Experimental evidence from H/D exchange studies for the failure of direct C-C coupling mechanisms in the methanol-to-olefin process catalyzed by HSAPO-34

Article abstract of DOI:10.1002/anie.200504372

(Chemical Equation Presented) Pool view preferred: In agreement with recent theoretical work, a new line of experimental evidence was obtained in support of the controversial claim that direct mechanisms do not couple methanol to ethylene in the catalyzed methanol-to-olefin process. The results preclude carbene routes and oxonium ylide mechanisms, among others (see picture), in favor of the hydrocarbon pool mechanism.

Full text of DOI:10.1002/anie.200504372

Angewandte
Chemie
remains a stimulus to mechanistic progress in heterogeneous
catalysis. The general features of this reaction on a working
catalyst (i.e., following a kinetic induction period) are widely
accepted as a hydrocarbon pool mechanism on the basis of
experimental[3–6] and theoretical work.[7,8] Prior to recognition
of the hydrocarbon pool mechanism, speculation centered on
any of a dozen or more “direct” mechanisms in which,
typically, an activated single carbon fragment reacted with
methanol or dimethyl ether (DME) to form ethylene, methyl
ethyl ether, or some other species with a “first” carbon–
carbon bond.[1] Note that there are still important additional
contributions that theory can make, such as a thorough
comparison of detailed hydrocarbon pool reaction steps to
putative direct reaction steps. Previously,[9] we used highly
purified reagents to call into question whether “direct”
mechanisms occur at all, even at very low rates during the
kinetic induction period. In this view, the original hydro-
carbon pool forms from impurities rather than a “direct”
route, and a mature hydrocarbon pool then develops from
side reactions of propene or other products.
Very recently, a pair of theoretical papers[10,11] examined
essentially all direct pathways from methanol to species with a
À
first C C bond. Those studies concluded that no combination
of direct steps exist that is capable of linking methanol with
ethylene, even at a temperature of 720 K. Some of the direct
routes examined involved the intermediacy of framework-
bound methoxy groups, Z(O)-CH3, which indisputably form
from a fraction of acid sites in the equilibration of methanol,
dimethyl ether, and water on the catalyst (Scheme 1).[6,12] The
Methanol-to-Olefin Process
DOI: 10.1002/anie.200504372
Experimental Evidence from H/D Exchange
À
Studies for the Failure of Direct C C Coupling
Mechanisms in the Methanol-to-Olefin Process
Catalyzed by HSAPO-34**
Scheme 1. Equilibration of methanol over framework acid sites in the
MTO catalyst, HSAPO-34.
David M. Marcus, Kelly A. McLachlan,
Mark A. Wildman, Justin O. Ehresmann,
Philip W. Kletnieks, and James F. Haw*
ease with which Z(O)-CH3 can be observed by spectroscopy
has probably contributed to the attraction of framework
methoxy groups as key intermediates in proposals of direct
reactions of methanol on HSAPO-34[13] or methyl halides on
basic catalysts.[14]
The conversion of methanol to ethylene and propene on
microporous solid acid catalysts[1,2] such as the silicoalumino-
phosphate HSAPO-34at temperatures of 573 K and above
In their theoretical study,[11] Waroquier and co-workers
reject equilibration of Z(O)-CH3 with carbene (Scheme 2) as
[*] D. M. Marcus, K. A. McLachlan, M. A. Wildman, J. O. Ehresmann,
P. W. Kletnieks, Prof. J. F. Haw
Department of Chemistry
University of Southern California
Los Angeles, CA 90089-1661 (USA)
Fax: (+1)213-740-0930
E-mail: jhaw@usc.edu
[**] This work was supported by the US Department of Energy (DOE)
Office of Basic Energy Sciences (BES; grant no. DE-FG03-
99ER14956).
Scheme 2. Putative dissociation of Z(O)-CH3 to carbene. Carbene
[15]
À
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
could also insert into an O D bond to form Z(O)-CH2D, and H/D
exchange would be evident.
Angew. Chem. Int. Ed. 2006, 45, 3133 –3136
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3133

Communications
a possible direct route owing to a theoretical barrier of
pairing of equal amounts of Z(O)-CH3 and Z(O)-CD3, and
242.1 kJmolÀ1 and a theoretical rate constant of 5.3 ” 10À7 at
720 K (reaction J1 in reference [11]). A second possibility
involving the conversion of Z(O)-CH3 and dimethyl ether to
trimethyloxonium ylide (Scheme 3, reaction X1 and analo-
gous reaction X2 in reference [11]) was also shown not to
exist.
the absence of C H activation is reflected in the absence of
À
[D1], [D2], [D4], and [D5] isotopomers. Figure 1b is the
opposite case, with complete, random H/D exchange resulting
in a statistical distribution of all isotopomers and a very
characteristic mass pattern. Figure 1c shows the case of
blending 90% no exchange and 10% complete exchange,
allowing for the possibility of exchange at
a hypothetical “minority site”. This case
demonstrates the accuracy of our fits—
HSAPO-34is a highly uniform material
and there is no evidence of any site
diversity. The experimental mass pattern
observed for the experiment in Scheme 4
as well as the fit to the data are shown in
Scheme 3. Putative formation of trimethyloxonium ylide from Z(O)-CH3 and
[D3]DME. Exchange of H/D labels would be facile if the ylide formed at all.
À
As all possible routes from Z(O)-CH3 to species with a C
À
C bond must involve C H bond activation, we devised a
simple and conclusive test of Scheme 2 and, by extension,
many other direct reaction mechanisms, based on the
measurement of H/D exchange between Z(O)-CH3 and
Z(O)-CD3 groups that coexist on the catalyst. Note that
observation of exchange would not prove any of these
mechanisms, but the opposite would be decisive.
The essence of our simple experiment is shown in
Scheme 4. By pulsing a particular isotopomer of [D3]DME
Scheme 4. Formation of framework-bound methoxy groups and their removal
without H/D exchange.
(CH3OCD3) onto the preferred methanol-to-olefin (MTO)
catalyst, HSAPO-34, at 573 K we were able to convert about
30% of the acid sites in a catalyst bed to methoxy groups,
equally Z(O)-CH3 and Z(O)-CD3. For the experiment
depicted in Scheme 4we raised the reactor temperature to
623 K and gave the intimate mixture of Z(O)-CH3 and Z(O)-
CD3 two hours to exchange labels, prior to sweeping the
methyl groups out of the catalyst as mixed isotopomers of
[Dn]DME using a pulse of HOD (i.e., 50% D2O). To ensure
an equal balance of hydrogen and deuterium, the catalyst bed
was also first treated with 50% D2O prior to activation and
formation of methoxy groups. As with our previous work,[9]
the use of highly purified reagents strongly suppressed the
formation of hydrocarbons in these experiments.
The mixture of [Dn]DME isotopomers released from the
catalyst was quantified using online GC-MS. Figure 1 reports
three hypothetical, simulated ion mass distributions for such
an experiment, corresponding to interesting exchange con-
ditions. Figure 1a shows the case of no exchange. As
suggested in Scheme 4, a distribution of 25% [D0]DME,
50% [D3]DME, and 25% [D6]DME arises from the chance
Figure 1. Ion mass distributions for specific mixtures of partially
deuterated dimethyl ether, modeling the experiment suggested in
Scheme 4. a) Simulated distribution for a mixture of 25% [D0]DME,
50% [D3]DME, and 25% [D6]DME (i.e., the case of no H/D exchange
beyond recombination of intact methyl and [D3]methyl groups to
dimethyl ether). b) Simulated distribution for complete randomization
À
of hydrogen and deuterium by C H bond activation. c) Simulated
distribution assuming that only 10% of methoxy groups undergo
complete exchange. d) Experimental distribution (and best-fit simula-
tion) for dimethyl ether liberated after 120 minutes at 623 K from a
catalyst bed deuterated to 50% and then treated with 0.50 equivalents
of CH3OCD3 to form equal concentrations of [D0]methoxy and
[D3]methoxy species. It is estimated that our detection limit for
methoxy H/D exchange corresponds to 1% of the acid sites in the
catalyst bed. The fit in Figure 1d is in excellent agreement with
experimental data and shows no evidence of H/D exchange what-
soever.
3134
www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 3133 –3136

Angewandte
Chemie
obtained from Cambridge isotopes (lot number R-444, PSO 4C-117),
Figure 1d. The experimental data in Figure 1d are in out-
standing agreement with the prediction for no H/D exchange
in Figure 1a. Indeed, measurements performed at various
times and temperatures (see Table 1) show that the low
entries for [D5]DME and [D2]DME in Figure 1d are fitting
as was D2O (99.9% deuterium content).
See Supporting Information for
a
schematic and detailed
description of our reactor. We used 300-mg catalyst beds of
HSAPO-34catalyst, rigorously calcined (873 K, 100 standard cubic
centimeters per minute (sccm) flowing air) for 1 h just prior to use.
The carrier gas was then switched to flowing He (600 sccm), and the
temperature was lowered to 648 K. Then, three pulses (3.3 equiv,
20 mL) of 50% D2O in H2O were applied at five-minute intervals to
bring the catalyst bed to 50% deuteration. The catalyst bed was held
at 648 K with He gas flow for 30 minutes to ensure complete removal
of water. At 573 K we loaded the catalyst with an equimolar mixture
of [H3]- and [D3]methoxy groups by pulsing two aliquots (0.25 equiv
per aliquot) of CH3OCD3 onto the catalyst bed. The reactor
temperature was then immediately raised to 623 K and gas flow was
maintained for 2 hours (the temperature was lowered back to 573 K
just prior to the end of this period), or the temperature was
maintained at 573 K for a variable time between 6 minutes and
2 hours, before hydrolyzing the chemisorbed methoxy species with a
single pulse (5 equiv) of 50% D2O in H2O to liberate dimethyl ether
for isotopic analysis by GC-MS. (Note: DME was chosen for this
analysis to avoid the complication of the readily exchangeable
hydroxy proton of methanol; i.e. CH3OD versus CH2DOH.)
An Agilent gas chromatograph equipped with dual 50-m
Supelco DH columns and both mass spectrometric and flame-
ionization detectors (FID) was used to analyze the reactor effluent
gases. The temperarture of the oven was held at 308 K (isothermal).
The mass spectrometer was used to obtain isotope distributions as
needed, while the FID provided quantified peak integrations (using
published response factors) as well as significantly lower detection
limits.
Table 1: H/D exchange as a function of time in framework-bound
methoxy groups on H0.50D0.50SAPO-34.
[Dn]DME
Distribution [%] at various times and temperatures
6 min, 15 min, 30 min, 60 min, 120 min, 120 min,
573 K 573 K
573 K
573 K
573 K
623 K
[D6]
[D5]
[D4]
[D3]
[D2]
[D1]
[D0]
26.5
1.4
0.3
48.0
1.5
0.0
26.8
1.3
0.0
48.4
1.0
0.0
26.9
1.3
0.0
48.2
1.4
0.0
26.9
0.8
0.0
48.7
0.7
0.0
26.9
0.6
0.3
49.1
1.0
0.0
26.3
1.3
0.0
48.1
1.5
0.0
22.4
22.5
22.2
22.9
22.2
22.8
errors caused by minor variations in the experimental spectra
and constraints in the model to preclude negative concen-
trations. The upper limit on the rate constant for the process
in Scheme 2 (J1 in reference [11]) at 623 K is estimated to be
2 ” 10À6, but it could also be much lower.
We also used GC-MS to determine that neither the
methanol or dimethyl ether that exit the reactor in the
seconds following the CH3OCD3 pulse showed unexpected
isotope exchange. Hence, methoxy groups are not deproton-
ated by methanol, dimethyl ether, or water at 573 K. Indeed,
Received: December 8, 2005
Revised: February 16, 2006
Published online: March 29, 2006
À
these observations show that C H bonds are not broken or
made in any reaction step that is reversibly coupled to
methoxy formation (see Scheme 3, or either X1 or X2 in
reference [11]).
Keywords: heterogeneous catalysis · isotopic labeling ·
.
methanol-to-olefin process · zeolites
We conclude that the only valid mechanisms for the
methanol-to-olefin conversion are hydrocarbon pool mecha-
nisms that describe primary synthesis of olefins on organic
reaction centers, such as methylbenzenes, coupled with well-
known secondary reactions of olefins. This is true not only for
a working catalyst but also for initiation reactions during the
kinetic induction period. All conceivable direct mechanisms
based on or in equilibrium with the formation of Z(O)-CH3
are invalidated by the failure of framework-bound methoxy
groups to undergo H/D exchange at high temperatures. In
agreement with recent theoretical studies, our experimental
studies preclude both the carbene route (Scheme 2) and the
methoxonium ylide route (Scheme 3). The conclusion of both
theory and experiment is that the direct mechanisms fail, and
chance introduction of a primordial hydrocarbon pool is the
only mechanism for MTO initiation in agreement with these
findings. The present work in combination with referen-
ces [10] and [11] disproves the central claims of reference [13]
that were recently reiterated in reference [16].
[1] M. Stöcker, Microporous Mesoporous Mater. 1999, 29, 3 – 4 8.
[2] J. F. Haw, W. Song, D. M. Marcus, J. B. Nicholas, Acc. Chem. Res.
2003, 36, 317 – 326.
[3] I. M. Dahl, S. Kolboe, J. Catal. 1994, 149, 458 – 464.
[4] B. Arstad, S. Kolboe, J. Am. Chem. Soc. 2001, 123, 8137 – 8138.
[5] P. W. Goguen, T. Xu, D. H. Barich, T. W. Skloss, W. Song, Z.
Wang, J. B. Nicholas, J. F. Haw, J. Am. Chem. Soc. 1998, 120,
2650 – 2651.
[6] W. Song, J. F. Haw, J. B. Nicholas, C. S. Heneghan, J. Am. Chem.
Soc. 2000, 122, 10726 – 10727.
[7] J. F. Haw, J. B. Nicholas, W. Song, F. Deng, Z. Wang, C. S.
Heneghan, J. Am. Chem. Soc. 2000, 122, 4763 – 4775.
[8] B. Arstad, J. B. Nicholas, J. F. Haw, J. Am. Chem. Soc. 2004, 126,
2991 – 3001.
[9] W. Song, D. M. Marcus, H. Fu, J. O. Ehresmann, J. F. Haw, J. Am.
Chem. Soc. 2002, 124, 3844 – 3845.
[10] D. Lesthaeghe, V. Van Speybroeck, G. B. Marin, M. Waroquier,
Chem. Phys. Lett. 2005, 417, 309 – 315.
[11] D. Lesthaeghe, V. Van Speybroeck, G. B. Marin, M. Waroquier,
Angew. Chem. 2006, 118, 1746 – 1751; Angew. Chem. Int. Ed.
2006, 45, 1714– 1719.
[12] T. R. Forester, R. F. Howe, J. Am. Chem. Soc. 1987, 109, 5076 –
5082.
[13] W. Wang, A. Buchholz, M. Seiler, M. Hunger, J. Am. Chem. Soc.
2003, 125, 15260 – 15267.
[14] D. K. Murray, J.-W. Chang, J. F. Haw, J. Am. Chem. Soc. 1993,
115, 4732 – 4741.
Experimental Section
[17]
HSAPO-34was synthesized using standard procedures,
followed
by calcination at 873 K. [D3]DME (specifically CH3OCD3) was
Angew. Chem. Int. Ed. 2006, 45, 3133 –3136
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3135

Communications
[15] a) W. Kirmse, Carbene Chemistry, 2nd ed., Academic Press, New
York, 1971; b) Carbenes Vol. I and II, (Eds.: R. A. Moss, M.
Jones, Jr.), Wiley, New York, 1974 and 1975.
[16] Y. J. Jiang, W. Wang, V. R. R. Marthala, J. Huang, B. Sulikowski,
M. Hunger, J. Catal. 2006, 238, 21—27.
[17] HSAPO-34was prepared according to a patented procedure:
B. M. Lok, C. A. Messina, R. L. Patton, R. T. Gajek, T. R.
Cannan, E. M. Flanigen, US Patent 4,440,871, 1984. X-ray
diffraction studies showed a pure crystalline phase with the CHA
structure. The product was calcined in static air at 873 K for 10 h
to remove the template agent and pressed into 10–20 mesh
pellets without binder or other additives.
3136
www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 3133 –3136