Concerted C-N and C-H bond formation in a magnesium-catalyzed hydroamination

Article abstract of DOI:10.1021/ja108881s

Coordinatively saturated To^MMgMe (1; To^M = tris(4,4-dimethyl-2-oxazolinyl)phenylborate) is an active precatalyst for intramolecular hydroamination/cyclization at 50 °C. The empirical rate law of -d[substrate]/dt = k′_obs[Mg

Full text of DOI:10.1021/ja108881s

Published on Web 11/30/2010
Concerted C-N and C-H Bond Formation in a Magnesium-Catalyzed
Hydroamination
James F. Dunne, D. Bruce Fulton, Arkady Ellern, and Aaron D. Sadow*
Department of Chemistry and U.S. Department of Energy Ames Laboratory, Iowa State UniVersity,
Ames, Iowa 50011, United States
Received October 1, 2010; E-mail: sadow@iastate.edu
Scheme 1. General Proposed Insertion Mechanism for
Intramolecular Hydroamination
Abstract: Coordinatively saturated ToMMgMe (1; ToM ) tris(4,4-
dimethyl-2-oxazolinyl)phenylborate) is an active precatalyst for in-
tramolecular hydroamination/cyclization at 50 °C. The empirical rate
law of -d[substrate]/dt ) k′obs[Mg]1[substrate]1 and Michaelis-Menten-
type kinetics are consistent with a mechanism involving reversible
catalyst-substrate association prior to cyclization. The resting state
of the catalyst, ToMMgNHCH2CR2CH2CHdCH2 [R ) Ph, Me,
s(CH2)5s], is isolable, but isolated magnesium amidoalkene does
not undergo unimolecular cyclization at 50 °C. However, addition of
trace amounts of substrate allows cyclization to occur. Therefore,
we propose a two-substrate, six-center transition state involving
concerted CsN bond formation and NsH bond cleavage as the
turnover-limiting step of the catalytic cycle.
dimethyl-2-oxazolinyl)phenylborato [ToM] magnesium(II) com-
pounds were chosen to provide the desired catalytic activity and
steric properties, based on recent reports of the first examples of
magnesium(II)- and calcium(II)-catalyzed intramolecular hydroami-
nation11 as well as extensive investigations of tris(pyrazolyl)borate
magnesium compounds.12 We have found that four-coordinate
magnesium alkyl compounds are precatalysts for hydroamination/
cyclization. Surprisingly, isolable magnesium amidoalkene com-
plexes are not kinetically competent for intramolecular cyclization
in the absence of additional aminoalkene. From this observation
and the catalytic rate law, we propose an alternative pathway that
invokes concerted C-N and C-H bond formation through a six-
centered transition state.
Migratory insertion of an olefin into a metal-nitrogen bond is
exemplified by d0 and fnd0 metal complex-catalyzed hydroamination/
cyclization of aminoalkenes.1 However, kinetic studies of these
catalytic hydroaminations only provide indirect access to the
proposed insertion step. The typical empirical rate law, -d[sub-
strate]/dt ) kobs[catalyst]1[substrate]0, is interpreted as proceeding
through turnover-limiting intramolecular insertion (Scheme 1).2 In
most cases this simplifies the overall mechanism because kobs is a
combination of elementary rate and equilibrum constants, and the
observed rate constant often includes competitive association and/
or substitution by product and substrate (i.e., inhibition).3 This amine
substrate coordination may be mechanistically significant, as at least
one additional amine is typically present in the catalyst resting
state.2,3 Additionally, kH/kD values ranging from 2 to 5 for H2NR
vs D2NR substrates suggest that N-H bond cleavage is also
involved in the rate-determining step and that C-N bond formation
is more complicated than classical 1,2-insertion.3,4
ToMMgMe (1) is prepared from HToM and MgMe2(O2C4H8)2.13
A crosspeak between the oxazoline nitrogen and the Mg-CH3 in
1
a H-15N HMBC experiment demonstrates that 1 is the desired
heteroleptic rather than a mixture of ToM2Mg (2) and MgMe2.
Neither THF nor dioxane coordinates to the magnesium(II) center
1
in 1, and no change in the H NMR spectrum is detected upon
Furthermore, 1,2-insertion reactions between alkenes and
[M]-NR2 bonds that afford isolable metal alkyl compounds are
rare. Group 4 examples are limited to alkyne insertions that provide
vinyl amine products,5 as does the sole group 6 example.6 Activated
olefins, such as norbornylene and acrylonitrile, react with amido-
iridium(III) and amido-platinum(II) compounds, respectively, and
these steps are useful in intermolecular hydroaminations.7,8 Re-
cently, palladium amido compounds were shown to react through
intramolecular and intermolecular olefin insertion pathways.9 We
are not aware of stoichiometric insertion reactions of d0 or fnd0
metal amides and alkenes that yield isolable metal alkyls. Despite
the limited stoichiometric examples, a wealth of data and compu-
tational models have supported the general insertion mechanism
for rare earth metal-catalyzed hydroamination (Scheme 1).10
To study the insertion step in hydroamination/cyclization, we
decided to investigate catalysts in which bonding of an additional
substrate might be limited by coordinative saturation. Tris(4,4-
thermolysis in benzene-d6 at 150 °C in a sealed NMR tube. An
X-ray structure of 1 (Figure 1) and solid angle determination show
that 75% of the space surrounding the Mg(II) center is occupied
by the ToM (60%) and CH3 (15%) ligands.14
Compound 1 is a catalyst for the cyclization of primary
4-aminoalkenes (3-5) and a secondary aminoalkene (6) affording
pyrrolidines (7-10) at 50 °C in benzene (Table 1). 2,2-Disubsti-
tution is required for cyclization to occur, and the unsubstituted
4-pentene-1-amine is not cyclized by 1 at 50 °C.
The lack of coordination of ethers to 1 suggests that there is no
open site for olefin coordination prior to insertion. As this is
uncommon for hydroamination catalysts, we were motivated to
further investigate the cyclization mechanism. The concentration
of substrate 3 was monitored over the course of the reaction, and
plots of ln[3] vs time are linear for >3 half-lives. A plot of kobs vs
[1] is linear giving the empirical rate law: -d[3]/dt ) k′obs[1]1[3]1
(k′(H) ) 6.9 × 10-3 M-1 s-1). Second-order rate laws for
obs
9
17680 J. AM. CHEM. SOC. 2010, 132, 17680–17683
10.1021/ja108881s  2010 American Chemical Society

C O M M U N I C A T I O N S
Figure 2. Plot showing relationship between initial rate -d[3]/dt and [3]ini
(Conditions: catalyst [11] ) 0.00493 M, C6D6, 60 °C). Each -d[3]/dt value
is obtained from a linear-least-squares fit of [3] vs time for 5% conversion.
The curve represents a nonlinear least-squares fit of the data to eq 1.
Figure 1. ORTEP diagram of ToMMgMe (1). Selected bond lengths (Å):
Mg1-N1, 2.122(1); Mg1-N2, 2.124(1); Mg1-N3, 2.090(1); Mg1-C16,
2.108(1).
Table 1. Hydroamination/Cyclization of Aminoalkenes Catalyzed
by ToMMgMe (1)a
catalyst + substrate yk1z catalyst · substrate
(2)
(3)
k-1
k2
catalyst · substrate
98 catalyst + product
In contrast, turnover-limiting metal-alkyl protonolysis would be
expected to show linear dependence on [substrate] over all
concentrations.
In order to better establish the nature of the cyclization step, the
catalyst, and the catalyst ·substrate intermediate, we investigated
the identities of the species in the reaction mixture. Inspection of
1H NMR spectra of catalytic reactions of a 10:1 ratio of subtrate 3
to precatalyst 1 or 11 in benzene-d6 reveals two ToMMg species.
The first compound, 12, was independently prepared in a stoichio-
metric reaction of 3 with the magnesium alkyls 1 or 11 (eq 4).
Related magnesium amidoalkenes 13 and 14 are also formed upon
treatment of 1 with the aminoalkenes 4 and 5, and compounds
12-14 are isolated and fully characterized. Pseudo-C3V symmetry
a Conditions: 10 mol % 1, C6D6, 50 °C. b Conversion (%) was
determined by H NMR spectroscopy. c 20 mol % 1.
1
aminoalkene cyclization are uncommon, but a few examples have
been reported.15 A plot of k(D)obs for cyclization of 3-d2 vs [1] gives
k′(D) ) 1.5 × 10-3 M-1 s-1. The ratio k′(H)obs/k′(D) is 4.6,
1
of 12-14, as indicated by H and 13C{1H} NMR spectra and IR
obs
obs
spectra (one νCN band), support the structure of eq 4 (see Supporting
Information (SI)).
revealing a significant primary isotope effect. This rate law and
isotope effect data are consistent with several possible mechanisms,
including turnover limiting intermolecular protonolysis of a mag-
nesium alkyl and a two-step sequence in which reversible
catalyst-substrate association precedes unimolecular cyclization.
If the reaction follows the latter pathway, the rate could show
saturation at high aminoalkene concentrations.16 Indeed, νini (νini
) -d[3]/dt at 5% conversion) vs [3]ini (Figure 2) increases until
saturation is observed.
The catalyst ToMMgNHCH2CPh2CH2CHdCH2 (12) is preformed
by reaction of ToMMgCH2C6Me2H3 (11) and 3 (see eq 4 below)
prior to catalytic cyclization, and its concentration equals the
measured concentration of C6Me3H3. A nonlinear regression analysis
correlates the data to eq 1.17
Previously, Cp′2LaNHR′(NH2R′) (R′ ) CH2CMe2CH2CHCH2) was
characterized at low temperature but could not be isolated as
cyclization occurs at ∼ -20 °C.3a Pyrrolide ToMMg(NC4H5-2-Me-
4,4-Ph2) (15) is the other ToMMg-containing species present in the
catalytic reaction mixture and is prepared by reaction of 1 and
diphenyl-pyrrolidine 7 (eq 5).
k2[12][3]ini
K' + [3]ini
k2[substrate][catalyst]
K' + [substrate]
-d[3]
dt
)
)
(1)
This rate law corresponds to the mechanistic sequence of eqs 2-3,
where k2 ) (1.4 ( 0.2) × 10-2 s-1 (cyclization) and K′ ) (k-1
k2)/k1 ) 0.24 ( 0.05 M (formation constant):
+
9
J. AM. CHEM. SOC. VOL. 132, NO. 50, 2010 17681

C O M M U N I C A T I O N S
Magnesium amidoalkene 12 was observed in the catalytic
reaction mixture, so its reactivity was tested by warming a benzene-
d6 solution of isolated 12 to 50 °C, and then 100 °C for 12 h (eq
6). Surprisingly, the amidoalkene was the only species observed
under these conditions, and neither free pyrrolidine 6 nor magne-
sium pyrrolide 15 was formed.
benzene-d6) were measured using selective population inversion
(SPI).19 From this analysis, the rate of dissociative exchange (kdiss
)
at 60 °C is 0.20 s-1, which is ∼14× faster than k2 (0.014 s-1).
Thus, the frequency of oxazoline dissociation is greater than the
TOF for aminoalkene cyclization.
The oxazoline dissociation rate constants kdiss are likely similar
in 2 and 12-14.20 Thus, an amine ligand is required for cyclization
but not for oxazoline dissociation, suggesting that aminoalkenes
3-5 interact with the magnesium(II) center under catalytic condi-
tions. The observations that (a) isolable compounds 12-14 do not
undergo intramolecular cyclization, (b) a second substrate molecule
is required for cyclization, and (c) the unobserved active species
ToMMgNHR′(NH2R′) (16-18; R′ ) CH2CR2CH2CHCH2) are at
least four-coordinate (i.e., coordination sites are limited) are not
consistent with C-N bond formation via migratory insertion of
the olefin into the Mg-N bond. Instead, we propose that C-N
bond formation involves a six-center transition state as shown in
Scheme 2. Related transition state structures, in which a proton is
transferred from carbon to nitrogen, are proposed for ketone
enolization by magnesium amides.21 Additionally, we have studied
diamido zirconium(IV) hydroamination/cyclization catalysts where
C-N bond formations appear to involve proton transfer from one
amido to the terminal olefinic carbon in a related six-centered
transition state.22
This result contrasts the catalytic cyclizations where C-N bond
formation occurs readily at 50 °C. Thus, the intermediate 12 is not
kinetically competent to mediate olefin insertion on its own.18
Addition of a slight excess (g1.01 equiv) of aminoalkene 3 to
1 provides 12 quantitatively after 15 min at rt. When this solution
of 12 and 0.01 equiv of 3 is warmed to 50 °C, cyclization provides
compound 15 (eq 7). These stoichiometric reactions and our kinetic
measurements argue strongly that C-N bond formation occurs from
an amine adduct of 12 (i.e., 12·3) during catalysis.
Scheme 2. Proposed Mechanism for ToMMg-Mediated Aminoolefin
Hydroaminationa
Three-coordinate magnesium might be required for insertion.11b
In this scenario, eq 2 might involve amine-assisted dissociation of
one oxazoline ring. Therefore ToM2Mg (2) was used to assess
oxazoline-magnesium lability and the steric constraints imposed
by the ToM ligand.
Homoleptic ToM2Mg contains two bidentate ligands. The mag-
nesium center is coordinatively saturated and protected on all sides
by the dimethyl groups of the oxazolines (Figure 3). This claim is
supported quantitatively by crystallographic solid angles for 2,14b-d
which are 5.70 and 5.77 steradians (45 and 46%), suggesting
exchange (see below) involves oxazoline dissociation. This struc-
tural study further shows that an intramolecular insertion process
in putative {κ2-ToM}MgNHCH2CR2CH2CHdCH2 (∼40% free
space) is not ruled out on steric grounds.
a R′ ) CH2CR2CH2CHdCH2, R ) Ph2, -(CH2)5-, Me2.
Our kinetic studies and the reactivity of isolated magnesium
amidoalkenes provide unambiguous evidence that two substrate
molecules are interacting with the magnesium site in the turnover-
limiting cyclization step. Coordinative saturation stabilizes the amine-
free magnesium amido complex, affording a distinct and unique
environment compared to other coordinatively unsaturated hydroami-
nation/cyclization catalysts. Interestingly, the resting states of many
d0 and fnd0 metal center hydroamination/cyclization catalysts, including
lanthanidocenes,3 highly enantioselective binaphtholates compounds,
and magnesium diketiminates,11 are amido-amine complexes {LnM}-
NHR(NH2R)n, and the rate laws for these systems typically follow
zero-order dependence on substrate concentration. Under saturation
conditions, the rate law of eq 1 reduces to this commonly observed
rate law, providing a possible connection between hydroaminations
mediated by 1 and other d0-metal based catalysts. Despite this
connection, the relative importance of the mechanism proposed in
Scheme 2 with respect to insertion (or other) pathways remains to be
investigated for catalysts with increased coordinative unsaturation.
Figure 3. ORTEP diagram of ToM2Mg (2). Methyl groups that undergo
chemical exchange are colored identically (bold red and blue italics).
Selected bond lengths (Å): Mg1-N2, 2.063(4); Mg1-N3, 2.046(5);
Mg1-N4, 2.070(5); Mg1-N5, 2.065(4).
Acknowledgment. We thank Dr. A. Bakac and Dr. K. Kristian
for valuable mechanistic discussions. Mr. J. Engelkemier is thanked
for preparation of 3-d2. The U.S. DOE Office of Basic Energy
Science (DE-AC02-07CH11358) and the ACS Green Chemistry
Institute-PRF provided financial support. James F. Dunne is
An EXSY experiment at 50 °C showed selective exchange
between methyl groups (C3, C9, and C13; C4, C8, and C14; see
Figure 3) further supporting a dissociative rather than associative
exchange mechanism. The rates of oxazoline exchange (35-72 °C;
9
17682 J. AM. CHEM. SOC. VOL. 132, NO. 50, 2010

C O M M U N I C A T I O N S
(11) (a) Crimmin, M. R.; Casely, I. J.; Hill, M. S. J. Am. Chem. Soc. 2005, 127,
2042–2043. (b) Crimmin, M. R.; Arrowsmith, M.; Barrett, A. G. M.; Casely,
I. J.; Hill, M. S.; Procopiou, P. A. J. Am. Chem. Soc. 2009, 131, 9670–
9685. (c) Horrillo-Mart`ınez, P.; Hultzsch, K. C. Tetrahedron Lett. 2009,
50, 2054–2056. (d) Neal, S. R.; Ellern, A.; Sadow, A. D. J. Organomet.
Chem. 2010, doi: 10.1016/j.jorganchem.2010.08.057.
supported by a GAANN Fellowship. Aaron D. Sadow is an Alfred
P. Sloan Fellow.
Note Added after ASAP Publication. Scheme 1 contained an error
in the version published ASAP December 1, 2010; the correct version
reposted December 15, 2010.
(12) (a) Han, R.; Looney, A.; Parkin, G. J. Am. Chem. Soc. 1989, 111, 7276–
7278. (b) Han, R.; Parkin, G. J. Am. Chem. Soc. 1992, 114, 748–757. (c)
Chisholm, M. H.; Eilerts, N. W.; Huffman, J. C.; Iyer, S. S.; Pacold, M.;
Phomphrai, K. J. Am. Chem. Soc. 2000, 122, 11845–11854.
Supporting Information Available: Procedures and characterization
data for compounds 1, 2, 11-15, representative kinetic plots, and X-ray
crystallographic data. This material is available free of charge via the
Internet at http://pubs.acs.org.
(13) Dunne, J. F.; Su, J.; Ellern, A.; Sadow, A. D. Organometallics 2008, 27,
2399–2401.
(14) (a) Solid angles of [ToM] and Me in ToMMgMe are 7.4 and 2 steradians,
as determined by the program Solid-G using X-ray crystallographic
coordinates. (b) White, D.; Coville, N. J. In AdVances in Organometallic
Chemistry; Academic Press: New York, 1994; Vol. 36, pp 95-158. (c)
Guzei, I. A.; Wendt, M. Dalton Trans. 2006, 3991–3999. (d) Guzei, I. A.;
Wendt, M. Solid-G; UW-Madison, WI, USA, 2004.
References
(15) (a) Hultzsch, K. C.; Hampel, F.; Wagner, T. Organometallics 2004, 23,
2601–2612. (b) Thomson, R. K.; Bexrud, J. A.; Schafer, L. L. Organo-
metallics 2006, 25, 4069–4071.
(16) Espenson, J. H. Chemical kinetics and reaction mechanisms, 2nd ed.;
McGraw-Hill: New York, 1995.
(17) (a) See SI for kinetic analysis. (b) Cornish-Bowden, A. Fundamentals of
Enzyme Kinetics, 3rd ed.; Portland Press: London, 2004; pp 137-141.
(18) A referee suggested an alternative possibility for the apparent inert nature
of 12, where insertion is reversible but favors the magnesium-amido
reactant. While this explanation cannot be ruled out for the stoichiometric
reaction, the isotope effect and substrate saturation data, taken together,
are not consistent with a reversible insertion under catalytic conditions.
(19) (a) Bain, A. D.; Cramer, J. A. J. Phys. Chem. 1993, 97, 2884–2887. (b)
Bain, A. D. CIFIT; Chemistry Department, McMaster University, 2003.
(c) Kristian, K. E.; Iimura, M.; Cummings, S. A.; Norton, J. R.; Janak,
K. E.; Pang, K. Organometallics 2009, 28, 493–498.
(1) (a) Hong, S.; Marks, T. J. Acc. Chem. Res. 2004, 37, 673–686. (b) Mu¨ller,
T. E.; Hultzsch, K. C.; Yus, M.; Foubelo, F.; Tada, M. Chem. ReV. 2008,
108, 3795–3892.
(2) Gagne, M. R.; Marks, T. J. J. Am. Chem. Soc. 1989, 111, 4108–4109.
(3) (a) Gagne, M. R.; Stern, C. L.; Marks, T. J. J. Am. Chem. Soc. 1992, 114,
275–294. (b) Gribkov, D. V.; Hultzsch, K. C.; Hampel, F. J. Am. Chem.
Soc. 2006, 128, 3748–3759.
(4) Stubbert, B. D.; Marks, T. J. J. Am. Chem. Soc. 2007, 129, 6149–6167.
(5) (a) Walsh, P. J.; Hollander, F. J.; Bergman, R. G. J. Am. Chem. Soc. 1990,
112, 894–896. (b) Leitch, D. C.; Turner, C. S.; Schafer, L. L. Angew. Chem.,
Int. Ed. 2010, 49, 6382–6386.
(6) Katayev, E.; Li, Y.; Odom, A. L. Chem. Commun. 2002, 838–839.
(7) (a) Casalnuovo, A. L.; Calabrese, J. C.; Milstein, D. J. Am. Chem. Soc.
1988, 110, 6738–6744. (b) Dorta, R.; Egli, P.; Zürcher, F.; Togni, A. J. Am.
Chem. Soc. 1997, 119, 10857–10858.
(8) (a) Cowan, R. L.; Trogler, W. C. Organometallics 1987, 6, 2451–2453.
(b) Cowan, R. L.; Trogler, W. C. J. Am. Chem. Soc. 1989, 111, 4750–
4761. (c) Seligson, A. L.; Trogler, W. C. Organometallics 1993, 12, 744–
751.
(9) (a) Neukom, J. D.; Perch, N. S.; Wolfe, J. P. J. Am. Chem. Soc. 2010, 132,
6276–6277. (b) Hanley, P. S.; Markovic, D.; Hartwig, J. F. J. Am. Chem.
Soc. 2010, 132, 6302–6303.
(20) Mg-N bonds are highly polarized, and dissociation rates of Mg-N bonds
in 1 and 11-13 are likely equal to or greater than rates for 2 based on
Mg-N bond length and strain associated with κ3-coordination.
(21) (a) Henderson, K. W.; Allan, J. F.; Kennedy, A. R. Chem. Commun. 1997,
1149–1150. (b) He, X.; Morris, J. J.; Noll, B. C.; Brown, S. N.; Henderson,
K. W. J. Am. Chem. Soc. 2006, 128, 13599–13610.
(22) Manna, K.; Xu, S.; Sadow, A. D. Angew. Chem., Int. Ed., accepted for
publication.
(10) Motta, A.; Lanza, G.; Fragala, I. L.; Marks, T. J. Organometallics 2004,
23, 4097–4104.
JA108881S
9
J. AM. CHEM. SOC. VOL. 132, NO. 50, 2010 17683