Fishing for catalysts: Mechanism-based probes for active species in solution

Article abstract of DOI:10.1002/1522-2675(20000906)83:9<2192::AID-HLCA2192>3.0.CO;2-G

A combination of derivatization with charged substrates and electrospray-ionization mass spectrometry is used to fish out the active species in a catalytic reaction. The observed species in the mass spectrometer corresponds to the resting state of the catalyst. Data for the ring-opening metathesis polymerization (ROMP) of norbornene by (Cy₃P)₂Cl₂Ru=CHPh (Cy = cyclohexyl) are used to illustrate the method.

Full text of DOI:10.1002/1522-2675(20000906)83:9<2192::AID-HLCA2192>3.0.CO;2-G

2192
Helvetica Chimica Acta ± Vol. 83 (2000)
Fishing for Catalysts: Mechanism-Based Probes for Active Species in
Solution
by Christian Adlhart and Peter Chen*
Laboratorium für Organische Chemie, ETH-Zürich, Universitätstrasse 16, CH-8092 Zürich
Dedicated to Albert Eschenmoser on the occasion of his 75th birthday
A combination of derivatization with charged substrates and electrospray-ionization mass spectrometry is
used to fish out the active species in a catalytic reaction. The observed species in the mass spectrometer
corresponds to the resting state of the catalyst. Data for the ring-opening metathesis polymerization (ROMP) of
norbornene by (Cy3P)2Cl2RuˆCHPh (Cy ˆ cyclohexyl) are used to illustrate the method.
We report an electrospray-mass-spectrometric method for the identification of an
uncharged active species in a solution-phase catalytic reaction based on derivatization
of the active species with a charged substrate. The identification of the active species, as
opposed to the procatalyst and any other related complexes that may be simultaneously
present, is nearly always a challenging task. As had been elegantly demonstrated by
Halpern [1], the most abundant complex in solution does not need to be a part of the
catalytic cycle; it may represent, in fact, a dead end. Moreover, mechanistic
considerations can suggest several active species that may be present. A method by
which their a priori different levels of activity or selectivity could be quantitatively
assayed would be of great utility. The present report represents a new methodology
which can both identify and assay catalytically active species in solution by selectively
transferring them to the mass spectrometer after trapping with a ꢀlabeledꢁ substrate.
The demonstration is done with the well-characterized (Cy3P)2Cl2RuˆCHPh olefin-
metathesis catalyst [2] as a model system.
Because electrospray (for a complete monograph on the technique, see [3])
transfers charged species from solution to the gas phase, a reaction which derivatizes
active species with a charge becomes a selective probe. In this sense, the method is
related to other derivatization strategies that pull out a desired species by way of
radioactive or fluorescent tags, or pendant binding groups, as labels. In the present
example, one fishes for the active species using a tagged substrate molecule as the
fishhook.
The test reaction is olefin metathesis or ring-opening metathesis polymerization
(ROMP), catalyzed by the neutral complex (Cy3P)2Cl2RuˆCHPh (Cy ˆ cyclohexyl;
1), presumably via a monophosphine active complex (Scheme 1) [2][4]. Two different
fishhooks, both cationized norbornenes, were synthesized straightforwardly according
to Scheme 2.
ROMP Reactions were performed by adding a solution of 4.7 mg of norbornene in
0.5 ml of CH2Cl2 to a solution of 8.2 mg of 1 in 0.5 ml of CH2Cl2, stirring at 258 for

Helvetica Chimica Acta ± Vol. 83 (2000)
2193
Scheme 1
Scheme 2a)
a) LiAlH4 (0.27 equiv.)/Et2O, 1 h 08, 2 h reflux, extraction; 93%. b) Br2P (0.55 equiv.)/cat. pyridine/Et2O, 08,
30 min 258, ice and 10% NaHCO3, 08, extraction, flash chromatography (silica gel; pentane; Rf 0.7); 27%.
c) Ph3P (3 equiv.)/DMF, 24 h 1508, precipitation with Et2O; 28%. d) Ion-exchange with Amberlite IRA 400
(ClÀ form; MeOH); quant. e) BnNH2 (1 equiv.), no solvent, 2 h 258, distillation, 2008, 8 mbar; 87%. f) LiAlH4
(0.26equiv.)/Et 2O, 1 h 08, 2 h reflux, H2O (1.2 equiv.), 08, extraction; 92%. g) K2CO3 (1.5 equiv.)/MeOH, MeI
(6equiv.), 12 h reflux, recrystallization from MeOH; 25%. h) Ion-exchange with Amberlite IRA 400 (ClÀ form;
MeOH); quant.
a) All chemicals were obtained from Fluka and used without any further purification. Bicyclo[2.2.1]hept-2-ene-
2-carbaldehyde was used as racemic endo/exo-mixture.
15 min, and finally, addition of 0.1 equiv. (relative to ruthenium) of the cationized
norbornene. The solution was then diluted 100-fold with CH2Cl2, and analyzed by
electrospray-ionization mass spectrometry in a modified Finnigan MAT TSQ-700
spectrometer as described in [5][6].
The resulting mass spectra are shown in Figs. 1 and 2. As mentioned previously, only
charged species, either cationic or anionic, depending on whether the spectrometer is
run in positive-ion or negative-ion mode, are transferred to the gas phase. Accordingly,
the mass spectra correspond to the selective detection of the catalyst-bound, charge-
labeled substrate (the fishhook). The observed species should represent the resting
state of the complex involved in the catalytic cycle. While our previous work has shown
that a cationic catalyst can be observed by electrospray-ionization mass spectrometry
[5], the mass spectrum in Fig. 1 shows the detection of an electrically neutral catalyst by
trapping with cationized substrate. The difference is significant. For a cationic catalyst,

2194
Helvetica Chimica Acta ± Vol. 83 (2000)
Fig. 1. ROMP of 5 equiv. of norbornene with 1 and 2a as fishhook. The polymer distribution with n ˆ 0 starts at
m/z 911. The ESI-MS was performed on a Finnigan MAT TSQ-700 by electrospraying at 10À5 m solution of the
labelled living catalyst in CH2Cl2 in high-mass mode. Tube lens voltage was 170 V. The two insets show the
experimental and computed isotope distributions for one of the oligomer peaks, which indicate, along with the
mass, that the complex has one ruthenium, one phosphine, two chlorines, and one unit of the cation in 2a.
all complexes, whether active or not, are detected. For a neutral catalyst and cationized
substrate, only those complexes which bound and reacted with the substrate are detected.
For a polymerization reaction, control experiments can be done by addition of
further monomer1) to establish that the detected species is still a competent catalyst
after reaction with the charge-labeled fishhook. For the particular case of ROMP by 1,
the reaction should be reversible, with a ring-closing metathesis, leading to cyclic
oligomers, known as ꢀback-bitingꢁ products, and a truncated catalyst-bound oligomer
chain. Indeed, reversibility was demonstrated in our previous reports [5], indicating
that the ion in the mass spectrometer was a competent carbene-ruthenium complex.
The mass spectrum in Fig. 1 also confirms that, subsequent to initiation, the catalytic
cycle proceeds through the monophosphine complex without further significant
involvement of a bis-phosphine complex, i.e., the resting state of the catalyst in ROMP
1
)
ROMP of norbornene by 1 has been shown to be living polymerization in which block copolymers can be
prepared by sequential addition of different monomers.

Helvetica Chimica Acta ± Vol. 83 (2000)
2195
is the monophosphine complex as opposed to the bis-phosphine resting state in acyclic
olefin metathesis. In the earlier study of the related complex (Cy2PR)2Cl2RuˆCHPh
‡
(R ˆ CH2CH2NMe3 ), the doubly-charged molecular ion was observed by direct
electrospray of a solution of the dicationic complex [5]. Loss of one phosphine in the
electrospray mass spectrometer required collisional activation, even with the additional
coulombic repulsion of the two positively charged phosphines. Moreover, a control
experiment, in which a solution of 1 was treated with (hex-5-enyl)triphenylphospho-
nium hexafluorophosphate (an acyclic metathesis reaction) and then electrosprayed,
‡
showed a peak due to (Cy3P)2Cl2RuˆCH(CH2)4PPh3 . The absence of bis-phosphine in
Fig. 1 within the 100 :1 signal-to-noise of the spectrum is consistent with our earlier
conclusion, based on kinetic arguments, that the resting state of the catalyst in ROMP is
an intramolecular p-complex, in which an olefinic moiety on the catalyst-bound
oligomeric or polymeric chain is bound to the metal center. With our earlier estimate of
3 kcal/mol for the intramolecular binding energy in the case of the ROM product with
norbornene, a simple equilibrium calculation (assuming, of course, that the mass-
spectrometric peak intensities represent solution-phase concentrations, and the CH2Cl2
itself does not bind significantly) suggests an upper-bound for the phosphine binding
energy2) of 8.5 kcal/mol, which is consistent with expectations of phosphine binding for
the sterically demanding tri(cyclohexyl)phosphine ligand.
One possible pitfall in the method is seen in Fig. 2. Clearly visible are two series of
peaks, both displaying the characteristic m/z 94 spacing of catalyst-bound norbornene
oligomers. One is tempted to attribute the presence of two series as evidence for two
propagating species in the ROMP reactions of 1, probed by addition of 2b. However, as
seen in Fig. 1, only one propagating species is evident when 2a is used as a probe. By
matching both the mass and the distribution of isotopic peaks, the second series in Fig. 2
can be definitively attributed to a structure with the composition corresponding to
(Cy3P)2Ru2Cl5(ˆCHPh)2(norbornene)n(2)2, which is present when 2b, but not 2a, is
used as a fishhook. We propose that the second series is a Cl-bound dimer of two
monophosphine complexes, perhaps m-Cl-bridged, associated norbornene units, and (in
total) two cationized norbornenes. The coulombic interaction holding together the Cl-
bound dimer is presumably more favorable for the small ammonium cation in 2b
compared to that for the larger phosphonium cation in 2a. While dinuclear species [7]
have been suggested in olefin metathesis, the particular series including 2b is clearly an
artifact created by the trapping method.
The ROMP reaction of norbornene by (Cy3P)2Cl2RuˆCHPh (1) has been used as a
model for the demonstration of a mass-spectrometric method by which an active
catalytic complex may be identified in solution. Synthesis of cationized norbornenes
has yielded selective probes that tagged the ROMP-active catalyst-bound oligomers in
the solution-phase polymerization. Transfer of only charged species to the gas phase by
electrospray completed the analysis. The method has potential for application in a wide
range of catalytic reactions, in which a generalized derivatization according to the
catalyzed reaction itself can lead to selective detection of active species.
2
)
The calculation was performed assuming DG ˆ À3 kcal/mol for the formation of the intramolecular p-
complex, equilibrium at 298 K in solution, a total concentration of Ru species (after dilution) of 10À4 m, and
[bis-phosphine]/[monophosphine] < 0.01.

2196
Helvetica Chimica Acta ± Vol. 83 (2000)
Fig. 2. ROMP of 5 equiv. of norbornene with 2b as fishhook. In addition to the first series starting at m/z 784, a
second series of a Cl-bound dimer of two monophosphine complexes starting at m/z 1606 is seen. The two insets
show experimental and computed isotope distributions for one of the peaks in the second series, which indicate,
along with the mass, that the complex has two rutheniums, two phosphines, five chlorines, and two units of the
cation in 2b.
REFERENCES
[1] J. Halpern, Science 1982, 217, 401.
[2] P. Schwab, R. H. Grubbs, J. W. Ziller, J. Am. Chem. Soc. 1996, 118, 100; E. L. Dias, S. T. Nguyen, R. H.
Grubbs, J. Am. Chem. Soc. 1997, 119, 3887; M. Ulman, R. H. Grubbs, Organometallics 1998, 17, 2484.
[3] ꢀElectrospray Ionization Mass Spectrometryꢁ, Ed. R. D. Cole, John Wiley, New York, 1997.
[4] D. M. Lynn, B. Mohr, R. H. Grubbs, J. Am. Chem. Soc. 1998, 120, 1627.
[5] C. Hinderling, C. Adlhart, P. Chen., Angew. Chem. 1998, 110, 2831; C. Adlhart, C. Hinderling, H. Baumann,
P. Chen, J. Am. Chem. Soc., in press.
[6] C. Hinderling, D. A. Plattner, P. Chen, Angew. Chem. 1997, 109, 272; C. Hinderling, D. Feichtinger, D. A.
Plattner, P. Chen, J. Am. Chem. Soc. 1997, 119, 10793; D. Feichtinger, D. A. Plattner, Angew. Chem. 1997,
109, 1796; D. Feichtinger, D. A. Plattner, P. Chen, J. Am. Chem. Soc. 1998, 120, 7175; Y. M. Kim, P. Chen, Int.
J. Mass Spectrom. 1999, 185-7, 871.
[7] E. L. Dias, R. H. Grubbs, Organometallics 1998, 17, 2758; T. Weskamp, F. L. Kohl, W. Hieringer, D. Gleich,
W. A. Herrmann, Angew. Chem. 1999, 111, 2573.
Received June 5, 2000