On the compatibility of Ruthenium metathesis catalysts with secondary phosphines

Article abstract of DOI:10.1021/om401021k

While attempts to incorporate secondary phosphine ligands into the first-generation Grubbs catalyst RuCl₂(PR₃) ₂(=CHPh) (GI, R = Cy) were frustrated by decomposition (for HPCy ₂) or bulk (for HP^tBu

Full text of DOI:10.1021/om401021k

Note
pubs.acs.org/Organometallics
On the Compatibility of Ruthenium Metathesis Catalysts with
Secondary Phosphines
Bianca J. van Lierop and Deryn E. Fogg*
Center for Catalysis Research & Innovation and Department of Chemistry, University of Ottawa, Ottawa, Ontario, Canada, K1N 6N5
*
S Supporting Information
ABSTRACT: While attempts to incorporate secondary phosphine ligands into the
first-generation Grubbs catalyst RuCl (PR ) (CHPh) (GI, R = Cy) were
2
3 2
t
frustrated by decomposition (for HPCy ) or bulk (for HP Bu ), the mixed-
2
2
t
phosphine complex cis-RuCl (HP Bu )(PPh )(CHPh) (Ru-2b) was accessible
2
2
3
in good yields from GI′ (R = Ph). Complex Ru-2b exhibited modest RCM activity.
Ligand exchange with pyridine revealed that this is due to preferential loss of the
secondary phosphine, rather than PPh , from Ru-2b.
3
econdary phosphines are intriguing as prospective ligands
in Ru-catalyzed olefin metathesis. Of particular interest is
their capacity for H-bonding interactions (Chart 1), which
Hz). While the magnitude of the 2JPP coupling constant i8s
1
S
consistent with trans-disposition of the two phosphine ligands,
some distortion of the usual square pyramidal geometry is
implied by the doublet multiplicity of the benzylidene proton
3
Chart 1. Secondary Phosphines Explored and Potential H-
Bonding Interactions
(20.16 ppm; d, J = 11 Hz), which indicates coupling to only
HP
one 31P nucleus. Attempts at more extensive characterization
were hampered by the instability of Ru-1a in solution. Even at
1
0 min, ca. 10% decomposition was evident by integration
against an internal standard, and no alkylidene signals remained
after 3 h in solution.
Decomposition may involve C−H activation (as established
9,10
for related Ru-PCy complexes)
or nucleophilic attack on
n
the benzylidene functionality. Hong and Grubbs demonstrated
that PCy can abstract the unprotected methylidene carbon in
3
11
RuCl (H IMes)(PCy )(CH ), and the reduced bulk at
2
2
3
2
could potentially improve the longevity of alkylidene2
phosphorus for HPCy2 may be sufficient to enable the
secondary phosphine to attack at the benzylidene moiety,
particularly in the four-coordinate species that mediates
intermediates in catalysis, including enoic alkylidenes
generated from acrylates (now key reagents in sustainable
3
−6
metathesis).
Secondary phosphines have gone largely
phosphine exchange.
Reaction of HP Bu with GI under the conditions of Scheme
unexplored in metathesis, however. We are aware of just one
t
t
report, a recent Exxon patent describing RuCl (HP Bu ) (
2
2
2 2
7
1
did not show such facile decomposition, but the equilibrium
CHR) (where R = CHCMe ). With the intention of
2
yield of Ru-1b was limited to ca. 30%. On the presumption that
a steric clash between the PCy and HP Bu groups might be
responsible, we turned to the PPh
alternative precursor (Scheme 2). Accordingly, we modified
clarifying the opportunities and limitations associated with
t
HPR ligands in this context, we explored the synthesis and
3
2
2
reactivity of their Ru-benzylidene derivatives.
3
complex GI′ as an
Addition of excess HPCy (Scheme 1) to the first-generation
2
Grubbs catalyst GI in C D resulted in complete reaction
within 10 min at RT. Consistent with formation of mixed-
6
6
a
Scheme 2. Reaction of GI′ with Secondary Phosphines
phosphine derivative Ru-1a, a pair of doublets was evident in
the 31P{ H} NMR spectrum (42.3 and 27.5 ppm; J = 297
1
2
PP
Scheme 1. Reaction of GI with Secondary Phosphines;
Decomposition Ensues for Ru-1a
a
Complex Ru-2a (R = Cy), probably a cis/trans mixture (see text),
decomposes rapidly.
Received: October 20, 2013
©
XXXX American Chemical Society
A
dx.doi.org/10.1021/om401021k | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Note
Grubbs’ classic one-pot route to GI (in which RuCl (PPh ) is
treated successively with phenyldiazomethane and PCy3),
Under these conditions, at a catalyst loading of 5 mol %,
2
3 3
12
RCM to afford the 16-membered lactone 2 proceeded with
15
instead adding HPR in the second step. A smaller, 2-fold
efficiency comparable to that found using GI. RCM yields
2
excess of phosphine was employed, given the anticipated ease
of displacement of PPh , vs PCy .
were limited to 90%, however, the balance being oligomeric
products. Unlike second-generation Ru catalysts bearing
3
3
t
16,17
Phosphine exchange with HP Bu was slow at RT, but
electron-rich N-heterocyclic carbene ligands,
Ru-2b is
2
proceeded to completion within 2 h in refluxing CH Cl . Pale
insufficiently reactive to attack the internal olefinic units, as
required to recycle the oligomers into 2. Attempted cross-
metathesis of anethole with methyl acrylate also failed.
2
2
t
green RuCl (HP Bu )(PPh )(CHPh) (Ru-2b) was isolated
2
2
3
in 74% yield, following reprecipitations and treatment with
1
3
Amberlyst-15 resin to remove excess phosphine. NMR and
microanalytical data are consistent with formulation as the
mixed-phosphine product Ru-2b, despite the nominal lability of
The low activity of Ru-2b has an additional, unexpected
dimension, which was revealed in attempts to generate a more
18
labile pyridine derivative. On dissolving Ru-2b in pyridine-d5,
we observed formation of two major alkylidene products, which
we assign as the desired mono- and bis-pyridine complexes.
More significantly, the characteristic upfield singlet for free
31
the PPh ligand. Of note is the 23 Hz coupling of the two
P
3
nuclei, as compared to the nearly 300 Hz coupling seen for Ru-
a. This feature indicates the surprising cis-disposition of the
1
t
31
t
t
HP Bu emerged in the P NMR spectrum, accompanied by
PPh and HP Bu groups, as also reported for the two HP Bu
2
3
2
2
7
only trace amounts of PPh . Evidently, the secondary
ligands in the Exxon work. Other distinctions relative to Ru-1a
include the upfield location of the benzylidene proton (16.02
ppm) and its appearance as a doublet of doublets due to
3
t
alkylphosphine HP Bu2 is more labile than the tertiary
arylphosphine PPh , its steric bulk perhaps contributing to its
3
3
1
3
ease of displacement. The modest metathesis activity described
splitting by both P nuclei ( J = 22 and 12 Hz).
HP
12
above is consistent with the weak donor ability of the PPh3
ligand retained in the active species. Catalyst decomposition via
Attempts to install HPCy onto GI′ resulted in even faster
2
decomposition than seen with GI. In an NMR experiment
using isolated GI′, nearly 80% loss of the alkylidene signal was
evident within 10 min. The balance consisted of two distinct
alkylidene multiplets centered at 17.5 and 19.4 ppm (see SI)
While attempts to identify the products were frustrated by
decomposition over 1 h, we tentatively assign the upfield signal
to cis-Ru-2a by analogy with the data for Ru-2b, and the
downfield signal to its trans isomer.
attack of the liberated HPR ligand on the methylidene moiety
2
may also contribute. Hong and Grubbs documented such
behavior for the much bulkier PCy ligand released from
3
11
RuCl (H IMes)(PCy )(CH ), as noted above.
2
2
3
2
The foregoing describes two limitations associated with
installation of secondary phosphines within Grubbs-class
t
metathesis catalysts. Use of HP Bu gives access to an isolable
2
t
RuCl (HP Bu )(PPh )(CHPh) product, but the secondary
2
2
3
With Ru-2b in hand, we turned to assays of its metathesis
performance. Probe reactions indicated that ring-closing
metathesis (RCM) of the benchmark substrate diethyl
diallylmalonate was recalcitrant at RT, with no sign of reaction
over 14 h using 1 mol % Ru-2b. The characteristic benzylidene
signals remained, however, indicating that the issue is low
reactivity, rather than catalyst instability. RCM to yield 1
proceeded at 40 °C, but the level of activity was relatively low,
ca. 10 h being required for complete RCM (Chart 2a). We
attribute this in part to the cis-dichloride geometry, known for
phosphine proves more labile than PPh . The active catalyst
3
thus corresponds to that generated from GI′ itself. While
t
t
bis(HP Bu ) analogues would enable access to RuCl (HP Bu )-
2
2
2
(
CHR) intermediates, the combination of low donicity and
lability is likely to limit catalyst productivity. Dicyclohexylphos-
phine, on the other hand, does not give access to stable
products. Decomposition of the HPCy derivatives may involve
2
C−H activation of the cyclohexyl ring or nucleophilic attack of
phosphine at RuCHPh. Studies examining the susceptibility
of the benzylidene and methylidene moieties to attack by these
and other nucleophiles are under way and will be reported in
due course.
14
nearly a decade to inhibit initiation of Ru metathesis catalysts.
a
Chart 2. RCM Activity of Ru-2b (40 °C, CH Cl )
2
2
EXPERIMENTAL SECTION
General Procedures. Reactions were carried out using standard
Schlenk or glovebox techniques at room temperature (RT; 21−23
C), unless otherwise indicated. NMR solvents were stored under N
■
°
2
over 4 Å molecular sieves for at least 24 h before use. Literature
1
9
20
procedures were used to prepare RuCl (PPh ) , PhCHN ,
2
3
3
2
1
2
12
RuCl (PCy ) (CHPh) (GI), RuCl (PPh ) (CHPh), and the
2
3
2
2
3 2
2
1
diene precursor to 2. Diethyl diallylmalonate (Aldrich) was freeze−
thaw degassed under N before use; secondary phosphines (Cytec)
2
were used as received. NMR spectra were recorded at 23 °C and
referenced to the residual proton or carbon signals of the deuterated
1
13
31
solvent ( H, C) or external 85% H PO ( P). Signals are reported
3
4
1
13
31
relative to TMS ( H and C) or 85% H PO ( P) at 0 ppm.
3
4
GC-FID analysis was carried out using a polysiloxane column (30 m
320 μm) with UHP-grade He as the carrier gas; column pressure
1.512 psi. The FID response was maintained between 50 and 2000
×
1
ρA. Retention times were confirmed by comparison with authenticated
samples (including for E/Z isomers). Calibration curves of peak areas
vs concentration were constructed in the relevant concentration
regime, to account for the dependence on detector response for
substrates, products, and the internal standards (IS) decane and
tetrahydronaphthalene (THN). Yields in catalytic runs were
a
Conditions: (a) 100 mM 1, 1 mol % Ru; (b) 5 mM 2, 5 mol % Ru.
B
dx.doi.org/10.1021/om401021k | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Note
determined from the integrated peak areas, referenced against IS, and
compared to the starting substrate:IS ratio.
Attempted CM Experiment. To a solution of anethole (148 μL,
0.495 mmol) and decane (104 μL, 1 equiv) in 1,2-dichloroethane
(DCE; 2.1 mL) were added methyl acrylate (180 μL, 4 equiv) and 84
μL (0.5 mol %) of a DCE solution of Ru-2b (10 mg, 15 mmol, 30
mM). No conversion was seen after heating at 70 °C in a glovebox for
1 h or after a further 11 h on a Schlenk line.
Attempted Synthesis of RuCl (HPCy )(PCy )(CHPh), Ru-1a.
2
2
3
In a representative procedure, a J. Young NMR tube was charged with
GI (20 mg, 24 μmol), C D (0.75 mL), and, as internal standards for
6
6
NMR analysis, Ph PO (6.8 mg, 24 μmol) and TMB (2,3,5-
3
trimethoxybenzene; 0.4 mg, 2 μmol). NMR spectra were acquired
to establish initial integration ratios, HPCy was then added (55 μL, 49
μmol, 2 equiv; glovebox), and the reaction was monitored in the NMR
probe. Mass balance at 10 min: 4% GI, 84% Ru-1a, 12% missing. The
ASSOCIATED CONTENT
Supporting Information
NMR spectra for Ru-2. This material is available free of charge
via the Internet at http://pubs.acs.org.
2
■
*
S
3
1
1
yield of Ru-1a dropped to 62% by 1 h. P{ H} NMR (C D , 75
6
6
2
2
MHz): δ 42.3 (d, J = 297 Hz, HPCy ), 27.5 (d, J = 297 Hz,
PCy ). H NMR (C D , 300 MHz): δ 20.16 (d, J = 10.9 Hz, 1H,
PP
2
PP
1
3
3
6
6
HP
3
3
AUTHOR INFORMATION
Corresponding Author
*E-mail: dfogg@uottawa.ca.
Notes
RuCHPh), 8.44 (d, J = 7.2 Hz, 2H, CHPh, H ), 7.28 (t, J
=
■
HH
o
HH
3
7
(
.2 Hz, 1H, CHPh, H ), 7.13 (t, J = 7.5 Hz, 2H, CHPh, H ), 4.80
p
HH
m
1
3
dd, J = 333 Hz, J = 2.9 Hz, 1H, HPCy ), 3.03−0.89 (m, 55H,
Cy). Use of 10 equiv of HPCy resulted in 100% conversion of GI
HP
HP
2
2
within 10 min, with 9% decomposition. No alkylidene signal was
observed after 3 h.
The authors declare no competing financial interest.
t
Synthesis of RuCl (HP Bu )(PPh )(CHPh), Ru-2b. A red-
2
2
3
ACKNOWLEDGMENTS
This work was funded by NSERC of Canada.
orange solution of PhCHN (246 mg, 2.09 mmol) in 2.8 mL of
■
2
hexanes was added dropwise by cannula to a cold (−78 °C), stirred
solution of RuCl (PPh ) (1.00 g, 1.04 mmol) in 29 mL of dry
2
3 3
CH Cl . After 20 min, the brown-green solution was warmed to 0 °C,
REFERENCES
2
2
■
stirred for a further 30 min to complete formation of GI′, and treated
(
1) Secondary dialkylphosphines, though relatively rare as ligands in
t
with HP Bu (424 μL, 2.29 mmol, 2.2 equiv) in 7 mL of dry hexanes.
2
catalysis, have been reported to function as highly efficient supporting
ligands in Pd-catalyzed Heck reactions. See: (a) Schnyder, A.;
Aemmer, T.; Indolese, A. F.; Pittelkow, U.; Studer, M. Adv. Synth.
Catal. 2002, 344, 495−498 and references therein. For Rh and Pd
complexes of ferrocenyl derivatives containing a secondary phosphine,
see: (b) Buergler, J. F.; Togni, A. Organometallics 2011, 30, 4742−
The reaction was heated at reflux for 2 h, then stripped of solvent.
Reprecipitation (CH Cl −pentane) afforded a gray-green solid, from
2
2
which residual phosphine was removed by stirring for 2 h with
Amberlyst-15 resin (798 mg, 3.75 mmol) in 10 mL of CH Cl . The
2
2
resin was then filtered off and washed with CH Cl (3 × 1 mL), and
2
2
the combined washings were concentrated to a minimum and treated
with hexanes to precipitate pale green Ru-2b. Yield: 517 mg (74%).
4
750. More commonly used are secondary phosphine oxides, which
prove versatile ligands in cross-coupling catalysis. For a recent review
of the latter topic, see: (c) Shaikh, T. M.; Weng, C.-M.; Hong, F.-E.
Coord. Chem. Rev. 2012, 256, 771−803. Secondary phosphine
chlorides were also recently reported to promote cross-coupling of
unactivated alkyl chlorides. See: (d) Ackermann, L.; Kapdi, A. R.;
Schulzke, C. Org. Lett. 2010, 12, 2298−2301.
3
1
1
2
t
P{ H} NMR (C D , 202 MHz): δ 75.3 (d, J = 24 Hz, HP Bu ),
6
6
PP
2
2
1
4
(
2
1
1
6.1 (d, J = 24 Hz, PPh ). H NMR (CDCl , 300 MHz): δ 16.02
PP 3 3
3
3
dd, J = 22 Hz, 12 Hz, 1H, RuCHPh), 8.46 (d, J = 7.6 Hz,
HP
HH
1
3
H, CHPh, H ), 7.8−7.0 (m, 18H, Ph), 3.75 (dd, J = 342 Hz, J
=
=
o
HP
HP
t
3
3
0 Hz, 1H, HP Bu ), 1.39 (d, J = 14 Hz, 9H, CH ), 0.71 (d, J
2
HP
3
HP
13
1
4 Hz, 9H, CH3). C{ H} NMR CDCl , 125.8 MHz): δ 306.7
3
(2) Choi, T.-L.; Lee, C. W.; Chatterjee, A. K.; Grubbs, R. H. J. Am.
Chem. Soc. 2001, 123, 10417−10418.
(
1
located via HMQC, RuCHPh), 149.2, 134.8, 132.7, 131.6, 130.6,
1
1
29.1, 128.3, 128.2, 36.4 (d, J = 22 Hz, CMe ), 35.8 (d, J = 22
(3) Miao, X.; Malacea, R.; Fischmeister, C.; Bruneau, C.; Dixneuf, P.
H. Green Chem. 2011, 13, 2911−2919.
PC
3
PC
2
2
Hz, CMe ), 32.0 (d, J = 2 Hz, CH ), 30.2 (d, J = 2 Hz, CH3).
3
PC
3
PC
Anal. Calcd. for C H Cl P Ru: C, 59.10; H, 6.02. Found: C, 58.92;
(4) Biermann, U.; Meier, M. A. R.; Butte, W.; Metzger, J. O. Eur. J.
Lipid Sci. Technol. 2011, 113, 39−45.
33
40
2 2
H, 6.19.
Attempted Synthesis of RuCl (HPCy )(PPh )(CHPh), Ru-
2
2
3
(5) Schmidt, B.; Geissler, D. ChemCatChem 2010, 2, 423−429.
(6) Lummiss, J. A. M.; Oliveira, K. C.; Pranckevicius, A.; Santos, A.;
dos Santos, E. N.; Fogg, D. E. J. Am. Chem. Soc. 2012, 134, 18889−
18891.
1
b. The corresponding reaction with HPCy failed. In an NMR-scale
2
experiment with trimethoxybenzene (TMB) as internal standard (see
Figure S1), HPCy (6.1 μL, 28 μmol) was added to a solution of G1′
2
(
10 mg, 13 μmol), TMB (0.64 mg, 3.8 μmol), and Ph PO (3.5 mg, 13
3
(7) Holtcamp, M. W.; Bedoya, M. S., P. U.S. Patent 8,524,930 B2,
2013.
μmol) in 0.75 mL of C D . After 10 min, no signal for GI′ remained.
6
6
Integration against TMB indicated 78% loss of total alkylidene.
Alkylidene multiplets at ca. 17.5 and 19.4 ppm were evident, but
elucidation of the products was frustrated by complete decomposition
after 1 h.
(
8) The Exxon patent described cis-disposition of the phosphine
t
ligands in RuCl (HP Bu ) (CHCHCMe ). See ref 7.
2
2
2
2
(9) Christ, M. L.; Sabo-Etienne, S.; Chaudret, B. Organometallics
1995, 14, 1082−1084.
Ligand Exchange of Ru-2b with Pyridine. A J-Young NMR
(10) Amoroso, D.; Yap, G. P. A.; Fogg, D. E. Can. J. Chem. 2001, 79,
958−963.
tube was charged with Ru-2b (7 mg, 10 μmol) and 0.75 mL of
1
31
pyridine-d to give a dark green solution. H and P NMR signals for
(11) Hong, S. H.; Wenzel, A. G.; Salguero, T. T.; Day, M. W.;
Grubbs, R. H. J. Am. Chem. Soc. 2007, 129, 7961−7968.
(12) Schwab, P.; Grubbs, R. H.; Ziller, J. W. J. Am. Chem. Soc. 1996,
118, 100−110.
5
t
free HP Bu emerged on standing at RT. After 6 h, ca. 5% Ru-2b
2
remained.
RCM Experiments. In a representative procedure, a 25 mL
Schlenk tube was loaded with diethyl diallylmalonate 1 (240 μL, 1.00
mmol), THN (139 μL, 1.00 mmol), and CH Cl (9.6 mL) to give a
(13) Monsaert, S. F.; Verpoort, F. W. C. WO Patent App. WO 2011/
091980 Al; CAPlus 383210.
2
2
final substrate concentration of 100 mM. An aliquot was analyzed
GC-FID) to establish initial integration ratios. Catalyst Ru-2b (6.7
mg, 0.01 mmol) was then added, and the stirred reaction was heated at
0 °C. Complete conversion required ca. 10 h: see Chart 2a.
Macrocyclization (Chart 2b) was carried out similarly using a 5 mM
diene solution and 5 mol % Ru-2b. The 16-membered lactone 2 was
obtained as a 2.7:1 mixture of E/Z isomers.
(14) For selected early reports demonstrating lower activity for
ruthenium metathesis catalysts containing cis-dichloride ligands,
including cases where initiation efficiency has been shown to be the
culprit, see: (a) Slugovc, C.; Perner, B.; Stelzer, F.; Mereiter, K.
Organometallics 2004, 23, 3622−3626. (b) Ung, T.; Hejl, A.; Grubbs,
R. H.; Schrodi, Y. Organometallics 2004, 23, 5399−5401. (c) Benitez,
D.; Goddard, W. A., III. J. Am. Chem. Soc. 2005, 127, 12218−12219.
(
4
C
dx.doi.org/10.1021/om401021k | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Note
(
d) Barbasiewicz, M.; Szadkowska, A.; Bujok, R.; Grela, K. Organo-
metallics 2006, 25, 3599−3604. (e) Ben-Asuly, A.; Tzur, E.;
Diesendruck, C. E.; Sigalov, M.; Goldberg, I.; Lemcoff, N. G.
Organometallics 2008, 27, 811−813.
(
15) Jafarpour, L.; Heck, M.-P.; Baylon, C.; Lee, H. M.; Mioskowski,
C.; Nolan, S. P. Organometallics 2002, 21, 671−679.
16) Conrad, J. C.; Eelman, M. D.; Duarte Silva, J. A.; Monfette, S.;
Parnas, H. H.; Snelgrove, J. L.; Fogg, D. E. J. Am. Chem. Soc. 2007, 129,
(
1
(
(
024−1025.
17) Monfette, S.; Fogg, D. E. Chem. Rev. 2009, 109, 3783−3816.
18) Sanford, M. S.; Love, J. A.; Grubbs, R. H. Organometallics 2001,
2
(
1
(
0, 5314−5318.
19) Hallman, P. S.; Stephenson, T. A.; Wilkinson, G. Inorg. Synth.
970, 12, 237−40.
20) Creary, X. Org. Synth. 1986, 64, 207−216.
21) Furstner, A.; Langemann, K. Synthesis 1997, 792−803.
(
̈
D
dx.doi.org/10.1021/om401021k | Organometallics XXXX, XXX, XXX−XXX