyst molecule, which equals
the experimental catalyst loading of the metathesis catalyst (0.1 mol%). The best fit was
achieved with turnover numbers of 15,000 for the metathesis and 12,000 for the
isomerization catalyst (figs. S31 and S32).
It should be noted that the simulated turnover number for the isomerization cannot be
directly compared with the values calculated for the sequential isomerizing ethenolysis or
the isomerizing hexenolysis. The loading of the isomerization catalyst had to be increased
to 0.4 mol% in the experiment to compensate for lower catalyst activity in the presence of
ethylene. Different loadings of isomerization and metathesis catalysts cannot separately
be parameterized in the simulation algorithm. Therefore, the simulations were conducted
a catalyst loading of 0.1 mol% as described above. Therefore, TON I from the simulation
(12,000) was corrected for the experimental catalyst loading to 3,000.
fig. S31. Experimental chain length distributions; MCL: 12.3; 11.8; 13.9.
fig. S32. Simulated distributions; TON M = 15,000; TON I = 3000; MCL: 7.6; 10.6;
13.6.
For the overlays in Figs. 2 and 4 in the main manuscript, the simulated data from figs.
S19 and S32 were used. Both curves were smoothed by a Savitzky-Golay filter for clarity
and to suppress statistical fluctuation. All simulated curves in the supplementary
materials are unaltered and contain minimal statistical fluctuations.
Result. By fitting the simulated curves to the experimentally obtained product
distributions of the product that fulfils EN 590, the individual TON for the isomerization
and metathesis catalyst were determined to be TON M = 15,000 and TON I = 3,000.
Analysis of the physical properties of the product blends.
Atmospheric distillation (EN ISO 3405)
The isomerizing hexenolysis, the sequential and the one-step isomerizing ethenolysis of
RME were conducted on large scale as described above. 100 mL of each sample were
analyzed by atmospheric distillation in an Anton Paar ADU 4 ISO 3405 distillation
apparatus to determine the boiling point curve according to the standard DIN EN 590 for
petrodiesel (table S5). The standard defines three thresholds:
At 250°C less than 65% of the sample is collected.
At 350°C at least 85% of the sample is collected.
At least 95% of the sample is collected at a maximum of 360°C
table S5. EN ISO 3405 distillation data of isomerizing metathesis reactions with
RME.
Recovery (%)
Recorded temperature (°C)a Diesel ICM with
1-hexene Seq. ICM with
Ethylene One step ICM with ethylene
IBP 177.8 121.0 73.7 111.6 5 197.2 171.6 125.4 155.6
10 205.7 190.7 157.1 168.2 15 213.0 209.8 184.3 182.0 20 221.2 225.9 202.8 195.5 25 229.6 239.1 217.3 207.9 30 237.3 250.2 228.9 219.8 35 245.0 259.7 240.0 231.4 40 252.3 268.8 250.5 242.1 45 261.1 277.3 261.5 252.2 50 269.6 285.7 271.3 263.7 55 277.4 294.5 280.6 274.4 60 286.1 302.8 289.6 283.6 65 294.5 310.6 299.9 293.8 70 303.6 319.2 309.4 302.8 75 311.6 327.5 318.1 311.3 80 320.8 336.3 326.9 319.3 85 329.5 346.3 336.0 327.8 90 339.6 354.1 346.5 338.0 93 346.9 359.5 355.3 346.4 95 353.4 decomposition 362.4 354.1
FBP 364.3 - 366.2 357.8 IBP = Initial boiling point, FBP = final boiling point, ICM = isomerizing cross-
metathesis, seq. = sequential. a) Recorded temperature corrected for atmospheric
pressure.
The hexenolysis product showed a recovery of only 93% at 360°C, which narrowly
misses the specified value of 95%. Towards the end of the distillation, it partially
decomposed with smoke formation. This is a common problem for biodiesel, caused by
oxidation of sensitive polyunsaturated fatty acid derivatives, and is usually addressed by
partial hydrogenation of the product fractions.
In contrast, the boiling point curve for the one-step isomerizing ethenolysis product had a
lower mean boiling temperature, resulting in the specified recovery of 95% at a
maximum of 360°C, thus meeting the boiling specifications.
movie S1. Webra “Winner” 2.5-cm3 self-igniting model diesel engine operated with
the fuel obtained via isomerizing ethenolysis of rapeseed methyl ester. The small two-
stroke diesel engine has ca. 10 bar compression pressure (c.f. ca. 25 bar for full-size
diesel engines), so that self-ignition requires additives. The standard fuel recipe involves
ca. 30% petrodiesel, 30% diethyl ether and 2% amyl nitrite as ignition booster, and 30%
two-stroke oil as lubricant. Instead, we used 60% isomerizing ethenolysis product, 30%
diethyl ether, and 10% rapeseed oil to run the engine. As can be seen in the video, the
engine ran steadily even at the difficult minimal throttle. At higher throttle, the engine
produced sufficient thrust to move the model car.
data file S1. MatLab simulation. This contains the MatLab code for the mathematical
modeling of the product distributions. This .m file can be opened by MatLab R2014b or
visualized with any plain-text editor.