Fatty Methyl Ester Hydrogenation to Fatty Alcohol

Part II: Process Issues

Ross D. Riekea,*, Deepak S. Thakurb, Brian D. Robertsb, and Geoffrey T. Whiteb
a
Procter & Gamble Company, Sharon Woods Technical Center, Cincinnati, Ohio 45241,
and bEngelhard Corporation, Research and Development, Beachwood, Ohio 44122

ABSTRACT: Fatty alcohols are produced by hydrogenating fatty (reduction of the cupric ions) under reaction conditions. The
methyl esters in slurry phase in the presence of copper chromite in situ reduction creates the active sites required for hy-
catalyst at temperatures of 250–300°C and hydrogen pressures of drogenolysis. The fatty alcohol producer therefore requires a
2000–3000 psi. The fatty methyl ester, catalyst, and hydrogen are catalyst with high activity, long life, and good separation
fed to the reactor cocurrently. The product slurry is passed properties. Therefore, research was undertaken to develop a
through gas–liquid separators and then through a continuous fil-
catalyst with higher hydrogenation activity and selectivity.
tration system for removal of the catalyst. A portion of the used
catalyst in crude alcohol is recycled to the hydrogenator. The
Catalyst samples were prepared by using procedures de-
overall efficiency of the process depends upon the intrinsic activ- scribed in the literature (3). CuCr-I was prepared by a simul-
ity, life, and filterability of the catalyst. The fatty alcohol producer taneous precipitation technique, which involves injecting two
therefore requires a catalyst with high activity, long life, and good streams, a metal-bearing (Cu and Cr) solution and ammonia,
separation properties. The main goal of the present laboratory in- into a tank at rates that maintain a constant pH. CuCr-II was
vestigation was to develop a superior copper chromite catalyst prepared by a sequential method in which ammonium chro-
for the slurry-phase process. Two copper chromite catalysts, pre- mate solution was added to copper nitrate solution. Catalytic
pared by different procedures, were tested for methyl ester hy- activity measurements were conducted in a batch autoclave
drogenolysis activity, reusability, and filtration characteristics. as described earlier (4).
The reaction was carried out in a batch autoclave at 280°C and
2000–3000 psi hydrogen pressure. The reaction rates were cal-
culated by assuming a kinetic mechanism that was first-order in RESULTS AND DISCUSSION
methyl ester concentration. The catalyst with the narrower parti-
cle size distribution was 30% more active, filtered faster, and Alcohol generation data as a function of time for the two cat-
maintained activity for several more uses than the catalyst with alysts are presented in Figure 1. This comparison shows that
the broader particle size distribution. X-ray photoelectron spec- the equilibrium conversion values are reached within
troscopy data showed higher surface copper concentrations for
the former catalyst.
JAOCS 74, 341–345 (1997).

KEY WORDS: Copper chromite, fatty alcohol, fatty-fatty ester,

filtration, hydrogenolysis, methyl ester, particle size distribu-
tion, slurry hydrogenation, wax ester.

Commercially, fatty alcohols are produced by methanolysis

of triglyceride or fatty acid feed stock, followed by catalytic
hydrogenolysis of the methyl ester (1). In the slurry process,
the overall alcohol production rate per kilogram of catalyst
depends upon the intrinsic activity, deactivation rate, and fil-
tration characteristics of the catalyst. Conversion levels are
maintained by continuously adding a small amount of fresh
catalyst in the recycle stream, and withdrawing spent catalyst
(2). The fresh catalyst in the oxide form is activated in situ

*To whom correspondence should be addressed at Procter & Gamble Com-

pany, Sharon Woods Technical Center, 11530 Reed Hartman Highway,
Cincinnati, OH 45241. FIG. 1. Dodecanol generation as a function of reaction time.

Copyright © 1997 by AOCS Press 341 JAOCS, Vol. 74, no. 4 (1997)
342 R.D. RIEKE ET AL.

FIG. 2. First-order rate constant plots.

FIG. 4. Dodecane generation as a function of conversion.

FIG. 5. Dodecanal generation as a function of reaction time.

FIG. 3. Dodecyl dodecanoate generation as a function of reaction time.

300–350 min for CuCr-I and 450–500 min for CuCr-II, indi- Dodecane generation, at equal conversion levels, is compared
cating that the former is more active than the latter. The first- for the two catalysts in Figure 4. The alkane is formed by var-
order rate equation plots, shown in Figure 2, clearly show that ious mechanisms discussed earlier (4). This is usually consid-
the hydrogenolysis rate is about 40% faster with CuCr-I than ered as yield loss. The results of Figure 4 show that alkane
with CuCr-II. formation is similar for both catalysts.
Fatty-fatty ester generation (dodecyl dodecanoate) is Dodecyl aldehyde and didodecyl ether are formed in small
demonstrated in Figure 3. Fatty-fatty esters (commonly re- amounts by dehydrogenation and dehydration reactions, re-
ferred to as wax esters or alkyl alkanoates) are formed in situ spectively. In addition, the ether may be formed via an inter-
by transesterification of methyl ester with fatty alcohol or via mediate hemiacetal in the hydrogenation of methyl ester. Re-
aldehyde reaction over copper chromite catalyst. The residual sults shown in Figure 5 indicate that dodecanal formation is
fatty-fatty ester content of the product is lower with CuCr-I independent of the type of catalyst used, while those shown
than with CuCr-II, which is consistent with earlier findings in Figure 6 suggest that ether formation can be minimized by
that the former catalyst is more active than the latter. using CuCr-I catalyst.

JAOCS, Vol. 74, no. 4 (1997)

 FATTY METHYL ESTER HYDROGENATION: PROCESS ISSUES 343

FIG. 6. Didodecyl ether generation as a function of reaction time. FIG. 7. Catalyst aging characteristics.

FIG. 8. Scanning electron microscopic analysis of two catalysts.

JAOCS, Vol. 74, no. 4 (1997)

344 R.D. RIEKE ET AL.

FIG. 9. Particle size distribution of two catalysts.

Catalyst aging data are presented in Figure 7. The catalyst while the CuCr-II catalyst consists of irregular-shaped parti-
samples were evaluated for five uses in a batch reactor at cles. Figure 9 shows that the particle size distribution of
280°C and 3000 psig hydrogen pressure. The results clearly CuCr-I is narrower than that of CuCr-II. The latter has a sig-
demonstrate that CuCr-I retains its activity (as indicated by nificant fraction of particles in the submicron size. The mor-
saponification values around 10) for five uses, while the phology differences and the fraction of small submicron-sized
CuCr-II sample shows high activity (low saponification val- particles lead to differences in filterability of the catalyst, as
ues) for the first three uses and starts deactivating after the shown in Figure 10. The filtration rates with CuCr-I are two
third use, reaching a saponification value of 20 mg KOH/g at times faster than the rates with CuCr-II. This property is ex-
the end of the run. This difference in activity is not surprising tremely useful in a continuous operation and improves over-
if one considers the physical properties of the two catalysts as all efficiency of the process.
shown in Table 1. CuCr-I has higher pore volume and higher
Cu-to-Cr ratio (surface rich in copper). This difference in
Cu/Cr atomic ratio accounts for the difference in hydrogena-
tion rates.
Scanning electron microscopic analysis (Fig. 8) reveals
the morphology differences between the two catalysts under
investigation: CuCr-I has spherical well-defined particles,

TABLE 1
Physicochemical Properties of Copper Chromite Catalysts
Physicochemical property CuCr-I CuCr-
II
Surface area (m2/g) 35 28
Total pore volume (cc/g) 0.19 0.15
Composition (wt%)
Copper 42.5 40.8
Chromium 32.5 31.8
XPS analysis
Cu/Cr atomic ratio
Before reduction 1.2 1.1 FIG. 10. Filtration data for two catalysts.

JAOCS, Vol. 74, no. 4 (1997)

 FATTY METHYL ESTER HYDROGENATION: PROCESS ISSUES 345

ACKNOWLEDGMENTS 3. Thakur, D.S., E. Palka, T.J. Sullivan, E. Nebesh, and B.D.

Roberts, U.S. Patent 5,134,108 (1992).
The authors thank both the Engelhard Corporation and the Procter 4. Rieke, R.D., D.S. Thakur, B.D. Roberts, and G.T. White, Fatty
& Gamble Company for their support and permission to publish. Methyl Ester Hydrogenation to Fatty Alcohol Part I: Correla-
tion Between Catalyst Properties and Activity/Selectivity, J.
Am. Oil Chem. Soc. 74, 333–339 (1997).
REFERENCES
1. Kruetzer, U.R., Manufacture of Fatty Alcohols Based on Nat-
ural Fats and Oils, J. Am. Oil Chem. Soc. 61:343–348 (1984).
2. Peters, R.A., Higher Aliphatic Alcohols, in Kirk-Othmer Ency-
clopedia of Chemical Technologies, John Wiley & Sons, Vol. 1,
4th edn., 1991, pp. 865–893. [Received July 17, 1996; accepted November 26, 1996]

JAOCS, Vol. 74, no. 4 (1997)