Fatty methyl ester hydrogenation to fatty alcohol part I: Correlation between catalyst properties and activity/selectivity

Article abstract of DOI:10.1007/s11746-997-0088-y

Fatty alcohols, derived from natural sources, are commercially produced by hydrogenation of fatty acids or methyl esters in slurry-phase or fixed-bed reactors. One slurry-phase hydrogenation of methyl ester process flows methyl esters and powdered copper chromite catalyst into tubular reactors under high hydrogen pressure and elevated temperature. In the present investigation, slurry-phase hydrogenations of C₁₂ methyl ester were carried out in semi-batch reactions at non-optimal conditions (i.e., low hydrogen pressure and elevated temperature). These conditions were used to accentuate the host of side reactions that occur during the hydrogenation. Some 14 side reaction routes are outlined. As an extension of this study, copper chromite catalyst was produced under a number of varying calcination temperatures. Differences in catalytic activity and selectivity were determined by closely following side reaction products. Both activity and selectivity correlate well with the crystallinity of the copper chromite surface; they increase with decreasing crystallinity. The ability to follow the wide variety of side reactions may well provide an additional tool for the optimized design of hydrogenation catalysts.

Full text of DOI:10.1007/s11746-997-0088-y

Fatty Methyl Ester Hydrogenation
to Fatty Alcohol Part I: Correlation Between
Catalyst Properties and Activity/Selectivity
a,
b
b
b
Ross D. Rieke *, Deepak S. Thakur , Brian D. Roberts , and Geoffrey T. White
a
Procter & Gamble Company, Sharon Woods Technical Center, Cincinnati, Ohio 45241,
and Engelhard Corporation, Research and Development, Beachwood, Ohio 44122
b
ABSTRACT: Fatty alcohols, derived from natural sources, are pathway for the hydrogenation of methyl ester to alcohol, it
commercially produced by hydrogenation of fatty acids or was necessary to develop an experimental reaction method
methyl esters in slurry-phase or fixed-bed reactors. One slurry-
phase hydrogenation of methyl ester process flows methyl es-
ters and powdered copper chromite catalyst into tubular reac-
tors under high hydrogen pressure and elevated temperature. In
the present investigation, slurry-phase hydrogenations of C12
methyl ester were carried out in semi-batch reactions at non-
optimal conditions (i.e., low hydrogen pressure and elevated
temperature). These conditions were used to accentuate the
host of side reactions that occur during the hydrogenation.
that would accentuate the generation of normally undesir-
able side reaction products. The side reaction products may
be obtained under widely varying conditions of hydrogen
pressure and reaction temperature used for the hydrogena-
tion of C12 methyl ester (Table 1). The ability to follow the
wide variety of side reaction products allows an accurate de-
velopment of hydrogenation reaction pathways and, in addi-
tion, provides a method for the more optimal design of cata-
Some 14 side reaction routes are outlined. As an extension of lyst surfaces.
this study, copper chromite catalyst was produced under a num-
ber of varying calcination temperatures. Differences in catalytic
EXPERIMENTAL PROCEDURES
activity and selectivity were determined by closely following
side reaction products. Both activity and selectivity correlate
well with the crystallinity of the copper chromite surface; they
increase with decreasing crystallinity. The ability to follow the
wide variety of side reactions may well provide an additional
tool for the optimized design of hydrogenation catalysts.
JAOCS 74, 333–339 (1997).
A series of slurry-phase, semi-batch hydrogenations of
methyl ester with copper chromite catalyst was carried out
in a fully computer-controlled Autoclave Engineers Micro-
Scale bench-top reactor system (4), referred to as a BTRS.
The experimental system is a small-scale, semi-batch reac-
tor with complete microprocessor control of temperature,
pressure, flow, agitation, and liquid sampling capability for
gas chromatographic analysis. The experimental conditions
were: reactor volume, 100 cc; construction material, 316
KEY WORDS: Copper chromite, fatty alcohol, fatty-fatty ester,
hydrogenation, methyl ester, slurry hydrogenation, wax ester.
Although there are extensive reviews on the hydrogenation
of methyl esters with copper chromite as the catalyst (1,2),
TABLE 1
these studies are limited to discussions of the major reac- Selectivity Differences vs. Reaction Conditions
tion, with little evaluation of side reaction products. In addi-
tion, reaction mechanisms are usually limited to speculation,
due to the lack of substantial side reaction data. It is well
known that the major reaction product in the slurry hydro-
genation of methyl ester (3), under high hydrogen pressure
and elevated temperature, is the fatty alcohol, along with
residual fatty-fatty ester and trace amounts of side reaction
products. Fatty-fatty esters, commonly referred to as wax
esters, are alkyl alkanoates. To describe a detailed reaction
Reaction conditions
H pressure
~2000 psig
280°C
1–6%
~3000 psig
260°C
~2.5%
2
Rx temperature
Catalyst slurry
Reaction products
Alcohol
Fatty-fatty ester
Aldehyde
Acetal
Ethers
Alcohol
Fatty-fatty ester
trace
—
trace
trace
—
Ketone
Fatty acid
Copper soap
Even-chain alkane
Odd-chain alkane
—
trace
—
*
To whom correspondence should be addressed at Procter & Gamble Com-
pany, Sharon Woods Technical Center, 11530 Reed Hartman Highway,
Cincinnati, OH 45241.
Copyright © 1997 by AOCS Press
333
JAOCS, Vol. 74, no. 4 (1997)

334
R.D. RIEKE ET AL.
stainless steel; feed stock, 60 g C12 methyl ester (98%); cop- in Figure 2. Except for the hemiacetal, which is inherently
per chromite catalyst, 1–6 g; hydrogen pressure, 2000 psig; unstable, all reaction products appearing in Figure 2 have
reactor hydrogen sweep, 75 cc/min; reactor temperature, been observed during semi-batch hydrogenations of dodecyl
2
80°C; impeller agitation, ~450 rpm; reaction time, 1–8 h; methyl ester. The ability to measure these products provides
sample frequency, 15 min, 30 min; component identifica- a refined tool for additional characterization of catalyst sur-
tion, gas chromatography (GC)/mass spectroscopy; sample face modification.
analysis, GC. Experimental runs consisted typically of pres-
Catalytic surface properties vs. reaction characteristics.
surizing a methyl ester/catalyst slurry up to 2000 psig with A series of plant-produced copper chromite catalysts were
hydrogen, followed by a continuous flow of hydrogen generated at various calcination temperatures, to produce the
through the reactor at 75 cc/min. The system temperature physical properties shown in Table 2. Powder x-ray diffrac-
was brought to 280°C during a 20-min heat-up cycle. Mi- tion patterns were developed for each of the five samples.
crosamples were taken at 15- or 30-min intervals, beginning These patterns appear in Figure 3. The diffraction patterns
with the initiation of system heat-up. All reaction conditions clearly show an increase in the overall background of the
were recorded by the on-line computer system provided by samples with decreasing calcination temperature, as well as
the BTRS. All reaction products were identified by a broadening of the diffraction peaks. A direct comparison
GC/mass spectroscopy and allowed reaction sample compo- of the samples generated at 399 and 538°C indicates virtu-
sitions to be followed by GC. Analysis of copper chromite ally a complete loss in resolution of certain reflections, espe-
physical properties was provided through Micromeritics cially in the 30 to 40° two-theta region. This is an indication
(
Norcross, GA). X-ray diffraction analysis was carried out of a decrease in surface crystallinity with decreasing calcina-
in-house (Procter & Gamble, Cincinnati, OH), with a Stoe tion temperature. There is also a significant increase in cata-
diffractometer.
lyst surface area with decreasing calcination temperature.
The fact that certain reflections are not observed in the higher
surface area samples may indicate changes in the crystalline
phase of the catalyst as the surface area increases (decreas-
RESULTS AND DISCUSSION
Reaction pathways. A typical dodecyl methyl ester hydro- ing calcination temperature). Such differences in surface
genation reaction profile, consisting of methyl ester deple- morphology suggest significant differences in catalyst activ-
tion, alcohol generation, fatty-fatty ester generation and de- ity and selectivity. Application of the hydrogenation condi-
pletion, and the generation of side reaction products, as listed tions described previously allows the study of such differ-
in Table 1, appears in Figure 1. A more detailed breakdown ences.
of a number of side reactions appears in the section entitled
Using the described variations of copper chromite catalyst,
Catalytic surface properties vs. reaction characteristics. The a series of hydrogenations of dodecyl methyl ester were car-
study of the hydrogenation of dodecyl methyl ester has re- ried out. The first series focused upon the rate of reaction,
vealed a number of interesting side reaction products. The specifically the depletion of dodecyl methyl ester and the gen-
presence of ethers supports the hemiacetal as the transient eration of dodecanol (Figs. 4 and 5). Catalysts calcined at
intermediate of the reaction. Potential reaction routes appear lower temperatures were more active than those calcined at
higher temperatures. The surface area of the catalyst in-
creased with decreasing calcination temperature. Considering
only surface area available for generation of hydrogen atoms,
the results of Figures 4 and 5 are expected. The profile for
generation and depletion of the fatty-fatty ester (dodecyl do-
decanoate) appears in Figure 6. However, the ratios of the do-
decyl methyl ester depletion rates are not equal to the ratios
of surface area. This suggests that different surface morpholo-
gies are present, as previously indicated by the powder x-ray
diffraction patterns of Figure 3.
Alkane generation, dodecane, is demonstrated in Figure 7.
Alkanes are the result of pyrolysis reactions during ester hy-
drogenation, as shown in the potential reaction routes of Fig-
ure 2. The alkene, dodecene, may be generated by thermal de-
composition of dodecyl dodecanoate, and may be followed
by the hydrogenation of the double bond to the alkane, dode-
cane. To a lesser extent, the alcohol product may dehydrate to
form the alkene, which can then be hydrogenated to form the
alkane. Both ether forms, the symmetrical and unsymmetri-
cal, may be hydrogenated, with some difficulty, to form an al-
FIG. 1. Typical fatty methyl ester hydrogenation.
cohol and the alkane.
JAOCS, Vol. 74, no. 4 (1997)

FATTY METHYL ESTER HYDROGENATION: CORRELATIONS
335
FIG. 2. Potential reaction routes in the hydrogenation of alkyl methyl ester.
The odd-chainlength alkane, undecane, appears at low provide a route to odd-chainlength undecane. The fatty-
concentrations during the latter portions of the hydrogena- fatty ester may thermally decompose to form dodecanoic
tion reaction (Fig. 8). Its generation depends upon thermal acid and alkene. Dodecanoic acid may undergo further
decompositions. As shown in Figure 2, there are a number thermal decomposition to form undecane and carbon diox-
of aldehyde (dodecanal) sources, generated from the two ide.
hemiacetal forms. Dodecanal may thermally decompose to
The primary source of the alkane tricosane (C H ) is
23 48
form the alkane, undecane, and carbon monoxide. Thermal through thermal decomposition of dodecyl dodecanoate. The
degradation of dodecyl dodecanoate has been shown to be fatty acid decomposition product, dodecanoic acid, may react
a source of the even-chainlength dodecane. It may also directly with copper chromite catalyst to form the copper
JAOCS, Vol. 74, no. 4 (1997)

336
R.D. RIEKE ET AL.
TABLE 2
CuCr Physical Properties vs. Calcination Temperature
Catalyst
number
Calcination temp.
(°C)
BET area
(m /g)
Pore vol.
(mL/g)
Pore size
(Å)
Catalyst dia.
2
(µ)
1
2
3
4
5
399
427
460
516
538
44.8
35.9
25.8
18.5
13.4
0.31
0.24
0.19
0.16
0.09
276
268
296
344
280
1.66
1.58
1.28
1.23
1.02
FIG. 3. Powder x-ray diffraction patterns of copper chromite catalyst.
soap, cupric dodecanoate. Cupric dodecanoate may thermally dodecyl methyl ether and the symmetrical didodecyl ether.
decompose to form carbon monoxide, cupric oxide, and tri- The generation profile of the didodecyl ether appears in Fig-
cosane (Fig. 9). Additional sources of the fatty acid, other ure 10. As shown in Figure 2, the reaction route of this
than from dodecyl dodecanoate, may be through hydrolysis species is through hemiacetal formation derived from the hy-
of the dodecyl methyl ester feed stock or through oxidation drogenation of dodecyl dodecanoate. The hemiacetal may be
of the aldehyde, dodecanal. However, as the copper chromite hydrogenated to form the symmetrical ether. It is also pro-
catalyst becomes fully reduced during hydrogenation, the posed that the hemiacetal may react with fatty alcohol to
source of water, for hydrolysis, and copper oxide, for oxida- form acetal, which, upon further hydrogenation, may form
tion, becomes minimized.
the symmetrical ether. Although it is difficult, the ether itself
Two forms of ether are generated during the hydrogena- may undergo further hydrogenation to form alcohol and
tion of dodecyl methyl ester. These are the unsymmetrical alkane.
JAOCS, Vol. 74, no. 4 (1997)

FATTY METHYL ESTER HYDROGENATION: CORRELATIONS
337
FIG. 4. Depletion of dodecyl methyl ester vs. catalyst calcination tem-
perature.
FIG. 6. The generation/depletion profiles for dodecyl dodecanoate vs.
catalyst calcination temperature.
FIG. 5. Dodecanol generation vs. catalyst calcination temperature.
FIG. 7. Dodecane generation vs. catalyst calcination temperature.
There is a wide variation in the levels of side reaction
It has been shown that, as the calcination temperature of
products that are created with catalysts generated under dif- copper chromite catalyst is lowered, there is a correspond-
ferent calcination temperatures. Confirmation of the differ- ing decrease in surface crystallinity. The decrease in crys-
ences in selectivity may be firmly established by plotting, for tallinity and the increase in catalyst surface area correspond
example, dodecanol product vs. dodecane. This is shown for well with increased catalytic activity and selectivity. The
hydrogenations with catalysts calcined in the two extreme generation of the primary product, alcohol, is favored along
cases of temperatures studied, 538 and 399°C (Fig. 11). If the with a suppression of all pyrolysis products (i.e., alkanes).
selectivity for both were the same, then the curves would have Over exceedingly long reaction residence times, side reac-
superimposed upon each other. No superimposition confirms tion products may be suppressed when a low-temperature-
the change in selectivity. This large selectivity change occurs calcined copper chromite catalyst is used. Figure 12 is a
for all side reaction products when plotted against the alcohol comparison of dodecanol generation and stability in semi-
product.
batch reactions beyond the point of maximum conversion.
JAOCS, Vol. 74, no. 4 (1997)

338
R.D. RIEKE ET AL.
FIG. 10. Didodecyl ether generation vs. catalyst calcination tempera-
FIG. 8. Undecane generation vs. catalyst calcination temperature.
ture.
FIG. 9. Tricosane generation vs. catalyst calcination temperature.
FIG. 11. Selectivity: dodecane vs. dodecanol at catalyst calcination tem-
peratures of 538 and 399°C.
The high-surface-area, low-calcination-temperature cata-
lyst does not allow the thermal degradation of the dode-
canol product through the alkane side reactions. However, ied further by hydrogenation reaction techniques here de-
significant degradation of dodecanol occurs when using a scribed.
catalyst of low surface area, produced through high calci-
nation temperature. Although not experimentally measured,
the data suggest that the low-surface-area catalyst (high
The authors thank Micheal Mootz, CRD X-Ray Analysis Labora-
calcination temperature) may have relatively large crys-
talline sites that are less active than the multiple, small
crystalline sites of high-surface-area catalyst (low calcina-
tional, Menlo Park, California. In addition, we thank both the Engel-
tion temperature). These differences in surface morpholo-
ACKNOWLEDGMENTS
tory, Miami Valley Laboratory, Procter & Gamble Company,
Cincinnati, Ohio, and Dr. Ronald Swidler (retired), SRI Interna-
hard Corporation and the Procter & Gamble Company for their sup-
gies need further analytical definition and should be stud- port and permission to publish.
JAOCS, Vol. 74, no. 4 (1997)

FATTY METHYL ESTER HYDROGENATION: CORRELATIONS
339
REFERENCES
1
2
3
.
.
.
Muttzal, K.M.K., High-Pressure Hydrogenation of Fatty Acid
Esters to Fatty Alcohols, Waltman, Delft, Netherlands, 1969.
Fatty Alcohols: Raw Materials, Methods, Uses, Henkel KGaA,
Düsseldorf ZR-Fettchemie, 1st edn., 1982.
Peters, R.A., Higher Aliphatic Alcohols, in Kirk-Othmer: Ency-
clopedia of Chemical Technologies, John Wiley & Sons, Vol. 1,
4
th edn., 1991, pp. 865–893.
4
.
Autoclave Engineers, Erie, Pennsylvania, Bulletin 1301, 1991.
FIG. 12. Dodecanol generation/stability vs. catalyst calcination temper-
ature.
[Received June 17, 1996; accepted November 26, 1996]
JAOCS, Vol. 74, no. 4 (1997)