Potassium 2,2,5,7,8-pentamethylchroman-6-oxide: A rationally designed base for Pd-catalysed amination

Article abstract of DOI:10.1002/chem.201103284

This base is just right! A new base, potassium 2,2,5,7,8- pentamethylchroman-6-oxide, for Pd-catalysed amination has been rationally designed that has sufficient basicity to efficiently deprotonate the intermediate aryl palladium ammonium complex, leading to far better rates than obtained with carbonate, which is known to be very mild. At the same time, this new base possesses minimal nucleophilicity to mitigate the degradation of base-sensitive functional groups in the starting materials and products, such as is commonly seen with potassium or sodium tert-butoxide. Copyright

Full text of DOI:10.1002/chem.201103284

DOI: 10.1002/chem.201103284
Potassium 2,2,5,7,8-Pentamethylchroman-6-oxide:
A Rationally Designed Base for Pd-Catalysed Amination
Ka Hou Hoi and Michael G. Organ*[a]
Metal-catalysed amination of aryl halides has become per-
haps the method of choice to prepare substituted aniline de-
rivatives.[1] Since the first amination of a non-metal amide
complex in 1983 using catalytic amounts of metal,[2] consid-
erable strides have been made in catalyst design, reaction
optimization, and mechanistic understanding of this valuable
transformation involving simple amines.[3] That said, the
nature of, and role/effect of the ancillary base that is neces-
sary to complete the catalytic cycle (Scheme 1)[4] has been
deprotonation most likely takes place on the surface of the
heterogeneous base.[9] The gentle nature of carbonates
makes them compatible with a broad spectrum of functional
groups (e.g., alcohols, amines, ketones, nitriles, esters,
amides), thus appealing for use in amination. However,
many amination protocols show a first-order dependence on
base with the consequence that couplings using carbonate
generally require very high reaction temperatures and pro-
longed reaction times.[10] Conversely, strong alkoxide bases,
such as tert-butoxide (pKa tert-butyl alcohol ca. 18), are also
widely employed in amination. They are quite soluble in or-
ganic media and highly effective deprotonating agents lead-
ing to greatly improved reaction rates when compared with
carbonate. For example, we recently demonstrated that Pd-
PEPPSI-IPent precatalyst (see structure 8; PEPPSI=pyri-
dine, enhanced, precatalyst, preparation, stabilisation, and
initiation;
IPent=N,N-bis(2,6-diisopentylphenylimidazoli-
um)) quantitatively couples 4-chlorocyanobenzene with
morpholine at 808C over approximately 3 h with Cs2CO3 as
base; the same reaction using KOtBu at room temperature
was complete within 10 s![10] However, KOtBu is poorly
functional group tolerant, which means that aminations em-
ploying it are limited to substrates that are relatively simple,
for example lacking the functionality listed above. To illus-
trate this point, consider the simple reaction in Scheme 2.
Compound 1, a typical substrate for amination, was exposed
to amination conditions using KOtBu in DME at 808C[10]
and this led to the complete degradation of 1.
Scheme 1. Putative Pd-catalysed amination mechanism.
far less studied experimentally.[5] Perhaps this is not surpris-
ing as bases that are available to the synthetic chemist are
pretty much assumed to be all known, and it is generally
perceived to be a matter of screening them to see which one
performs the best for a particular application. Certainly
cation effects can be quite pronounced, but the anionic
ꢀbusiness endꢁ of the salt pair for amination has been limited
primarily to carbonates and alkoxides.[6,7]
When considering the general, putative amination mecha-
nism (Scheme 1), the baseꢁs role, on the surface, would be
deprotonation of the aryl palladium ammonium complex.
Assuming a pKa value of approximately 8–10 for this
proton, carbonate is barely suitable for deprotonation (pKa
HCO3 ca. 10.3),[8] which is further exacerbated by its negli-
À
gible solubility in organic solvents. Particle size, hence sur-
face area, has been shown to be critical to the success of
amination protocols employing carbonate suggesting that
Scheme 2. Base compatibility test with methyl 4-chlorobenzoate.
With the goal of increasing substrate scope, while keeping
reaction times as short as possible, we took a step back and
re-evaluated the structure of the base as it pertains to reac-
tivity in metal-catalysed amination. While soluble alkoxides
clearly are well capable of deprotonation leading to desira-
ble kinetics, their relatively strong nucleophilicity diminishes
their selectivity, which curtails their wide-spread use. If the
[a] K. H. Hoi, Prof. M. G. Organ
Department of Chemistry, York University
4700 Keele Street, Toronto, ON, M3J 1P3 (Canada)
Fax : (+1)416-736-5936
E-mail: organ@yorku.ca
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201103284.
804
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 804 – 807

COMMUNICATION
base could be softened, while simultaneously diminishing
nucleophilicity, a more balanced base for amination could
be realised. To this end, bases with conjugate acid pKa
values between 11 and 15 were targeted. We investigated
1,8-diazabicycloACHTUNGTRENNUNG[5.4.0]undec-7-ene (DBU) (pKa guanidinium
ca. 13.5) and potassium trimethylsilanoxide whose pKa is in
the same vicinity as carbonate (pKa silanol ca. 11) but it is
fully soluble. Neither base showed any hint of reactivity in
preliminary amination studies. Both bases should be suitable
for deprotonation; it is possible that catalyst poisoning, in
particular with DBU, could be the source of the problem. In
any case, we continued our search and next considered
phenoxide-derived bases. While this choice might appear
esoteric, there is precedence for the use of phenoxide itself,
or substituted derivatives thereof, in amination[5] and amino
carbonylation[11] reactions. Phenol has a pKa of 9.95, which
we deemed to be lower than desired (i.e., simple phenoxide
would be too weak a base), and further phenoxide is a
known nucleophile in Pd-catalysed substitution reactions
that could lead to selectivity and product contamination
problems.[11] Interestingly, while the methoxy group in ani-
sole is viewed as a strong electron donor in electrophilic ar-
omatic substitution, p-OCH3 only increases the pKa of
phenol to 10.2. Perhaps even more enlightening is that the
pKa of p-cresol is essentially the same (10.19) as that of p-
methoxyphenol. This would suggest that the lone pairs of
the methoxy substituent are not drawn strongly into conju-
gation with the ring. The presence of the phenolic oxygen
atom would disfavour delocalisation of electron density into
the ring, thus minimising the destabilisation brought about
by the OCH3 group that would increase the pKa, thus im-
proving basicity. Further, A1,3-type strain builds between the
ring and the methyl group when one of the lone pairs on
oxygen is brought into conjugation with it. We reasoned if
the alkoxy oxygen atom could be fixed conformationally
such that one lone pair was locked into full conjugation with
the aromatic ring that we could ꢀforceꢁ the pKa value up,
thus increasing the rate of amination where deprotonation
was rate limiting.[10] To this end we turned to nature and ex-
amined the structure of a-tocopherol (vitamin E), whose
pKa has been highly studied as a function of its biological
activity. In water, the pKa of a-tocopherol has been shown
to vary between 11 and 13 with different surfactant-like ad-
ditives.[12] Inspired by this, we examined a truncated version
of a-tocopherol, specifically 2,2,5,7,8-pentamethyl-6-chroma-
nol (4a). Using proprietary in-house software, we calculated
the pKa of 4a to be 11.4; as a control, phenol was deter-
mined to be 9.9, which compares well with the experimental
data.[13] When the potassium chromanoxide salt 4b was al-
lowed to react under the same conditions as KOtBu with 1
(Scheme 2), there was no degradation whatsoever of the
sensitive starting material. As part of our design, the methyl
groups that flank the phenoxide site offset the stonger basic-
ity of a-tocopherol by sterically reducing its nucleophilicity,
which seems to be validated by this result.
Figure 1. Comparison of the rates of amination of p-chlorotoluene with
morpholine by Pd-PEPPSI-IPent using KOtBu, Cs2CO3, and the K (4b),
Na (4c), and Li (4d) salts of 2,2,5,7,8-pentamethyl-6-chromanol.
monly used amination bases using precatalyst 8[14,15]
(Figure 1). In the absence of sensitive functionality, the im-
pressively high reactivity of KOtBu in combination with 8 is
clear. Also clear is the vastly slower reactivity of Cs2CO3.
We were delighted that 4b was not only suitable for the re-
action, but that it showed vastly improved kinetics relative
to Cs2CO3 completing the transformation in approximately
2 h. The sodium salt 4c had similar reactivity relative to 4b,
while the lithium salt (4d) showed essentially no reactivity.
This could be due to poor solubility of 4d, however addition
of a strong lithium chelater ([12]crown-4) did not signifi-
cantly improve the situation.
To further evaluate our base design, we stripped off the
second ring providing 9b and this was accompanied by a sig-
nificant drop in reactivity (Figure 2). While the alkoxide end
of 9b would be similarly hindered, 9b is, overall, less cum-
bersome than 4b and this should improve the baseꢁs perfor-
mance from a steric point of view. Thus, the diminished per-
formance would be consistent with the basis for the reactivi-
ty of 4b being primarily electronic (i.e., pKa) and that in the
absence of the conformational lock the alkoxy group does
little to enhance the phenolꢁs basicity. Removal of the flank-
ing methyl groups (i.e. 9a) renders the base totally ineffec-
tive for this transformation and the starting materials were
untouched after 24 h.
Having established the mildness of 4b with base-sensitive
compounds, and its very good reactivity as a base in the con-
trol amination reaction, it remained to demonstrate its utili-
ty in the amination of base-sensitive aryl chlorides and
amines (see Tables 1 and 2). For demonstration purposes,
the reaction was performed with substrates possessing a
With the compatibility study results in hand, we attempt-
ed a test coupling to see how 4b compared with other com-
Chem. Eur. J. 2012, 18, 804 – 807
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
805

M. G. Organ and K. H. Hoi
Table 2. Amination reactions involving base-sensitive substrates by the
Pd-RuPhos catalyst using KOtBu and 4b. Reported yields [%] were de-
termined on isolated material following silica gel chromatography; reac-
tions were performed in duplicate.
taining the base-sensitive group was not detected in the
crude product mixtures, indicating that it had been fully de-
composed under these standard amination conditions. Nota-
bly, 4b efficiently produced the highly hindered product 13
(Table 1), which typically requires forcing conditions, such
as butoxide base. Additionally, starting materials that con-
tain alcohols (e.g., leading to 15, Table 1), which will be de-
protonated before coupling begins with either KOtBu or 4b,
also produce high product yields.
In conclusion, we have rationally designed a base from
first principles for Pd-catalysed amination. The base, potassi-
um 2,2,5,7,8-pentamethyl-6-chromanoxide (4b),[17] has the
necessary basicity (pKa ca. 11.4) to facilitate the efficient de-
protonation of the corresponding aryl palladium ammonium
complexes to give desirable kinetics, while having dimin-
ished nucleophilicity to mitigate undesired side reactions
with base-sensitive functional groups. The base has shown
excellent reactivity and selectivity with both highly reactive
phosphine- and NHC-based Pd catalysts indicating that 4b
will have high compatibility, thus good generality and broad
uptake in metal-catalysed amination reactions.
Figure 2. Comparison of the rates of amination of p-chlorotoluene with
morpholine by Pd-PEPPSI-IPent using various phenoxide bases.
ketone (10, 14), ester (12), carbamate (15), or nitrile (11)
using two structurally diverse, highly reactive catalyst sys-
tems for amination, the NHC complex 8 (Table 1) and the
phosphine-based Pd-RuPhos catalyst (Table 2).[17] In most
cases, when KOtBu was used there was minimal (or no)
amination product observed and the starting material con-
Table 1. Amination reactions involving base-sensitive substrates by Pd-
PEPPSI-IPent using KOtBu and 4b. Reported yields [%] were deter-
mined on isolated material following silica gel chromatography; reactions
were performed in duplicate. [a] 3 equiv of base used. [b] Aryl bromide
was used.
Experimental Section
General procedure for Pd-PEPPSI-IPent-catalysed amination: In air, a
vial (3 mL screw-cap threaded) equipped with a stir bar was charged with
8 (15.8 mg, 4 mol%). Caesium carbonate (489 mg, 1.5 mmol) or KOtBu
(84.2 mg, 0.75 mmol) were weighed out in the glovebox and transferred
into the reaction vial under argon. The vial was sealed with a Teflon-
lined screw cap and purged with argon (3ꢃ). When potassium phenoxide
bases were used, the corresponding phenol (0.75 mmol) was added to a
vial containing an equimolar amount of KH in DME (0.25 mL), and the
mixture was stirred until the effervescence subsided (ca. 10 min). The
aryl halide (0.5 mmol), amine (0.75 mmol) and DME (0.25 mL) were
added subsequently by using a syringe. Alternatively, if the aryl halide
was a solid at room temperature, it was introduced into the reaction vial
prior to purging with argon. The solution was stirred at 808C for the indi-
cated time, after which it was diluted with diethyl ether (2 mL) and fil-
tered through a plug of silica covered with celite. The reaction vial and
806
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 804 – 807

Rationally Designed Base for Pd-Catalysed Amination
COMMUNICATION
silica pad were rinsed with additional diethyl ether (10 mL) and the or-
ganic layers combined. The solvent was removed in vacuo and the resi-
[5] a) L. M. Alcazar-Roman, J. F. Hartwig, J. Am. Chem. Soc. 2001, 123,
12905–12906; b) S. Shekhar, J. F. Hartwig, Organometallics 2007, 26,
340–351.
due purified by using
a Biotage Isolera UV-VIS Flash Purification
System with prepacked silica cartridges.
[6] M. G. Organ, G. A. Chass, D.-C. Fang, A. C. Hopkinson, C. Valente,
Synthesis 2008, 2776–2797.
General procedure for Pd-RuPhos-catalysed amination: The above pro-
[7] For examples of amination using LHMDS, see: a) I. C. Lerma, M. J.
Cawley, K. Arentsen, J. S. Scott, S. E. Pearlson, J. Hayler, S. Cad-
dick, J. Organomet. Chem. 2005, 690, 5841–5848; b) K. W. Ander-
son, R. E. Tundel, T. Ikawa, R. A. Altman, S. L. Buchwald, Angew.
Chem. 2006, 118, 6673–6677; Angew. Chem. Int. Ed. 2006, 45, 6523–
6527; c) D. Maiti, B. P. Fors, J. L. Henderson, Y Nakamura, S. L.
Buchwald, Chem. Sci. 2011, 2, 57–68.
cedure for 8 was used with the following exceptions: [Pd2ACTHNUGTRNE(UNG dba)3] (9.2 mg,
2 mol%; dba=dibenzylideneacetone) and the RuPhos ligand (9.3 mg,
4 mol%) were loaded into a reaction vial in a glove box and the vial was
sealed with a Teflon-lined screw cap and purged with argon (3ꢃ). DME
(0.25 mL) was added and the solution was stirred for 10 min to allow the
Pd-RuPhos complex to form. From this point forward, all other reaction
components were introduced as detailed above for 8 and the reactions
were conducted and analysed also as outlined above.
[8] All pKa values cited in this manuscript were recorded in water and
compared directly on that basis.
[9] C. Meyers, B. U. W. Maes, K. T. J. Loones, G. Bal, G. L. F. Lemiꢄre,
R. A. Dommisse, J. Org. Chem. 2004, 69, 6010–6017.
[10] K. H. Hoi, S. C¸ alimsiz, R. D. J. Froese, A. C. Hopkinson, M. G.
Organ, Chem. Eur. J. 2011, 17, 3086–3090, and references therein.
[11] J. R. Martinelli, T. P. Clark, D. A. Watson, R. H. Munday, S. L.
Buchwald, Angew. Chem. 2007, 119, 8612–8615; Angew. Chem. Int.
Ed. 2007, 46, 8460–8463.
Acknowledgements
This work was supported by the NSERC (Canada) and the ORDCF (On-
tario). We thank Dr. Jeff Zablocki of Gilead Pharmaceuticals for per-
forming pKa calculations. We are grateful to Professor Yasunoi Minami
of Chuo University (Tokyo) for his useful suggestions on this project.
[12] C. J. Drummond, F. Grieser, Biochimica et Biophysics Acta 1985,
836, 275–278.
[13] The calculations were performed by Dr. Jeff Zablocki at Gilead
Pharmaceuticals, Inc., Foster City, California.
Keywords: amination · bases · N-heterocyclic carbenes ·
palladium · synthetic methods
[14] For other cross-coupling reactions demonstrating the high catalytic
performance of Pd-PEPPSI-IPent, see: a) M. G. Organ, S. Calimsiz,
M. Sayah, K. H. Hoi, A. J. Lough, Angew. Chem. 2009, 121, 2419–
2423; Angew. Chem. Int. Ed. 2009, 48, 2383–2387; b) S. C¸ alimsiz, M.
Sayah, D. Mallik, M. G. Organ, Angew. Chem. 2010, 122, 2058–
2061; Angew. Chem. Int. Ed. 2010, 49, 2014–2017; c) M. Dowlut, D.
Mallik, M. G. Organ, Chem. Eur. J. 2010, 16, 4279–4283; d) C. Val-
ente, M. E. Belowich, N. Hadei, M. G. Organ, Eur. J. Org. Chem.
2010, 23, 4343–4354.
[1] a) Handbook of Organopalladium Chemistry for Organic Synthesis
(Ed.: E. I. Negishi), Wiley-Interscience, New York, 2002; b) A. de
Meijere and F. Diederich in Metal-Catalyzed Cross-Coupling Reac-
tions, 2nd ed., Wiley-VCH, Weinheim, 2004; c) J. F. Hartwig in Or-
ganotransition Metal Chemistry, University Science Books, Califor-
nia, 2010.
[2] N. V. Kondratenko, A. A. Kolomeitsev, V. O. Mogilevskaya, N. M.
Varlamova, L. M. Yagupolꢁskii, J. Org. Chem. USSR. 1986, 22,
1547–1554. Translated from: Zh. Org. Khim. 1986, 22, 1721–1729.
[3] a) A. S. Guram, R. A. Rennels, S. L. Buchwald, Angew. Chem. 1995,
107, 1456–1459; Angew. Chem. Int. Ed. Engl. 1995, 34, 1348–1350;
b) J. Louie, J. F. Hartwig, Tetrahedron Lett. 1995, 36, 3609–3612.
[4] S. Shekhar, P. Ryberg, J. F. Hartwig, J. S. Mathev, D. G. Blackmond,
E. R. Strieter, S. L. Buchwald, J. Am. Chem. Soc. 2005, 127, 3584–
3591.
[15] Pd-PEPPSI-IPent is now commercially available through Sigma–Al-
drich (Catalogue Number 732117).
[16] All compounds prepared in this study were evaluated by 1H and
13C NMR spectroscopy and mass spectrometry. The spectra thus ob-
tained were consistent with spectral data reported in the literature.
[17] 2,2,5,7,8-Pentamethyl-6-chromanoxide is now commercially avail-
able through Sigma–Aldrich (Catalogue Number 430676).
Received: October 18, 2011
Published online: December 16, 2011
Chem. Eur. J. 2012, 18, 804 – 807
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
807