close

Вход

Забыли?

вход по аккаунту

?

Characterization by Mssbauer Spectroscopy of Iron Carbide Phases in a Highly Active Carbon Matrix Catalyst for Medium Pressure Fischer-Tropsch Synthesis.

код для вставкиСкачать
high energy. For geometry (3) this interaction is lost by symmetry and, thus, during the CoCp transit, (6) rises to high energy, ultimately becoming MO (8) of structure (3) (predominantly 271. in character). On the other hand, the antibonding
between 2e, and 1nsvanishes so that the LUMO (7) is lowered in energy and in addition begins to mix into itself 271, in
a bonding fashion with respect to 2e,, so that finally MO (9)
results.
alkyne ligand via (10) (23 kcal/mol). Thus, one would expect, that if the energy manifold leading from (1) to (4)were
accessible, then “cross products” of the type given by (11)
should be observed. Such is not the case for R = D[lA1,and
when R=SiMe,f’b’ Vollhardt and Fritch find (11) to be
formed with a higher activation energy than that for the
(I)-+ (1‘) rearrangement. The reason for the enormous energetic requirements of a CoCp shift (4) -,( 4 ) lies in the fact,
that all dominant frontier orbital interactions in (4) between
2e, and 2e, of CoCp and the highest occupied TT and the lowest unoccupied T* orbitals of two diacetylenes break down
for symmetry reasons in (5).
I71
I I71
1101
I VI
Our preliminary results therefore make all pathways of a
diethynylcyclobutadiene-CoCp rearrangement as shown in
Scheme 1 appear rather improbable.
A further mechanistic hypothesis for this unusual reaction
could involve the direct conversion of (I), R=II, to (12),
mentioned as “structural and electronic counterpart” of (3)
by Vollhardt and Fritch“”].
The symmetry imposed barrier of the CoCp migration is
calculated as 47 kcal/mol. The relatively high temperatures
used in the experiments and the fact that the CoCp unit need
not remain on the mirror plane bisecting the tricyclooctatetraene (which technically avoids the HOMO-LUMO correlation) would still make the pathway via (3) an attractive process. However, the conversion of (1) to (2) is also strongly
symmetry forbidden. The orbitals of the diethynylcyclobutadiene ligand are shown on the right side of Scheme 2. Of
these 171, and 271, form ideal bonding combinations with 2e,
and 2e,, respectively. The m-orbitals of both valence isomers
of the pure ligand correlate with each other nicely during the
cycloaddition and, therefore, they must as well in the CoCpcomplexes. The problem lies in the orthogonal (“in plane”)
71- and m*-orbitals of the two ethynyl groups of ( I ) , the S and
A combinations of which are easily constructed. Of these orbitals two are important, the S-combination of the m*-levels
and the A-combination of the r-levels. As the two ethynyl
groups approach each other to cycloadd to a four membered
ring, the symmetric combination of the m* orbitals finally
becomes a bonding a-orbital, the A-combination of the alkyne T-MO’S correlates to an antibonding a*-orbital of the
new ring. The transformation of (1) into (2) therefore is not
unlike the symmetry forbidden [m2s+ ,2,] dimerization of
ethylenef6].The activation barrier found for the rearrangement (1)+(1‘) of 37 kcal/mo1flb1seems to be too small for the
process to occur by way of (2) and (3). It can also be shown,
that the direct conversion of (1) into (3) [CoCp migration
and simultaneous ring closure] is doubly symmetry forbidden: two occupied A-levels become empty, while two unoccupied M O s of S symmetry correlate to occupied orbitals of
the product.
The direct retrocycloaddition of (1) to (4)is also strongly
symmetry forbidden. A a*-orbital of the cyclobutadiene ring
develops into a filled m-level of the alkyne ligands, a a-MO
correlates to n* orbitals. The reverse of this process has been
Our preliminatreated by Mango et al. in qualitative
ry calculations furthermore put a higher barrier on going
from (4)to ( 4 ) (60 kcal/mol) than that for the rotation of an
Angew. Chem. Int. Ed. Engl. 19 (1980) No. 9
Received: January 14, 1980 [Z 558 IE]
German version: Angew. Chem. 92, 747 (1980)
CAS Registry number:
(1). 67378-04-5
111 a) J. R. Frilch, K. P. C. Vollhordt, J. Am. Chem. Sac. 100, 3643 (1978); b) Angew. Chem. 91,439 (1979); Angew. Chem. Int. Ed. Engl. 18. 409 (1979).
[2] N . T. Anh, M. Elian, R. Hoffmann, J. Am. Chem. Sac. 100, 110 (1978).
I31 a) T. A. Albright, P. Hofmann, unpublished; b) T A . Albright, P. Hofmann, R.
Hofmann, P. Lillya, P. Dobosh, in preparation; c) P. Hofmann, R Hoffmann.
J. Am. Chem. Sac. 98, 598 (1976).
141 R. Hoffmann, J. Chem. Phys. 39, 1397 (1963). H,,-matrix elements: J. H. Ammeler, H:B. Burgi, J. C. Thibeaull, R. Hoffmonn. J. Am. Chem. Sac. 100,
3686 (1978). Parameters and geometrical details: P. Hofmann el 01.. unpublished.
151 Valence orbitals of a CoCp unit: M. Elian, M. M. L. Chen, D. M . P. Mingos,
R. Hoffmonn, Inorg. Chem. I S , 1148 (1976).
[61 R. B. Woodword, R. Hofmann, Angew. Chem. XI, 797 (1969); Angew. Chem.
Int Ed. Engl. 8, 781 (1969).
171 F. D. Mango, Fortschr. Chem. Forsch. 45, 39 (1974); F. D. Mango, J. H
Schachlschneider, J. Am. Chem. Sac. 91, 1030 (1969); F. D. Mango, Tetrahedron Lett. 1973. 1509.
Characterization by Mossbauer Spectroscopy of
Iron Carbide Phases in a Highly Active Carbon
Matrix Catalyst for Medium Pressure FischerTropsch
By Hartwig Schayer-Stahl [‘I
As crucial intermediates in the synthesis of hydrocarbons
from CO and HZon heterogeneous catalysts containing iron,
cobalt, and nickel (Fischer-Tropsch synthesis), the formation
of metal carbide phases was reported as long ago as the
1920’s by F. Fischer et at.[’],partly on the basis of results of
[*I
Dr. H. Schafer-Stahl
Fakultat fur Chernie der Universitat
Postfach 5560, D-7750 Konstanz (Germany)
[**I This work was supported by the Deutsche Forschungsgemeinschaft. 1 wish
to thank Prof. Dr. H . H. Brintzinger for valuable discussions.
@ Verlag Chemie, GmbH, 6940 Weinheim, 1980
0570-0X33/X0/0909-0729
$02.50/0
729
hydrolysis experiments with such catalysts (cf. also r21). This
concept is still a matter of contr~versyf~.~].
We now report observations on a highly active matrix catalyst which show that
the catalytically active sites are formed at iron carbide crystallites.
Reduction of iron halide-graphite intercalation compounds leads to destruction of the ordered host lattice, giving
a carbon layer matrix with a greatly increased surface area in
which highly dispersed iron particles are embedded whose
diameter is shown by Mossbauer spectroscopy and electron
m i c r o s c ~ p y [to
~ ,be
~ ~smaller than 2000 pm (20 A). This matrix catalyst mediates Fischer-Tropsch synthesis at temperatures as low as 140 "C in the medium pressure range''].
Under these conditions the original iron particles undergo
complete transformation. Superparamagnetic iron oxide and,
above all, ferromagnetic e- and X-iron carbide phases are recognizable in the Mossbauer spectrum. If the temperature of
Fischer-Tropsch synthesis is raised to 185 "C then the unstable iron oxide phases disappear and only magnetically ordered E- and X-iron carbide phases are present in the catalyst.
The activity of the catalyst increased concomitantly. Subsequent heating to 320 "C in a gas mixture having a high concentration of H2 effects complete phase transition to the xiron carbide phase (magnetically ordered at room temperature) which is then also stable at 140 "C (Fig. 1). This X-iron
carbide phase is the most active form of the investigated matrix catalyst in the temperature range around 140°C. Since
the phase is magnetically ordered at room temperature but
fails to show sharp X-ray reflections, the particle diameter
can be estimated as between several tens and several
hundreds of A.
The correlation between catalytic activity and content of
iron carbide phases, and particularly the X-iron carbide
phase, observed in all the operations described above can be
regarded as unequivocal proof that such carbide phases are
the actual catalytically active component, at least below
200 "C.
The matrix catalyst under investigation is particularly suitable for a structure-reactivity correlation. Owing to the extremely low diameter of the particles, which are essentially
isolated from one another by the disordered carbon layer
matrix and the exceptionally low reaction temperatures thus
possible, the formation, transformation, and growth of the
reactive phases can be recognized from the sequence of magnetically disordered iron, superparamagnetic iron oxide
alongside &-iron carbide, iron carbide mixtures, and ultimately pure X-iron carbide. Our results differ in this point
.'~
from those of similar studies on a-Fe p a r t i ~ l e d ~whose
larger diameters necessitated higher formation temperatures,
at which the temperature-dependent sequential formation of
oxide and carbide phases and their varying catalytic activities are unobservable.
The high proportion of surface atoms in the highly dispersed particles described above also has the effect that the
periphery of the particles makes a significant contribution to
the Mossbauer spectra relative to the bulk. Since none other
than carbide phases, and especially no a-Fe phasel3,'1, can be
observed at maximum activity, it may be assumed that X-iron
carbide is also present in the periphery of the reactive particles. It remains to be established how the carbide centers
activate the CO molecule for reduction and chain propagation reactions and whether analogies exist with recently described reactions of CO at iron carbide clusters in homogeneous sol~tion[~l.
Procedure
-
-
,
- 7.60
- 3.80
0.00
Y
[mm/r]
-
3.80
7.60
Fig. 1. Mossbauer spectrum of the matrix catalyst after 600 h operation at 140 "C
in its most active form. Almost the entire iron content is present in the form of
ferromagnetic x-carbide. (Recorded under protective gas with an ELSCINT
AMEN instrument, calibration with Ne-He laser interferometer and a-Fe foil.
Chemical shift relative to center of gravity of the a-Fe spectrum.)
However, if the catalyst is heated to 320°C in a gas mixture having a low H2 content then the X-iron carbide phase is
partly degraded to ferromagnetic Fe304;it no longer exhibits
Fischer-Tropsch activity at 140 "C. A residual catalytic activity can be detected at 185 "C; in the course of several days
the X-iron carbide phase is regenerated with a concomitant
increase in catalytic activity. A similar increase in activity is
observed when the partially oxidized catalyst is reduced with
hydrogen before it is re-used. After such regeneration small
amounts of ferromagnetic Fe304and of a-Fe are recognizable in the Mossbauer spectrum. Accordingly, the original activity of the pure X-iron carbide Phase is not quite reached.
This oxidation-reduction cycle is reproducible.
730
@ Verlag Chemie, GmbH, 6940 Weinheim, 1980
All operations were performed in an argon atmosphere
with exclusion of air and moisture. Flattened and comminuted lithium wire (11 g) and biphenyl (65 g ) are stirred in tetrahydrofuran (THF) (500 ml) at - 80 "C until dissolution is
complete. FeC13.C6., (50 g) is cautiously added with further
cooling. If the green color of the solution disappears further
lithium wire is added until the green color is retained. After
10 minutes' stirring at room temperature the solvent and volatile components are stripped off in uacuo and sublimed off
up to 100 "C, respectively.-Some 20 cm3= 10 g of the matrix catalyst prepared with biphenyllithium and having an
iron content of 0.98 g ( e0.0175 mol Fe) is examined with regard to its catalytic activity.
The Fischer-Tropsch reaction was carried out in a differential reactor with a membrane pump in the external synthesis gas cycle at pressures between 10 and 32 bar and a flow
rate of 18 I/h. The reactivity of the matrix catalyst follows
from the decrease in pressure of the synthesis gas with time.
The product spectrum (GC-MS) was similar to that obtained
on use of standard precipitation catalysts; it did not alter significantly during the catalyst transformations described
above.
Received February 13, 1980;
supplemented. June 18, 1980 [Z 560 IE]
German version: Angew. Chem. 92, 761 (1980)
CAS Registry numbers:
No compounds indexed
[I]F. Fischer, H . Tropsch, Ber. Dtsch. Chem. Ges. 59, 830 (1926); F. Fischer,
Bnhr, Gesammelte Abh. Kennt. Kohle 8, 255, 269 (1924-27).
Vielstich, Chem;Ing.-Tech. 52, 327 (1980).
121 M . Rizschel,
[3) D. J. Dwyer, G. A. Somorla;, J. Catal. 52, 291 (1978).
0570-0833/80/0909-0730
H.
w
$ 02.50/0
Angew. Chem. Int. Ed. Engl. I 9 (1980) No. 9
141 G. Henrici-Olive, S. Olive, J. Catal. 60, 481 (1979); R. J. Madon, ibid. 60, 485
(1979): P. A . Jacobs, H. H. Nijs, J. J. Verdonck, J. B. Uylterhoeven, Preprints,
Symposia Anaheim, CA 23, 469, ACS Div. Petrol. Chem.; H. Kolbel, D.
Schneidt, Erdol, Kohle, Erdgas, Petrochem. 30, 139 (1977).
[Sl B. Cleveland. C. R. Guamieri, J. C. Walker, Bull. Am. Phys. SOC. 15, 108
(1970).
(61 H. Schafer-Stahl, Chem.-Ing-Tech., in press.
(71 Increasing temperature, synthesis gas composition 2H,/CO (30 bar). For
comparison: Fischer-Tropsch catalysts based on graphite intercalation compounds reduced with potassium at 300 "C are catalytically active only above
300°C; M. fchikawa, M . Sudo, M . Soma, T. Onishi, K . Tamaru. J. Am.
Chem. SOC.91, 1538 (1969); V. J. Mashinskir, K A . Postinikov, Yu. N . Novikov. A. L. Lapidus, M. E. Vol'pin, Yo T. Eidus, Izv. Akad. Nauk SSSR, Ser.
Khim. 1976, 2018; S. Parkash, S. K . Chakrabartfy,J. G. Hooley, Carbon 15,
307 (1977).
[Sl J. A. Amelse, J B. Burt, L. H. Schwarlr, J. Phys. Chem. 82, 558 (1978); G. B.
Raupp, W. N Delgass, J. Catal. 58, 337, 348, 361 (1979).
191 J. S. Bradley. G. B Ansell, E W. Hill, J. Am. Chem. SOC.101. 7417 (1979).
Facile Synthesis of a-and P-0-Glycosyl Imidates;
Preparation of Glycosides and Disaccharides'"]
Aryl'
HA+ Aryl'
(zk R
= Bn,
(3): R = Bn,
(4): R = Bn,
f S k R = Bn,
Aryl'
Aryl'
Aryl'
Aryl'
= CsH5, Aryl' = p-CH3-CSH4
= Aryl' = CsH5
(6): R = Bn
f7): R = Ac
= Aryl' = p-C1-C,H4
= C,H5, Aryl' = p-(CH,)$-C,H,
By Richard R. Schmidt and Josef Michell']
Glycosides and saccharides are largely synthesized via haloses and their activation by heavy metal salts-especially
silver saltsI'-21.The disadvantages of this method are self-evident. Simplification of this approach should come from the
synthesis of sterically pure, readily isolable intermediates
with other leaving groups not requiring activation by heavy
metal s a l t ~ l ~Suitable
.~'.
candidates would seem to be, e. g., aand p-glycosyl imidates, since @-glycosyl imidates-prepared from a-haloses with silver salts-undergo acid-catalyzed reaction to give good chemical and stereochemical
yields of a-glycosides and a-~accharides~~~.
Thus a facile synthesis of a- and p-glycosyl imidates is called for''].
Ketenimines and nitriles containing electron-withdrawing
substituents are known to afford imidates directly on reaction with alcohols[61.We shall now demonstrate for the case
of the C-I -unprotected glucopyranose (1) that this reaction
can be applied to cyclic hemiacetals. Use of sodium hydride
as base and aryl-substituted ketenimines gave exclusively the
p-imidates (2)-(5) whereas the same reaction both with
benzyl- and with acetyl-protected glucopyranose ( I ) and
trichloroacetonitrile led diastereospecifically to the a-imidates (6) and (7) (Table 1). Both the 8-imidates and the aimidates, now prepared for the first time, could be conveniently isolated.
RO
"RO O
D
6R
OR
(&1j-(8e),
(a): R'
[hj:
(8h): R = Bn
= -CO-CH,
'
( 9 ~ ) - ( 9 h ) :R = Bn
( 9 1 ) - ( 9 k ) : R = Ac
(f):R' =
R' = p-CGH4-CH3
(c): R' =
( d ) : R' =
OBn
(h): R' =
H3C CH3
Table 1. 0-Glycosyl imidates (2)-(7)
(e): R1 =
prepared [a]
O
(2)
(3)
(41
(5)
(6)
84
83
93
38
96
6.11
6.11
6.05
6.1 1
6.56
5.22
5.22
5.04
5.30
8.60
7.0
7.0
7.2
7.5
3.5
(7)
85
6.60
8.77
3.5
+
+
+
1670
1650
1670
1670
1670
(Y~H:
+
+
33.0
33.0
34.5
68.7
61.5
Q
(i): R' = C & ,
2.0
1.75
1.88
1.6
1.0
3320)
1680
t103.0
1.2
( Y ~ H 3330)
:
(YCO:
1755)
[a] Abbreviations: Ac= acetyl; Bn= benzyl. All compounds gave correct elemental analyses. [h] Isolated yields. [c] 80 MHz spectra in CDCI, with tetramethylsilane as internal standard; 6 values, coupling constants in Hz. [d] [cm - 'I, film between NaCl plates. [el In CHCI,.
['I
Prof. Dr. R. R.Schmidt, DipLChem. J. Michel
Fakultat fur Chemie der Universitat
Postfach 5 5 60, D-7750 Konstanz (Germany)
["I This work was supported by the Deutsche Forschungsgemeinschaft and the
Fonds der Chemischen Industrie.
Angew Chem Inr. Ed. Engl. 19 (1980) No 9
As expected, the beruyl-protected p-imidates (2)- (5) undergo acid-catalyzed reaction with hydroxy components in
dichloromethane to form mainly or exclusively the a-linkage
product (8) while the a-imidate (6) correspondingly affords
the @-linkageproduct (9) (see Table 2). The use of carboxylic
acids [- @a), (SJ)], p-cresol [-+(8b)],steroid alcohols [ + (ad),
(9d); (8e), (9e)], and carbohydrates [-(8c), (9c); (9g); (8h),
(9h)l demonstrates the considerable scope of this simple
method of glycosidation; however, the stereochemical result
has not yet been optimized in all cases. The particularly
readily accessible acetyl-protected a-imidate (7) reacts on ca-
0 Verlag Chemre, GmbH, 6940 Wernherm, 1980
0570-0833/80/0909-0731
$02.50/0
731
Документ
Категория
Без категории
Просмотров
0
Размер файла
349 Кб
Теги
matrix, pressure, tropsch, fischer, phase, medium, spectroscopy, synthesis, activ, iron, mssbauer, characterization, carbon, catalyst, highly, carbide
1/--страниц
Пожаловаться на содержимое документа