Mechanism of the isomerization of 1-alkene during iron-catalyzed Fischer-Tropsch synthesis

Article abstract of DOI:10.1006/jcat.2001.3175

The deuterium/hydrogen exchange reaction was conducted under iron-catalyzed Fischer-Tropsch reaction conditions using a mixture of deuterated octane, nonane, decane, tridecane, and pentadecane as the probes. The deuterium/hydrogen exchange did not occur in alkanes under these conditions. Under the same reaction conditions, 1-octene-d₁₆ was used as the probe to study the isomerization of 1-alkene. The 1-octene-d₁₆ was reduced to deuterated octane, and isomerized to deuterated trans-2-octene and cis-2-octene with nearly equal amounts of the two isomers. Products from cracking and isomerization to internal octenes other than the 2-octene isomers did not occur to a measurable extent. When 1-octene-d₁₆ was used as the probe, there was H/D exchange in octenes and octane; the lowest deuterium-containing isomer products were the d₁₁ isotopomers, and there were no d₁₆ isomers in the 2-octenes.

Full text of DOI:10.1006/jcat.2001.3175

Journal of Catalysis 199, 202–208 (2001)
doi:10.1006/jcat.2001.3175, available online at http://www.idealibrary.com on
Mechanism of the Isomerization of 1-Alkene during Iron-Catalyzed
Fischer–Tropsch Synthesis
Buchang Shi, Robert J. O’Brien, Shiqi Bao, and Burtron H. Davis1
Center for Applied Energy Research, University of Kentucky, 2540 Research Park Drive, Lexington, Kentucky 40511
Received July 24, 2000; revised January 11, 2001; accepted January 11, 2001; published online March 27, 2001
conditions (11, 12). They found that there was a H/D ex-
change between the aldehyde and ethanol, but they did not
use those probes to further study the details of the mech-
anism of FT reactions. It has been reported (13–17) that
when H2 was replaced by D2 to hydrogenate CO, an inverse
kinetic isotope effect was obtained. Most of the deuterium
tracer experiments conducted so far have not addressed the
key mechanistic problems of the FT reactions, such as the
readsorption of 1-alkene. Miller and Moskovits (18) used
ethylene to study the mechanism of the FT reaction and
found that the deuterium from the deuterated ethylene was
incorporated into the alkanes and alkenes almost entirely
as CD2 units. The authors published their results in a note
in 1989; however, further details of this study have not been
published since that time.
Recently, we have reported that product hold-up has a
very important impact on the data interpretation in isotopic
tracer studies of FT reactions (19, 20). We reported that due
to the nature of isotope tracer experiments in FT synthesis,
this product accumulation factor could be minimized, but
could not be removed completely in 14C tracer studies. Be-
cause of this, the deuterium tracer technique becomes an
attractive alternative.
The belief that H/D exchange might occur under FT re-
action conditions (8) is the main reason why just a few
deuterium tracer experiments have been conducted even
though the deuterium tracer technique is one of the most
frequently used methods in homogeneous organic reac-
tions. It issurprisingthat no experiment hasbeen performed
to show whether there is a H/D exchange reaction among
the major products under FT reaction conditions. If there
is no H/D exchange in alkanes, it will be possible to study
the chemical behavior of 1-alkene under FT reaction con-
ditions using the deuterated 1-alkene as the probe.
There are many ways for double bond isomerization to
take place including in photochemical reactions and on
treatment with metal ions (21). In the presence of metal
catalysts, the hydrogenation of 1-alkene is often accom-
panied by double bond migration and cis/trans isomeriza-
tion. For the homogeneous catalyzed reaction, at least two
mechanisms have been recognized for the isomerization of
The deuterium/hydrogen exchange reaction was performed un-
der iron-catalyzed Fischer–Tropsch reaction conditions using a
mixture of deuterated octane, nonane, decane, tridecane, and
pentadecane as the probes. The results indicated that deu-
terium/hydrogen exchange does not occur in alkanes under these
conditions. Under the same reaction conditions, 1-octene-d16 was
used as the probe to study the isomerization of 1-alkene. The
1-octene-d16 was reduced to deuterated octane, and isomerized
to deuterated trans-2-octene and cis-2-octene with approximately
equal amounts of the two isomers. No other internal deuterated
octene and C7 products were obtained. The deuterium/hydrogen
exchange reaction occurred in deuterated 1-octene as well as in the
productsformed from it. Themaximum number ofdeuterium atoms
that can be replaced by hydrogen is 5. There is no C8D16 isotopomer
of 2-octenes. To account for these experimental facts, a mechanism
was proposed for the isomerization of 1-alkene under iron-catalyzed
c
Fischer–Tropsch reaction conditions.
2001 Academic Press
Key Words: deuterium; Fischer–Tropsch synthesis; iron catalyst;
isotopic tracer; isomerization, alkene.
INTRODUCTION
In studying the mechanism of Fischer–Tropsch (FT) syn-
thesis, isotopic tracer techniques have been widely used
and, based on the results, several very important conclu-
sions have been reported (1–9). The most frequently used
isotopes in FT reactions are carbon-14- and carbon-13-
labeled compounds. Almost all of the possible carbon-14-
and carbon-13-labeled compounds suitable for studying the
mechanism of FT reactions, such as alcohols, alkenes, CO,
CO2, aldehyde, and many others, have been used as probes
(10). Compared to the number of carbon-14 and carbon-
13 tracer experiments, very few deuterium tracer experi-
ments have been conducted to study the mechanism of FT
reactions.
Kruykov and co-workers used 14C- and deuterium-
labeled aldehyde and ethanol to study the H/D ex-
change between these two compounds under FT reaction
1 To whom correspondence should be addressed. Fax: (859) 257-0302.
E-mail: davis@caer.uky.edu.
202
0021-9517/01 $35.00
c
Copyright
2001 by Academic Press
All rights of reproduction in any form reserved.

AN IRON-CATALYZED FISCHER–TROPSCH SYNTHESIS
203
the alkene in the presence of a metal complex: the metal ion medium-pressure FT synthesis conditions. The results nor-
hydride addition–elimination mechanism and the -allyl mally associated with metals may not apply here because
complexmechanism (21, 22). For metalliccatalyststhe bond (1) we are dealing with an iron carbide catalyst rather than
breaking is usually considered to be homolytic so that the a metal catalyst and (2) CO is present at reasonably high
breaking of the hydrogen bond forms two hydrogen atoms concentrations.
with an electron that is utilized in bonding to the metal
In the iron-catalyzed FT reactions, a mixture of alkane,
surface. Thus, metal-catalyzed hydrogenation and/or iso- 1-alkene, cis- and trans-2-alkenes, other internal alkenes,
merization reactions are usually considered to be, or to isoalkanes, and oxygenates of each carbon number may
resemble, a mechanism that involves a radical intermedi- be formed. Alkenes can be reabsorbed to produce higher
ate (23).
carbon number hydrocarbons (1–4, 6, 7, 10, 34–48). The
Since the pioneering work of the Farkas brothers, the fraction of the internal alkenes that are primary products
addition and exchange of H/D has been widely utilized to remains an unanswered question. Many investigators con-
study the mechanism of metal-catalyzed conversions of hy- sider 1-alkene to be the primary product and that internal
drocarbons (24). A vast literature has developed that de- alkenes arise from isomerization products.
scribes conversions with most metal catalysts. However,
In this study, we want to address the questions of whether
Maitlis and co-workers (8) indicate that “deuterium is there is H/D exchange in alkanes under the iron-catalyzed
not suitable as a label since scrambling occurs on heat- FT reaction conditions and the extent of isomerization of
ing deuterio-organics, especially over metal under hydro- 1-alkens product.
gen.” It is true that in most cases the H/D exchange re-
action occurs over metals, but there are exceptions. Our
work using platinum catalysts shows, however, that H/D
exchange is not a simple process. Thus, at 482 C, we ob-
serve that an alumina support, either acidic or nonacidic,
catalyzes H/D exchange in n-octane as rapidly as, or more
than Pt at levels normally found in naphtha reforming cata-
lysts. With Pt supported on silica, the exchange of D with
n-octane was rapid at lower temperatures but exchange was
much less at 482 C. There is little H/D exchange between
octane-d18 and methylcyclohexane or between octane-d0
and methylcyclohexane-d14 with Pt/SiO2 at a temperature
as high as 482 C (25). There is also little H/D exchange
in octane and in methylcyclohexane over Pt/SiO2 at 482 C
when D2 gas was passing through the reaction zone, even
though all of the products, such as ethylbenzene, xylenes,
and toluene, contain deuterium in a statistic distribution
(25). The lack of H/D exchange in the reagents defines the
mechanism for dehydrocyclization (25–30). Richard and
co-workers (31) reported that over a supported ruthenium
catalyst at 150 C and 0.45 MPa (44.23 atm) total pressure
(H2), less than 5% of the deuterium in benzene-d6 was ex-
changed after 75 min (88% conversion to products). These
results clearly show that the H/D exchange reaction does
not always occur under catalytic conditions at high temper-
ature, at least not so rapidly that it masks the chemistry
defined by kinetic-controlled reactions. Metal catalysts are
noted for their ability to effect the hydrogenolysis of alka-
nes as well as to catalyze H/D exchange. Thus, Monnier
et al. (32) found that, while an iron Fischer–Tropsch cata-
lyst would catalyze the hydrogenolysis of n-hexadecane,
the presence of a low partial pressure of CO would inhibit
this reaction so that it was not detectable. As another ex-
ample, Huang et al. (33) found that n-octacosane did not
undergo measurable hydrogenolysis following more than
150 h of exposure to a doubly promoted iron catalyst under
EXPERIMENTAL
Deuterated octane-d18, nonane-d20, decane-d22, tridec-
ane-d28, and pentadecane-d32 were purchased from Cam-
bridge Isotope Laboratories, Inc. The deuterium contents
(according to the manufacturer) are 99% , 98% , 99% , 98% ,
and 98% , respectively. Other chemicals were obtained from
Aldrich and used without further purification.
The synthesis of 1-octene-d16 was conducted by a method
modified from a previous report (49). To a mixture of 5 g
of octanol-d17 and 3 g of pyridine in 150 ml of diethyl ether,
3 g of acetyl chloride in 50 ml of diethyl ether was added
dropwise with excellent mechanical stirring. After the ad-
dition was completed, the reaction mixture was warmed to
boiling and kept for an additional 30 min. When the reac-
tion mixture was cooled, 100 ml of 10% HCl was added.
The organic layer was separated, and the water layer was
extracted with ether (50 ml
were combined and washed with 10% HCl (10 ml
3). The organic fractions
2)
and then with water (10 ml), and then dried over anhydrous
MgSO4. After evaporation of the solvent, 7.9 g of octyl-d17
acetate was obtained without further purification. The py-
rolysis of octyl-d17 acetate was conducted at 480 C at a flow
rate of 0.54 ml/h in a glass reactor. To the mixture of the
products collected was added 2 ml of H2O, and the organic
fraction was separated. After distillation, 2.5 g of octene-d16
was obtained. The deuterium content of this compound was
determined by GC/MS to be 98.5% .
The FT reaction was carried out in a 1-L continuously
stirred autoclave reactor (CSTR) as described previously
(37–46). The iron catalyst (100 Fe : 4.4 Si : 1.4 K atomic ra-
tio) was activated using CO at 270 C and the synthesis was
conducted at 0.12 MPa (11.9 atm) and 270 C using a 10 wt%
catalyst slurry (50). Three traps follow the reactor and are
held at 160 C, 60 C, and 0 C. For the run that included the

204
SHI ET AL.
addition of deuterated alkanes (referred to as the alkane
addition run), a mixture of octane-d18, nonane-d20, decane-
d22, tridecane-d28, pentadecane-d32, and o-xylene (total
10 ml) was added to the H2/CO feed to the 1-L autoclave
during a 5-h period. Immediately prior to addition of the
tracers, the three products traps were emptied. At the end
of the addition of the deuterated alkanes, the product traps
were again drained and the contents were analyzed us-
ing our normal procedure as well as the GC/MS analysis.
In the run with 1-octene-d16 a similar procedure was fol-
lowed. 1-Octene-d16 (2.5 g of 1-octene-d16 diluted with 5 g of
o-xylene) was added to the reactor over 5 h. At the end of
the addition of 1-octene-d16, the product traps were drained
and drained again after another 19 h. The carbon added in
1-octene is 10.3% of the total hydrocarbon products (FT
hydrocarbons plus added 1-octene).
The deuterium content of the tracer compounds and its
products was determined using GC/MS. The data were cor-
rected for the 13C content of the products. In the cases when
quantitative analysis of the relative amount of isotopomer
of a compound was not possible by GC, the GC/MS method
was used. Because of the inverse isotope effect of the
deuterated compound on gas chromatography (51–57), the
relative amount of the total area of the molecular ion of
each isotopomer was used to calculate the molar ratio (58).
To compare the relative amount of different compounds us-
ing the total area of the molecular ion of each compound,
the corrections were made using a known mixture as the
standard.
FIG. 1. Deuterium distribution in octane before and after the alkane
addition run.
probes after reaction are the same as those before reac-
tion. These results clearly show that there is no measurable
H/D exchange in alkanes under iron-catalyzed FT reaction
conditions where CO conversion is not complete.
As can be seen from the above results that there is no
measurable H/D exchange in the deuterated alkanes. The
deuterated alkane did not exchange its deuterium with its
ordinary counterpart; it did not exchange its deuterium
with alkene, which is one of the major products of the
FT reaction; and it did not exchange its deuterium with
H2, which is a reactant for the FT reaction. The deuterium
content of recovered tracer compound would decrease if
exchange occurred, and it did not. Since there is no H/D
exchange in the alkane, the opportunity to use deuterium
tracer techniques to study the mechanism for FT reaction is
appropriate.
It is reasonable to assume, as shown in the following,
that for a given deuterated 1-alkene, the deuterium on car-
bon C4 and higher carbon positions will not be subject to
H/D exchange and its behavior will be similar to that of
the paraffin under iron-catalyzed FT reaction conditions.
RESULTS AND DISCUSSIONS
No H/D Exchange in Alkanes
Deuterated octane, nonane, decane, tridecane, and pen-
tadecane were selected as the probes in studying the H/D
exchange reaction under Fe-catalyzed FT reaction condi-
tions. The deuterium content of each starting compound as
well as the five deuterated probes recovered after 5 h run
under these conditions was determined using GC/MS. As
can be seen from the data in Table 1 and the representa-
tive data shown in Fig. 1, the D% and D/mol of the alkane
TABLE 1
Deuterium Contents in Deuterated Alkanes
After reaction
Before reaction
D% D/Mol
(5 h on stream)
Deuterated
alkane
On the other hand, whether the deuteriums on carbon C1
to C3 positions will undergo H/D exchange if isomerization
occurs depends upon the mechanism that applies. If the iso-
merization occurs through a -allyl complex mechanism,
there should be little or no H/D exchange; if it is through
a metal ion hydride addition–elimination mechanism, H/D
exchange should occur.
D%
D/Mol
Octane
Nonane
Decane
Tridecane
Pentadecane
99.6
97.7
98.8
97.9
96.7
18.0
19.5
21.7
27.4
30.8
99.6
97.8
98.9
97.7
96.3
18.0
19.6
21.8
27.4
30.8

AN IRON-CATALYZED FISCHER–TROPSCH SYNTHESIS
205
TABLE 2
is no measurable amount of other internal octenes. This
confirmed the assumption that only the deuterium on C1
to C3 carbons in 1-octene-d16 will be subjected to H/D ex-
change. If 3-octenes and 4-octenes were also produced, the
deuterium on C4 and C5 carbons would also be subjected
to H/D exchange and the H incorporation would exceed
5. The lack of 3-octene is believed to be the result of low
adsorption rates of 2-octene as compared to 1-octane (65)
and is in agreement with exchange up to only 5 H.
In the double bond migration of 1-alkene under acidic or
basic conditions, Zaitsev’s rule applies and the double bond
is located between the carbons with the fewest hydrogens.
For example, in the case of base-promoted isomerization
of 1-decene, the mixture of products contains not only 1-
decene and cis- and trans-2-decene but also the cis and trans
isomers of 3-, 4-, and 5-decene as well as branched alkene as
a result of the rearrangement of carbocation (21). Since 3-
or 4- or branched deuterated octenes were not obtained, the
isomerization of 1-octene under iron-catalyzed FT reaction
conditions through carbonium or carbocation intermediate
is ruled out.
Mole Percentage of Hydrogenation and Isomerization of Added
1-Octene-d16 under Iron-Catalyzed Hydrocarbon Synthesis Condi-
tions
Mol% (5 h on stream)
Mol% (24 h on stream)
1-Octene
Octane
trans-2-Octene
cis-2-Octene
85.9
8.0
3.0
66.5
20.8
7.1
3.1
5.6
The Mechanism of Isomerization of 1-Octene-d16 under
Iron-Catalyzed FT Reaction Conditions.
a. The product distribution. The data in Table 2 pro-
vide the composition of the deuterium-containing C8 prod-
ucts collected at the end of the 5-h addition of 1-octene-d16
and the sample collected 19 h later. 1-Octene-d16 was iso-
merized to deuterium-containing cis-plus trans-2-octenes
and hydrogenated to deuterium-containing n-octane. The
extent of conversion, as expected, was greater in the sample
collected after 19 h than in the sample collected immedi-
ately after termination of 1-octene-d16 addition. The ratio
of cis- to trans-2-octene is essentially 1, indicating no steric
preference for the abstraction and addition of the H (D)
that leads to the 2-octene products.
The cis : trans ratio of the isomerization product may de-
pend upon the catalyst (59–61). The results and interpreta-
tions of the relative ratio of trans-/cis-2-isomers are often
contradictory(22). The high ratio oftrans-/cis-2-butene (62)
and the high ratio ofcis-/trans-2-isomers(63, 64) mayresults
from a similar mechanism. Since the experimental ratio of
cis-/trans-2-octene formed from the added 1-octene-d16 is
close to 1, the trans : cis ratio of the 2-alkene product alone
cannot define which mechanism is applicable.
In the C7 hydrocarbon region in gas chromatography,
there is no indication of any deuterated compound. Thus,
hydrogenolysis to produce C7 isomers does not occur to a
measurable extent. This result is in agreement with earlier
studies (33, 66).
The deuterated compounds with molecular ion (M+) 139
to 141 were found in the C9 hydrocarbon region of the gas
chromatograph. These compounds could be either C9 alka-
nes or C9 alkenes with up to 15 deuterium atoms. This result
indicates that the added 1-octene can initiate chain growth
under iron-catalyzed FT reaction conditions, in agreement
with the results of carbon-14-labeled decene tracer stud-
ies under similar conditions (40, 42, 67). However, under
these experimental conditions, the C9 products containing
deuterium were too low for the relative rates of hydrogena-
tion to chain growth to be easily determined. The data show
that the hydrogenation rate was greater than chain initia-
tion rate.
Figure 2 shows the chromatogram of the C8 products dur-
ing the isomerization of 1-octene-d16 under iron-catalyzed
FT reaction conditions. Peaks for only four significant com-
pounds were obtained. Compared to these four peaks, there
b. Mechanism for deuterium distribution in octenes
and octanes. For the isomerization of 1-alkene through
a
-allyl complex, there should be several important
characteristics: more trans-2-alkene than cis-2-alkne will
be produced, the reaction will involve a 1,3-hydrogen shift,
and no external hydrogen is required. A typical mechanism
was demonstrated with Fe3(CO)12, which shows that dou-
ble bond migration does not need external hydrogen and
that the reaction takes place through a 1,3-hydrogen shift
(68). For a mixture in which 1-octene-d16 and 1-octene-d0
is present, as would be the case when we add 1-octene-d16
to the synthesis gas which produces 1-octene-d0, the
product exiting the reactor would contain 1-octene-d16 and
2-octene-d16 combined with 1-octene-d0 and 2-octene-d0 if
the isomerization occurred only by internal D (or H) shift.
FIG. 2. Total ion chromatograph of deuterium-containing C8 prod-
ucts that are in the reactor effluent product at the end of the addition of
1-C8D16
.

206
SHI ET AL.
TABLE 3
In heterogeneous catalysis, hydrogen is almost always
considered to be adsorbed by homolytic bond cleavage; that
is, hydrogen dissociates to form two hydrogen atoms that
bond with the surface utilizing an electron associated with
the metal (69). It is the exception when hydrogen is ad-
sorbed by heterolytic bond cleavage; one rare example of
this is adsorption on ZnO where H + is formed and adsorbs
on a surface oxygen ion and the H that forms is bonded
to the surface zinc ion (70). Likewise, the involvement of
molecular hydrogen in hydrogenation is rare (71).
Deuterium Distribution in Octenes
1-Octene
trans-2-Octene
cis-2-Octene
Starting 5 h on 24 h on 5 h on 24 h on 5 h on 24 h on
reagent stream stream stream stream stream stream
D%
D/Mol
C8D16
98.5
15.8
77.6
20.8
1.6
0
92.7
14.8
31.5
33.5
23.2
10.7
1.0
87.1
13.9
11.5
21.8
28.3
28.9
7.2
88.6
14.2
0
41.1
36.4
20.8
1.7
83.5
13.4
0
17.4
28.4
32.4
16.6
5.2
88.5
14.2
0
85.1
13.6
0
C8D15
H
39.6
39.3
18.0
3.1
24.5
29.7
28.6
17.2
0
C8D14H2
C8D13H3
C8D12H4
C8D11H5
The exchange of H and D in 1-octene could also be ex-
plained by a mechanism that involves homolytic adsorption
of H2 with the formation of M–H, where M is an active site
in iron carbide, and then follows a reaction pathway that
is the same as that shown in Eq. [1]. The only difference
between this reaction pathway and the hydride ion path-
way is the step leading to the formation of M–H: in the
hydride ion mechanism the vinyl D is abstracted by M as a
D whereasin the homolyticbond-breakingmechanism the
vinyl D is abstracted by M as D . The homolytic bond cleav-
age mechanism is considered to apply to the H/D exchange
of n-octane with a Pt-silica catalyst (25–30). However, the
exchange leads to a distribution of isotopomers that is
statistically based on the H/D ratio in the feed so that, for
the conversion of an equimolar mixture of n-octane-d18 and
n-octane-d0, the isotopomer present in the largest amount
will be n-octane-d9. For the homolytic bond cleavage to ap-
ply in the present case, there must be some reason for the
exchange to be limited to 5 D and, considering the experi-
mental error of the GC/MS analysis and the small amount
of the H5 isotopomer, the limit may be 4 D. A reasonable
explanation is the relative adsorption of 1- and 2-octene. In
competition with CO, 1-octene is expected to be preferen-
tially adsorbed compared to 2-octene. Thus, it is reasonable
to expect 1-octene-d16 to be adsorbed through reaction with
M–H that is formed by homolytic bond cleavage of hydro-
gen to then undergo the reactions shown in Eq. [1]. If this is
the case, desorption of 1- and 2-octene must compete favor-
ably with the rate of hydrogenation and with the isomeriza-
tion of 2-octene to 3-octene. The homolytic bond cleavage
mechanism agrees with the data generated in this study and
appears to be the most reasonable reaction mechanism.
The deuterium distribution in octane (Table 4) is similar
to that of 1-octene and higher than that of the 2-octenes.
This result indicates that octane, once formed, does not
continue to undergo H/D exchange. Since no d16 isomer
of 2-octene was found, 44% of the C8D16H2 isomer of n-
octane was produced exclusively from the hydrogenation
of 1-octene; the remaining isotopomers can be generated
from both 1-octene and 2-octene. Similar to the octenes, d11
is the lowest deuterium-containing isotopomer of n-octane
(C8D11H7); i.e., the exchange of up to 5 D with H to produce
octene isotopomers followed by the addition of 2 H to form
n-octane.
0
0
0
2.4
0
0
The data in Table 3 indicate that there is H/D exchange
in the 1-octene recovered from the reaction mixture. The
feed 1-octene-d16 contains 98.5% D (15.8 D/molecule)
whereas the sample collected at the end of the 5-h period of
addition contains 92.7% D (14.8D/molecule) and therefore
shows an exchange of about 1 D/molecule. The sample
collected 19 h after terminating the addition of 1-octene-d16
contained 87.1% D (exchange of 2 D/molecule). The
highest hydrogen-containing isotopomer is 1-octene-d11h5.
With respect to the mechanisms for the isomerization
reaction, the most important finding in this study is that
there isno detectable d16 isotopomer in 2-octene. The added
1-octene contained 78% of the d16 isotopomer and the
1-octene in the sample recovered after the 5-h period con-
tained 32% of the d16 isotopomer. This makes the lack of
any d16 isotopomer in the recovered 2-octenes significant
since this requires that, for the formation of cis- and trans-2-
octene, at least one D be replaced by H. If the isomerization
was exclusively by the -allyl complex, the amount of the
2-octene-d16 isotopomer must be between 32 and 78% . If
even 1.5% of the d16 isotopomer was formed, it would have
been observed in the GC/MS analysis that was utilized. This
requires that the -allyl complex pathway, if it contributes
at all, must be less than about the 2% limitation.
The results are consistent with a mechanism that in-
volves the metal ion hydride addition–elimination mech-
anism (Eq. [1]), in which the external hydrogen leads to
H/D exchange in the 1-alkene through Eq. [1].
M
H
*
=
RCD2CD CD2 )
RCD2C(M)D CD2H
M
D
*
=
RCD CDCD2H
[1]
)
If the process shown in Eq. [1] reverses several times, the
maximum number of deuteriums in the 1-alkene exiting
the reactor that can be replaced by hydrogen is five, since
only the deuteriums in the C1, C2, and C3 positions can be
exchanged. Thismechanism would require heterolyticbond
breaking during the chemisorption of hydrogen.

AN IRON-CATALYZED FISCHER–TROPSCH SYNTHESIS
207
TABLE 4
the 2-octene isomers did not occur to a measurable ex-
tent. When 1-octene-d16 was used as the probe, there is
H/D exchange in octenes and octane; the lowest deuterium-
containing isomer products are the d11 isotopomers, and
there are no d16 isomers in the 2-octenes. To account for
all of these experimental facts, the mechanism for isomer-
ization of 1-alkene in an Fe-catalyzed FT reaction is pro-
posed to be the metal ion hydride or metal–hydrogen atom
addition–elimination mechanism, with the latter consid-
ered to be most likely.
For the metal-catalyzed FT synthesis, several mecha-
nisms have been proposed. These mechanisms differ in the
details of initiation, propagation, and termination steps.
However, despite these differences, most of these mech-
anisms agree on the starting hydrogen-containing species
as M–H or Hads. It is reasonable to assume that the metal-
adsorbed hydrogen species that plays a very important role
in FT reactions is also the one that promotes the isomeriza-
tion of 1-alkene.
Deuterium Distribution in Octane
5 h on stream
24 h on stream
D%
D/Mol
94.7
15.2
44.0
37.2
9.1
9.7
0
87.1
13.9
12.4
26.9
22.8
24.3
7.8
C8D16H2
C8D15H3
C8D14H4
C8D13H5
C8D12H6
C8D11H7
0
5.8
As indicated in Eq. [1], in the formation of 2-alkene from
1-alkene, the half-hydrogenated intermediate was formed.
This half-hydrogenated intermediate, which is the common
intermediate in both isomerization and hydrogen addition,
may revert to the starting 1-alkene, yield 2-alkene, or be hy-
drogenated to alkane. In the case of 1-octene-d16, whether
the half-hydrogenated intermediate is structure A or a
mixture of structure A and structure B depends upon the
relative ratio of the surface pool of H/D.
ACKNOWLEDGMENTS
This work was supported by U.S. DOE Contract No. DE-FC26-
98FT40308 and the Commonwealth of Kentucky.
REFERENCES
1. Hall, W. K., Kokes, R. J., and Emmett, P. H., J. Am. Chem. Soc. 82,
1027 (1960).
2. Hall, W. K., Kokes, R. J., and Emmett, P. H., J. Am. Chem. Soc. 79,
2983 (1957).
3. Kummer, J. T., Dewitt, T. W., and Emmett, P. H., J. Am. Chem. Soc.
70, 3632 (1948).
4. Kummer, J. T., and Emmett, P. H., J. Am. Chem. Soc. 75, 5177 (1953).
5. Kummer, J. T., Podgurski, H. H., Spencer, W. B., and Emmett, P. H.,
J. Am. Chem. Soc. 73, 564 (1951).
6. Blyholder, G., and Emmett, P. H., J. Phys. Chem. 63, 962 (1959).
7. Blyholder, G., and Emmett, P. H., J. Phys. Chem. 64, 470 (1960).
8. Maitlis, P. M., Long, H. C., Quyoum, R., Turner, M. L., and Wang,
Z.-Q., Chem. Commun. 1 (1996).
9. Turner, M. L., Long, H. C., Shenton, A., Byers, P. K., and Maitlis,
P. M., Chem. Eur. J. 1, 549 (1995).
10. Raje, A., and Davis, B. H., in “Fischer–Tropsch Synthesis. Mechanism
Studies Using Isotope, Catalysis” (J. J. Spring, Ed.), Vol. 12, p. 52. The
Royal Chem. Soc., Cambridge, 1996.
11. Kryukov, Y. B., Bashkirov, A. N., Fridman, R. A., Liberov, L. G.,
Smimova, R. M., and Pegov, A. A., Neftekhimiya 5, 62 (1965).
12. Kryukov, Y. B., Bashkirov, A. N., Fridman, R. A., Liberov, L. G.,
Smimova, R. M., Lysenka, G. T., and Pegov, A. A., Neftekhimiya 4,
869 (1964).
13. Sakharoff, M. M., and Dokukina, E. S., Kinet. Katal. 2, 710 (1961).
14. Mckee, D. M., J. Catal. 8, 240 (1967).
15. Luytene, L., and Jungers, J. C., Bull. Soc. Chim. Belg. 54, 303 (1945).
16. Dalla Betta, R. A., and Shelef, M., J. Catal. 49, 383 (1977).
If the amount of H is much greater than that of D, the inter-
mediate will be structure A only. The further hydrogenation
of A will give C8D16H2 while -elimination will produce ei-
ther 1-octene with a 33% possibility for the d16 isotopomer;
or 2-octene with the possibility of a d16 isotopomer being
zero. In this tracer experiment, the ratio of H atoms (from
H2) to D atoms (from 1-octene-d16) is more than 80 : 1, even
assuming all D can equilibrate during the 5-h addition pe-
riod. For the 24-h sample, the ratio is close to 900 to 1. This
ensures that M–H will dominate in the isomerization and
hydrogenation reactions. If the ratio of the surface pool of
H/D was small, a significant amount of structure B would
be produced, which could generate the d16 isotopomer of
2-octene.
CONCLUSIONS
Under iron-catalyzed FT synthesis conditions, the 1- 17. Wilson, T. P., J. Catal. 60, 167 (1979).
18. Miller, D., and Moskovits M., J. Am. Chem. Soc. 111, 9250 (1989).
octene was hydrogenated to octane, isomerized to trans-
and cis-2-octenes with the formation of approximately
equal amounts of these isomers, and was also incorporated
into higher carbon number hydrocarbons. Products from
cracking and isomerization to internal octenes other than
19. Shi, B., and Davis, B. H., “Abstracts of papers, 217th ACS Meeting,
Anaheim, CA, March, 1999.” Am. Chem. Soc., Washington, DC, 1999.
20. Shi, B., and Davis, B. H., Catal. Today 58, 255 (2000).
21. March, J., “Advances in Organic Chemistry,” pp. 581–585. Wiley-
Interscience, New York, 1992.

208
SHI ET AL.
22. Olah, G. A., and Molnar, A., “Hydrocarbon Chemistry,” pp. 118–128.
Wiley, New York, 1995.
44. Tau, L-M., Dabbagh, H. A., and Davis, B. H., Energy Fuels 5, 174
(1991).
23. Bond, G. C., “Catalysis by Metals,” Academic Press, New York, 1962.
24. Farkas, A., “Orthohydrogen, Parahydrogen and Heavy Hydrogen,”
Cambridge Univ. Press, Cambridge, UK, 1925.
45. Tau, L-M., Dabbagh, H. A., Halasz, J., and Davis, B. H., J. Mol. Catal.
71, 37 (1992).
46. Xu, L., Bao, S., Houpt, D. J., Lambert, S. J., and Davis, B. H., Catal.
Today 36, 347 (1997).
25. Shi, B., and Davis, B. H., J. Catal. 147, 38 (1994).
26. Shi, B., and Davis, B. H., J. Catal. 157, 626 (1995).
27. Shi, B., and Davis, B. H., J. Catal. 162, 134 (1996).
28. Shi, B., and Davis, B. H., in “Proceedings, 11th International Congress
on Catalysis, Baltimore, 1996” (J. W. Hightower, W. N. Delgass,
E. Iqlesia, and A. T. Bell, Eds.), p. 1145. Elsevier, Amsterdam, 1996.
29. Shi, B., and Davis, B. H., J. Catal. 170, 89 (1997).
47. Herrington, E. F., Chem. Ind. 347 (1946).
48. Iglesia, E., Reyes, S. C., and Madon, R. J., J. Catal. 129, 238 (1991).
49. Shi, B., Ji, Y., Dabbagh, H. A., and Davis, B. H., J. Org. Chem. 59, 845
(1994).
50. O’Brien, R. J., Xu, L., Spicer, R., and Davis, B. H., Energy Fuels 10,
921 (1996).
30. Shi, B., and Davis, B. H., J. Catal. 168, 129 (1997).
31. Richard, M. A., Taube, D. J., Yee, D. K., Durleux, D. O., De Deken,
51. For a review and the references cited, see: van Hook, W. A., ACS Adv.
Chem. Ser. 89, 99 (1969).
J. C., Alcala, P., and Moreno, H., Prepr. Div. Fuel Chem. Am. Chem. 52. Bruner, F., Cartoni, G. P., and Liberti, A., Anal. Chem. 38, 298 (1966).
Soc. 39, 1833 (1991).
32. Monnier, J., Quilliam, M., and Anderson, R. B., Can. J. Chem. Eng.
64, 469 (l986).
33. Huang, C. S., Dabbagh, H., and Davis, B. H., Appl. Catal. 73, 237 (1991)
34. Podgurski, H. H., Kimmer, J., DeWitt, T., and Emmett, P. H., J. Am.
Chem. Soc. 72, 5382 (1950).
53. van Hook, W. A., J. Phys. Chem. 71, 3270 (1967).
54. Mohnke, M., and Heybey, J., J. Chromatogr. 417, 37 (1989).
55. Shi, B., and Davis, B. H., J. Chromatogr. A 654, 319 (1993).
56. Shi, B., Keogh, R. A., and Davis, B. H., J. Chromatogr. A 678, 97 (1994).
57. Shi, B., and Davis, B. H., J. Chromatogr. A 731, 351 (1996).
58. Shi, B., and Davis, B. H., submitted.
35. Kokes, R. J., Hall, W. K., and Emmett, P. H., J. Am. Chem. Soc. 79,
2089 (1957).
59. Bond, G. C., Phillipson, J. J., Wells, P. B., and Winterbottom, J. M.,
Trans. Faraday Soc. 60, 1847 (1964).
36. Pichler, H., Schulz, H., and Elstnor, M., Brennst. Chem. 48, 78 (1967).
37. Ross, R., Robinson, R., Tau, L.-M., and Davis, B. H., J. Catal. 105, 335
(1987).
38. Dabbagh, H., Tau, L.-M., Bao, S., Halasz, J., and Davis, B. H., in “Cata-
lysis 1987” (J. Ward, Ed.), pp. 61–72. Elsevier, Amsterdam, 1988.
39. Tau, L.-M., Dabbagh, H., Bao, S., Halasz, J., Chawla, B., and Davis,
B. H., in “Proceedings, 9th International Congress on Catalysis, Cal-
gary, 1988” (M. J. Phillips and M. Ternen, Eds.), Chem. Institute of
Canada, Ottawa, 1988.
60. Bond, G. C., Webb, G., and Wells, P. B., Trans. Faraday Soc. 64, 3307
(1968).
61. Nishimura, S., Sakamoto, H., and Ozawa, T., Chem. Lett. 855 (1973).
62. Schunn, R. A., Inorg. Chem. 9, 2567 (1970).
63. Kanai, H., J. Chem. Soc., Chem. Commun. 203 (1972).
64. Cairns, M. A., and Nixon, J. F., J. Organomet. Chem. 87, 109
(1975).
65. Madon, R. J., Reyes, S. C., and Iglesia, E., J. Phys. Chem 95, 7795
(1991).
40. Tau, L.-M., Dabbagh, H. A., and Davis, B. H., Energy Fuels 4, 94 66. Ko¨lbel, H., Ludwig, H., and Hammer, H., J. Catal. 1, 156 (1962).
(1990). 67. Shi, B., Bao, S., and Davis, B. H., unpublished results.
41. Dabbagh, H. A., Chawla, B., Halasz, J., and Davis, B. H., Fuel Sci. 68. Casey, C. P., and Cyr, C. R., J. Am. Chem. Soc. 95, 2248 (1973).
Technol. Int. 8, 719 (1990).
42. Tau, L.-M., Dabbagh, H., and Davis, B. H., Catal. Lett. 7, 141 (1990).
69. Rennard, Jr., R. J., and Kokes, R. J., J. Phys. Chem. 70, 2905 (1966).
70. Kokes, R. J., and Dent, A. L., Adv. Catal. 22, 1 (1972).
43. Tau, L.-M., Dabbagh, H., Bao, S., and Davis, B. H., Catal. Lett. 7, 127 71. Yamaguchi, T., and Hightower, J., J. Am. Chem. Soc. 99, 4201
(1990). (l977).