close

Вход

Забыли?

вход по аккаунту

?

Fei-Catalyzed Gas-Phase Oxidation of Ethane by N2O.

код для вставкиСкачать
slowly (3 h) warmed to room temperature and then the volatiles were removed
in vdcuo. Following treatment with pentane, filtration, and evaporation, 2.4 g
(96 % ) of diazo 4a was obtained. Carbene 4 b was prepared by heating a toluene
solution (10 mL) of 2.4 g of diazo 4a, at 35'C for 1 h.
N2O
d
k , = 0.4
CAS Registry numbers:
l a , 97135-48-3; Ib, 115797-19-8; lc, 130195-59-4; 2a, 130148-60-6; Zc,
130148-62-8; 2d, 130168-01-3; Ze, 130148-61-7; 3a, 130148-63-9; 3b, 13014866-2; 4a. 130148-64-0; 4b, 130148-67-3; 4e, 130148-70-8, Sa, 130148-65-1 ; 5b,
130148-68-4; 5e, 130148-71-9; 6a, 130168-02-4, 6b, 130148-69-5; 6e, 13014872-0; 29Si,14304-87-1.
By Detlef Schroder and Helmut Schwarz *
Although thermodynamically favored, many oxidation
reactions do not occur at room temperature unless the kinetic barriers are lowered by catalysts. This fact is a prerequisite
for the origin and maintenance of life,['] but it also poses an
'obstacle to the in vitro transformation of simple substrates
to "more valuable" products.
Here we describe a model reaction in which oxidation of
ethane by N,O [Eq. (a)] is catalyzed by Fee. Although the
process is strongly exothermic (AHr= - 121 kcal mol-I), it
does not occur in the absence of Fee.
+ CH,CHO
+ H,O + 2 N,
(a)
The oxidation takes place in the gas phase and the individual steps of the FeO-catalyzed cycle (Scheme 1) can be studied by Fourier transform ion cyclotron resonance (FTICR).'~]
The following observations were made concerning the individual reactions:
1 . In the reaction of FeOO with C,H,, already investigated by Freiser et al.,"] the encounter complex [FeO/C,H,]@*
[*] Prof. Dr. H. Schwarz, Dip1.-Chem. D. Schroder
Institut fur Organische Chemie der Technischen Universitit
Strasse des 17. Juni 135. D-1000 Berlin 12 (FRG)
I**] This work was supported by the Deutsche Forschungsgemeinschaft. the
Fonds der Chemischen Industrie, and the Gesellschaft von Freunden der
Technischen Universitit Berlin.
Angew. Chmr Itrr. Ed. Etigl. 29 (I9901 N o . 12
0 VCH
4
k,=1.8
0
FeCO
0
Fe-OH,
FeOH@
0
FeC,H,
I
FeCHF
Scheme 1. Catalytic cycle for the FeO-mediated oxidation of C,H, by N,O.
The rate constants kj, given in cm3 molecule-' s - I , have to be multiplied by
10-10.
FeO@
Dedicated to Projessor Albert Eschenmoser on the occasion
of his 65th birthday
+ 2 N,O
L
is not observed. Instead, the products of three reaction paths
were found [Eq. (b)]. The corresponding rate constants ki are
given in Scheme 1. All processes are strongly exothermic.[81
Fe@-Catalyzed Gas-Phase Oxidation of Ethane
by N,O**
C,H,
k , =0.7
"C2H,0"
Received: July 6, 1990 [Z 4057 IE]
German version: Angew. Chem. f 0 2 (1990) 1486
[l] a) 0. Glemser. R. Mews, Angew. Chem. 92 (1980) 915; Angew. Chem. Inr.
Ed. En$ 19 (1980) 883; b) G. Bertrand, J. P. Majoral. A. Baceiredo, Acc.
Chem. Res. I Y (1986) 17; c) D. Christen, H. G. Mack, C. J. Marsden. H.
Oberhammer, G. Schatte, K. Seppelt, H. Willner, J Am. Chem. Sor. 109
(1987) 4009.
[2] a ) A. Igau. H. Grutzmacher, A. Baceiredo, G. Bertrand, J Am. Chem. Soc.
i1lJ (1988) 6463; b) A. Igau. A. Baceiredo, G. Trinquier, G. Bertrand,
Angew. Chem. 101 (1989) 617; Angeu. Chem. Int. Ed. Engl. 28 (1989) 621
13) a) H. Keller, M. Regitz, Terruhedron Lett. 29 (1988) 925; b) H. Keller, G.
Maas. M. Regitz, ihid. 27(1986) 1903; c) T. Facklam. 0.Wagner, H. Heydt,
M . Regitz. Angew. Chem. 102 (1990) 316; Angew. Chem. Int. Ed. EngI. 29
(1990) 314; d) H. Keller, Disserration. Universitit Kaiserslautern, 1988;
e) T. Facklam, Disserrnrion, Universitat Kaiserslautern, 1989.
[4] T. Aoyama, S. hove, T. Shioiri, Tetrahedron Leu. 25 (1984) 433.
[5] J. Boske, E. Niecke, E. Ocando-Mavarez, J. P. Majoral, G. Bertrand, h o g .
Chetn. 25 (1986) 2695.
N2
+ C,H,
67 Yo
AHr
[kcal mol-
+ H,O
Fe(OH,)@ + C,H4
-31
FeO + CH,CH,OH
- 13
Fe(C,H4)0
'1
-36
(b)
A double resonance experiment demonstrates that FeO@
is the direct precursor of all products shown in Equation (b):
irradiation at the resonance frequency of FeO@ results in
suppression of the reactions. Collisional activation and labeling experiments provide indications of the structures of
the Fe(C,H,)@ and Fe(OH,)@ ions. The iron-hydrocarbon
complex at m/z 84 specifically loses C,H, upon collisional
activation with argon; this finding is consistent with an
Fe(ethylene)@structure. The dominating reaction of the ion
at m/z 74 with argon is loss of H,O; in addition, ca. 10 % loss
of H' or OH' is observed. If, instead of an Fe(OH,)@ complex, an Fe(H)OHe structure were present, preferential H'
loss would be expected owing to the dissociation energies
D(MO-X) (X = H, OH).[91Moreover, the thermochemical
data favor the structure Fe(OH,)@ over Fe(H)OHe by more
than 14 kcal mol-'.[*] The role of Fe(OH,)@ as a "sink" in
the catalytic cycle will be discussed in greater detail later.
The encounter complex (FeO/C,H,)@* presumably corresponds to Fe(C,H,)OHe*, as indicated by the product distribution obtained from the reaction of FeOO with
CH,CD,: decomposition of (FeO/CH,CD,)@* yields almost exclusively HDO and C,H,D, as neutral molecules.
2. Reaction of "isolated" Fe(C,H,)@ with N,O results in
liberation of N, [Eq. (c)]. Since the liberated N, takes away
~ r l a g . ~ g e s ~ ~ I l .mhH,
~ ~ h n W-6940
//t
Weinheim, 1990
0571)-0N33~90~1212-i431
$3.50+ .ZS/O
1431
part of the excitation energy as translational and vibrational
energy, it is not surprising that a signal for Fe(C,H,O)@
is observable. However, the main products are Fee
and “C,H,O”. The two reactions account for 88% of the
product distribution; the products corresponding to k , in
Scheme 1 (C < 12%) are formed in nearly equal amounts
(X = CO, OH, C,H,, CH,).
16 VO
Fe(C,H,)@
72%
+ N,O
Fe(C,H,O)@
Fee
+ C,H40
FeX@
3. The ion Fe(C,H,O)@ is most likely a ferraoxacyclobutane, since, upon collisional activation, the “isolated” ion
decomposes practically exclusively to FeCHp/CH,O and
Fe@/C,H,O. The formation of FeCHp would not be possible from either an Fe(CH,CHO)@ or an Fe(CH, =CHOH)@
complex. That the metallacycle does not decompose to
FeO@/C,H, upon collisonal activation is due to the product
stability for the two other conceivable cycloreversions;
FeO@/C,H, is at least 20 kcal mol-’ less favorable than
FeCHf)/CH,O.
4. The central question posed by the reaction
N,O + Fee + C,H,O [Eq. (c)] concerns the
FeC,Hp
constitution of “C,H,O”. Chemically reasonable structures
for the neutral compound include oxirane, vinyl alcohol, and
acetaldehyde. As shown by Scheme 2, the formation of each
+
Fe(C,H,)a+
0
/ \
Fe(butadiene)@
+ H,C-CH,
@Fe-0
I
I
H,C-CH,
+ C,H,
(e)
away a considerable part of the energy as kinetic energy. This
assumption is supported by the observation that an encounter complex is not detectable in the reaction FeO@+
C,H, + Fe(C,H,O)@, although this process is 9 kcal mol-’
less exothermic; instead, only subsequently formed products
are observed. If at least 9 kcal mol-’ of internal energy is
transferred to N, as translational energy on formation of
Fe(C,H,O)@ from Fe(C,H,)@ and N,O, however, then
oxirane is necessarily ruled out for thermochemical reasons
as the oxidation product of C,H,. This supposition is supported by an independent experiment: Fee, thermalized
with argon (p N 5 x l o - * mbar) for about 20 s, reacts with
added oxirane to give, among other products, FeO@ and
C,H, (25 %); this means that Fe(C,H,)@/N,O would react
via Fe@/C,H,O to give FeOe/C,H, if C,H,O corresponded
to oxirane. This is not observed in the experiment [Eq. (c)].
The decision in favor of acetaldehyde is facilitated by a
“mechanistic criterion”. Armentrout, Beauchamp et a1.16~
have indicated that the weakest bond in a metallacycle of
type 1 is the metalkarbon bond. Cleavage of this bond
could lead, via diverse ion/dipole complexes, to a “metalbound” diradical 3 (Scheme 3). Experimental data and MO
6,
N,O
Fe
+ “C,H,O”
-N>
Fe(C,H,)
0
+ N,O
280
FeOa+ C,H,
F>*’+
271
A
1
270
2
3
4
Scheme 3.
Fe“+ CH,=CHOH
256
Fe@+ C H p O
240
<236
0
Fe-0
I I
H,C-CH,
>222
Scheme 2
of the three isomers starting from Fe(C,H,)@/N,O would be
exothermic. Since neutral species cannot be directly observed
in ICR experiments, indirect evidence is necessary. Nonetheless, the appearance of a weak-intensity signal at m/z 43
(CH3CO@)proves that, in the course of this reaction, oxygen
must have been transferred to the C,H, unit.
The enthalpy of formation AK of the metallacycle
Fe(C,H,O)@ can be estimated from the occurrence or absence of the ligand displacement reactions (d) and (e):
222 < AH: (FeC,H,O)@ < 236 kcal mol-’.
The observation of the metallacycle Fe(C,H,O)@ despite
the strong exothermicity of the reaction [Eq. (c)] must be due
to the fact that the simultaneously formed nitrogen carries
1432
$>#VCH Verlu~.~ge,srli.srhuft
mhH. W-6940 Weinlwm, 1990
calculations[‘ indicate that a free diradical OCH,CH,
should isomerize to CH3CH0. If this also holds true in the
presence of Me, then complex 4 would be formed in the
present case. It should be mentioned that the side product
Fe(C0)” (4%) in reaction (c) could be formed from 4 by
reductive elemination of CH,. The main path, however,
would be loss of CH3CH0 and regeneration of Fe@,which,
in turn, could initiate a new cycle. Beyond any doubt, Fee
junctions as a catalyst in reaction ( a ) . From the rate constants given in Scheme 1 , each Fee ion is estimated to convert about 2.5 molecules of C,H,. Where are the “sinks”
that keep the turnover number so small? (1) The metallacycle 1 (M = Fe) not only decomposes to Fe@and CH3CH0
with a rate constant of k , = 0.4 x lo-’* cm3 molecule-’
s-’, but also affords other side products ( k , = 0.1 x
lo-’’ cm3 molecule-’ s-’); common to all is the loss of
Fe@. (2) A strong sink is the reaction path with the rate
constants k , : Fe(OH,)@ undergoes rapid further reaction
with N,O to give Fe(OH)p, which, however, only reacts
with C,H,/N,O at an unmeasurably slow rate; thus,
Fe(0H)P represents a dead end. It is noteworthy that FeOO
reacts with C,H, whereas Fe(0H)P reacts only sluggishly:
although both iron-containing ions have a formal oxidation
state of + 3, the oxidation state of the metal is not the only
governing factor for an efficient activation.
0570-0833190/1?12-1432 $3.51)+ .2S/O
Angen. Chem. Int. Ed. Engl. 29 11990) No. 12
Finally, the result of a further test should be mentioned.
This test my be regarded as a measure of the reliability of the
kivalues in Scheme 1 (which themselves were derived from
the reaction of “isolated” intermediates). Use of the soderived rate constants for determination of the time dependence of the ions Fee, FeO@, Fe(C,H,)@, and Fe(0H)Y
appearing in the catalytic cycle gives the data shown in Figure 1. Theory (solid line) and experiment are in good agree-.
[Sj a) M. M. Kappes. R. H. Staley, J. Am. Chcm. Soc. 103 (1981) 1286:
b) M. M. Kappes, R. H. Staley, J Phys. Chem. 85 (1981) 942.
[6] S. K. Loh, E R. Fisher, L. Lian, R. H. Schultz, P. B. Armentrout, J. f / i p
Chem. 93 (1989) 3159.
[7] T. C . Jackson. D. B. Jacobson, B. S . Freiser, J. Am. Chem. Soc. /06(1984)
1252.
[8] All reaction enthalpies were estimated by using known AH< values (kcal
mol-’): Fee 282; FeOe 259; Fe(C,H,)e 260; Fe(H,O)e 196; Fe(H)OHe
211 -223; Fe(butadiene)e 261; N,O 20; C,H, 12; C,H, - 20; HzO - 57;
k.H,CH,O - 12; CH,=C(H)OH - 26; CH,CHO - 44; CH,CH,OH
- 56,CH,=CH-CH=CH2 27.
[Y] T. F. Magnera, D. E. David, J. Michl, J Am. Chem. SOC.//1(1989) 4100.
1101 a) L. F. Halle, P. B. Armentrout, J. L. Beauchamp, Orgunomelul1ic.r 2
(1983) 1829; b) H. Kang, J. L. Beauchamp, J. Am. Chem. Soc 108 (1986)
5663; c) E. R. Fisher, P. B. Armentrout, J. P h w . Chem. 94 (1990) 1674.
[ l l ] a)A. Lifshitz, H. Ben-Hamon, J. Chem f1r.v.s. X7 (1983) 1782; b ) J. G.
Serafin, C. M. Friend, J Am. Chem. Soc. I 1 t (1989) 6019; c) F. W. McLafferty, Science (Wushingron D.C.) 247 (1990) 925; e) C. Wesdemiotis. B.
Leyh, A. Fura, F. W. McLafferty, J. Am. Chem. Soc., in press.
FeO@Activates Methane **
By Detlef Schroder and Helmut Schwarz *
Dedicated to Professor Rolf Huisgen
on the occasion of his 70th birthday
0
50
-
100
t Is1
150
200
Fig. 1 Time dependence of ion abundance for Fee, FeOe, Fe(C,H,)e, and
Fe(OH)y in the range 0 < I < 200 s. The solid lines are based on a numerical
solution of the respective differential equations using the k , values given in
Scheme 1 (see text). .,m.o,o = measured values.
ment in view of the known errors involved in FTICR intensity measurements. Comparison of theory with experiment
reveals the following: (1) No noticeable excitation occurs on
“isolation” of the intermediates. (2) Additional intermediates need not be considered in modeling the catalytic cycle.
Received. June 15. 1990 IZ 4017 IE]
German version: A n g e x Chem. 102 (1990) 1466
CAS Registry numbers:
[FeOCH,CH,Ie. 129877-78-6; C,H,. 74-84-0; N,O, 10024-97-2. Fee, 1406702-8; FeOe, 12434-84-3; [Fe(H,C =CH,)Ie, 109801-95-8.
[ I ] A. Eschenmoser, Angeiv. Chem. 100(1988) 5 ; Angew Chem. Int. Ed. Engl.
27 (1988) 5 .
[2] Conditions: Fee ions were generated from a stainless steel target in the
external ion source of a Spectrospin CMS 47X mass spectrometer[3] by
laser desorption/ionization using an Nd-YAG laser (Spectron Systems;
.; = 1064 nm)[4]. The metal ions were transferred to the actual ICR cell by
means of a system of electrical potentials and lenses. After deceleration,
they were trapped in the magnetic field of a superconducting magnet (Oxford Instruments; maximum field strength 7.05 T). The isotope “Fern was
“isolated” by a double resonance experiment and allowed to react with
N1O ( p = 8 x
mbar) to give FeOe [ S ] . Since this reaction is strongly
exothermic (1.8 eV)[6], the products were formed with considerable excitation energy. which, however, was removed by collisions with the environment FeO“‘ was “reisolated” and allowed to react with C,H,
( p = 1.2 x
mbar) as described by Frerser[7]. The main product was
FeC,Hp (70%). which, after renewed “isolation”. was allowed to react
again with N,O. Mainly (>SO%) “C,H,O” and Fee were thereby
formed. The regenerated Fee completes the catalytic cycle and is available
for a new cycle. It should be noted that the rate constants k , are independent of the “age” of Fee within the error limits of f 15%
[3] Detailed description: a) K. Eller, H. Schwarz, In!. J. Muss Spertrom. fon
frociwe.s 93 (1989) 243; b) K. Eller, W. Zummack, H. Schwarz. J. Am.
Chm. Soc. 112 (1990) 621.
141 a ) B. S Freiser, Tulimtu 32 (1985) 697. b) B. S Freiser. Anal. Chim. Arm
178 (1985) 137.
Angm
(%mi.
Int. Ed. Engl. 29 ,1990) N o . 12
The metal-mediated activation of methane is the most important of all CH bond activations.“] This is due not only to
the great economic interest in the direct conversion of methane to “valuable” products such as methanol, but also to the
inherent scientific challenge, which still remains to be met in
a general way. The approach used succesfully for many substrates, namely, to obtain information on the detailed steps
of the activation of CHjCC bonds by “naked” transitionmetal ions,I2]has failed for the simple reason that the oxidative addition of most metal ions M@(M = Ti, V, Co, Fe, Nb,
Rh, Sc, Y, La, Lu, Co, Ni, Zn, and U)[31to CH, [Eq. (a)] is
end other mi^.'^'
CH,
+ Me
+ H-Ma-CH,
+products
(a)
The only exceptions are those cases where either M0 is
electronically excited (e.g., Ti0 or Cre I6])or the barrier to
oxidative addition is overcome by kinetic excitation of
M@.12b.71The reaction of “naked” metal atoms M
(M = Mn, Fe, Co, Ni, Zn) with CH, in a noble-gas matrix
is also only possible after photochemical excitation of M.[81
The often useful approach of activating CHjCC bonds not
by single metal ions but rather by homo- or heteronuclear
ionic clusters[91has also proved fruitless.[”]
A conceivable way of favorably influencing the reaction
enthalpy for activation of methane is the use of MOO instead
of M@.[”]Indeed, Freiser et
were able to increase the
propensity of alkanes to react by use of FeO@ instead of
Fee. The authors reported, however, that “cold” FeOe did
not react with CH, in the gas phase, even though all processes shown in Equation (b) are exothermic.[131
We have not been able to confirm this
FeO*,
generated under FTICR (Fourier transform ion cyclotron
resonance) conditions[’41from Fee and N,O, affords all the
products shown in Equation ( b ) in a ratio 0 f 5 7 : 2 : 4 1 . “ ~~The
possibility that excited FeO@ might be generated[”] and
[*] Prof. Dr. H. Schwarz, DipLChem. D. Schroder
Institut fur Organische Chemie der Technischen Universitit
Strasse des 17. Juni 135. D-I000 Berlin 12 (FRG)
I**]
This work was supported by the Deutsche Forschungsgemeinschaft. the
Volkswagen-Stiftung, and the Fonds der Chemischen Industrie.
X j VCH Verluxsgesellschafi mhH. W-6940 Weinhrim. 1990
0570-0833/90~1212-1433$3.50+ . 2 , 0
1433
Документ
Категория
Без категории
Просмотров
0
Размер файла
347 Кб
Теги
oxidation, ethan, gas, fei, phase, n2o, catalyzed
1/--страниц
Пожаловаться на содержимое документа