Intramolecular C-H bond activation through a flexible ester linkage

Article abstract of DOI:10.1002/anie.201203163

Replacing the director: A bipyridinyl ligand with an aliphatic side chain determines the regioselectivity of copper-catalyzed C-H oxidation by intramolecular effects. Because the aliphatic chain is attached through an ester linkage, the catalytic cycle can in principle be closed by transesterification. Ion-mobility mass spectrometry and isotopic labeling provide mechanistic insight not available from direct mass spectrometry experiments. Copyright

Full text of DOI:10.1002/anie.201203163

Angewandte
Chemie
DOI: 10.1002/anie.201203163
CꢀH Activation
ꢀ
Intramolecular C H Bond Activation through a Flexible Ester
Linkage**
Christopher J. Shaffer, Detlef Schrçder,* Christoph Gꢀtz, and Arne Lꢀtzen*
Dedicated to Professor FranÅois Diederich on the occasion of his 60th birthday
The selective functionalization of alkanes is a hallmark
challenge within transition-metal catalysis. This important
field continues to be advanced by fundamental investigations
[
1]
of ion/molecule reactions in the gas phase. Conceptually,
+
+
+
+
diatomic metal-oxide ions (e.g., MgO , VO , FeO , NiO ,
+
and PtO ) have provided many of the most impressive results
in this field and have even demonstrated remarkable reac-
[
2]
+
tivity with methane; recently, also the reactive CuO cation
has been studied.
[
3]
The strength of these ions as functional oxidants, however,
is often in contrast to their moderate selectivity in reactions
with more complex substrates. In an effort to improve
selectivity through the application of adaptable ligands, we
[
4]
Scheme 1. Synthesis of 2 and electrospray ionization to generate
+
2
/Cu(NO ) .
3
+
have studied the reaction of phenanthroline-ligated CuO
[5]
[10]
with propane. Despite the improvement brought by ligation,
the selectivity could not be improved past an effective 1:5
preference for the activation of primary versus secondary
CꢀH bonds.
linkage. Synthetically, 2,2’-bipyridine-5-carboxylic acid (1)
is easily converted into the ester 2 (Scheme 1).
Electrospray ionization (ESI) of a solution of 2 in
methanol/water (1:1) containing an equimolar amount of
+
In an effort to increase the regiocontrol and to provide
perspectives for a possible transfer to the condensed phase,
we have looked towards synthetic advances in regioselective
CꢀH functionalization. Specifically, chemists have long used
copper nitrate generates the ion 2/Cu(NO ) in large abun-
3
[
11–15]
dance.
Collision-induced dissociation (CID) of mass-
selected ions is used to monitor the fragmentation processes
and the reactions preceding fragmentation. The major ionic
+
the distance limitations inherent in intramolecular reactions
product upon CID of 2/Cu(NO ) (m/z 437) corresponds to 1/
3
+
to control the regioselectivity of CꢀH functionalization
Cu(NO ) (m/z 325) which arises by ester cleavage concom-
3
[
6–9]
reactions.
Unfortunately, these concepts are also limited
itant with elimination of octene. Competitive with the ester
[6]
by weak activation interactions or are dependent on
irregular directing-group effects within the carbon frame-
fragmentation are losses of NO C and NO C, respectively.
2
3
+
While the latter process leads to the copper(I) complex 2/Cu
[
7a]
+
work.
(m/z 375), the loss of NO C leads to an ion X with m/z 391. By
2
Herein we offer a means of intramolecular regioselective
monitoring the appearance of these ions upon CID we
control in CꢀH bond activation combined with the potency
observed that although ester cleavage is favored, NO C and
2
and catalytic turnover delivered from transition-metal catal-
ysis. To this end, a bipyridine unit as a tunable ligand is
coupled with a long-chain alkyl group through an ester
NO C losses occur at similar threshold collisional energies
3
(Figure S1 in the Supporting Information).
+
Upon CID of mass-selected X , two major fragments
were observed, both occurring at higher collisional energies
+
than that required for the fragmentation of 2/Cu(NO ) .
[
*] Dr. C. J. Shaffer, Dr. D. Schrçder
3
+
Much like in 2/Cu(NO ) , the predominant dissociation
Institute of Organic Chemistry and Biochemistry
Academy of Sciences of the Czech Republic
Flemingovo nꢀm e˘ stꢁ 2, 16610 Prague 6 (Czech Republic)
E-mail: schroeder@uochb.cas.cz
3
+
pathway leads to the ester-cleavage product 1/Cu (m/z
+
263). However, the loss of water to produce [XꢀH O]
2
(
m/z 373) competes to a significant extent (Figure 1). The
C. Gꢂtz, Prof. Dr. A. Lꢂtzen
Kekulꢃ-Institute of Organic Chemistry and Biochemistry
University of Bonn
Gerhard-Domagk-Strasse 1, 53121 Bonn (Germany)
E-mail: arne.luetzen@uni-bonn.de
occurrence of dehydration suggests that, at some point, ion
+
X
or a transient ion of the same mass can promote an
activation of CꢀH bonds in ligand 2. A plausible suggestion is
+
that the loss of NO C from 2/Cu(NO ) leads to the copper–
2
3
+
III
oxo species 2/CuO (m/z 391) with a formal Cu center. The
[
**] This work was supported by the Academy of Sciences of the Czech
reactive metal-oxide cation may then activate a CꢀH bond of
Republish (RVO 61388963) and the European Research Council
[16]
(
AdG HORIZOMS).
the alkyl chain, followed by reductive elimination to afford
I
Cu complexes of the corresponding alcohols, for example, the
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201203163.
+
+
+
regioisomers 3/Cu , 4/Cu , and 5/Cu (Scheme 2, all also
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
+
Figure 1. Plot of the energy-dependent CID of X showing the competi-
+
tion of ester cleavage to yield 1/Cu and with neutral octene as well as
+
dehydration to afford the ion [XꢀH O] .
Figure 2. Region of the parent ion and the dehydration product upon
+ +
2
CID of the deuterium-labeled species [D ]-X , [8,8,8-D ]-X , and [7,7-
1
7
3
+
D ]-X .
2
+
[5]
activation of propane by (phenanthroline)CuO
the ratios
shown in Figure 2 imply a selectivity of 58% for dehydrogen-
ation across C(7)/C(8), 30% for dehydrogenation across C(6)/
C(7), and 12% for more internal positions. This selectivity is
significant as there is no directing effect other than the
distance from the ester linkage.
Ion-mobility mass spectrometry (IMMS) provides
a unique opportunity to investigate structural effects in gas-
[
17]
phase ions.
separates ions not just by mass, but also by shape.
Accordingly, we extended the above studies with IMMS
In addition to conventional MS, IMMS
[
18]
+
Scheme 2. Oxygen insertion into a CꢀH bond of 2/CuO to yield the
complexes of the corresponding alcohols 3–5.
[
19,20]
investigations of the ions generated by ESI.
For similar types of ions, plotting their arrival times (ta)
versus m/z in most instances leads to linear correlations, as
m/z 391), from which water elimination to the corresponding
olefin complexes can proceed.
+
+
found for the ions 1/Cu(NO ) (m/z 325), 2/Cu (m/z 375),
3
+
Solely reliant on ion masses and collision energies, this
experiment is not able to distinguish whether the insertion
and 2/Cu(NO ) (m/z 437). However, the arrival times for
3
+
+
ions X and [XꢀH O] with m/z 391 and 373, respectively, are
2
+
process occurs with the formation of X or is concurrent with
significantly shorter than expected from this correlation
(Figure 3). The greater mobilities of these ions can be
+
elimination and formation of [XꢀH O] . Moreover, the mass
2
spectrometric data does not allow the direct determination of
the site of CꢀH bond activation, because a measurable mass
difference occurs only after the elimination has taken place;
for example, we cannot distinguish between insertion of
oxygen in an w-CꢀH bond followed by 1,2-dehydration and
(
w-1)-CꢀH bond activation with subsequent loss of water.
In order to determine the origin of the hydrogen atoms
+
involved in the loss of water from X , we investigated
isotopologues of 2 with deuterated aliphatic chains. At first,
exclusive loss of D O upon perdeuteration of the octyl group
2
+
(
[D ]-X ) confirms the ester moiety as the source of the
1
7
hydrogen atoms involved in water formation and not the 2,2’-
biypyridine moiety (Figure 2a).
Deuterium labeling at the terminal position in [8,8,8-D ]-
3
+
X
results in an approximate 1:1 ratio of H O and HDO
Figure 3. Measured arrival times (t
ions generated by ESI of dilute solutions of Cu(NO ) in methanol/
water with either ligand 2 (marked as *) or the authentic alcohol
regioisomers 3–5 (marked as +, &, and , respectively) or the
in ms) of various mass-selected
a
2
losses. The substantial amount of HDO loss reveals oxidation
at the terminus of the alkyl chain. Deuteration of the adjacent
position in [7,7-D ]-X leads to a 1:5 ratio of H O and HDO
3 2
&
+
2
2
corresponding olefins 6 and 7 (both marked as ꢄ) as a function of the
mass-to-charge ratio; at m/z 373, the ꢄ und * symbols overlap. In
loss, suggesting a high selectivity for CꢀH bond activation at
6
3
65
the three most terminal carbon atoms. Based on the the
some cases, the corresponding Cu and Cu isotopes are both plotted
to demonstrate the linearity of mass effects.
kinetic isotope effect of k /k = 1.4 found for CꢀH bond
H
D
2
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
rationalized by assuming a recoil of the functionalized alkyl
chain to the copper center, such that the shapes of these
species become more compact. Interestingly, the shorter
+
arrival time of X suggests that Cu coordination occurs in this
ion as well. As it is unlikely that a purely aliphatic chain
coordinates with copper, we hypothesized that we already
+
observe a CꢀH-activated product, rather than 2/CuO .
Accordingly, we synthesized the authentic ligand components
for each of the proposed oxidation products, that is, ligands 3–
5
. Next, each of these ligands was measured separately in
+
+
Scheme 3. Dehydration of 3/Cu to create the regioisomers 6/Cu and
7
a solution of Cu(NO ) with an excess of NH OH·HCl to
encourage the formation of the corresponding Cu complexes
+
+
3
2
2
/Cu ; note that water elimination from 4/Cu might also lead to
I
+
+
+
6
/Cu and loss of H O from 5/Cu gives rise to 7/Cu , respectively.
2
+
+
+
3/Cu , 4/Cu , and 5/Cu , (all
m/z 391). The resulting ions
show subtle variations in
arrival times, but cumula-
tively overlay the broad
arrival time observed for
+
X , which we hence assign
to a mixture of these hydrox-
ylation products. This assign-
ment is further supported by
B3LYP/6-311 + G(2d,p) cal-
[21]
culations,
which predict
the singlet states of the Cꢀ
+
H-insertion products 3/Cu ,
+
+
4
6
/Cu , and 5/Cu to be 61,
ꢀ1
4, and 59 kcalmol , respec-
tively, lower in energy than
+
the triplet state of 2/CuO .
Based on the computed geo-
metries, the cross sections of
+
+
+
3
/Cu , 4/Cu , and 5/Cu are
estimated as 161, 160, and
Scheme 4. Gas-phase reactivity and theoretical completion of catalytic cycle.
2
1
58 ꢀ , respectively, com-
2
pared to a value of 187 ꢀ
for 2/CuO .
+
[22]
The signifi-
+
cantly larger cross section of 2/CuO rules out that this
Scheme 4). Transesterification with excess alcohol in solution
+
structure would correspond to X . These results imply that Cꢀ should make it possible for the tethered alcohol to be
+
H bond activation is facile and occurs already with the loss of
replaced by a fresh reductant leading to 2/Cu , from which re-
oxidation closes the catalytic cycle. The ester fragmentation
observed in the mass spectrometric experiments, therefore, is
not a detriment, because in order for this proof of concept to
be effective in solution, ester exchange must be competitive
with oxidation. In this context, we note that after a day of
storage, the solutions of 2 and Cu(NO ) in methanol/water
[
23]
NO C and not only during CID of m/z 391. Hence, it appears
2
+
that NO C loss from 2/Cu(NO ) leads to the high-valent
2
3
+
copper–oxo species 2/CuO only as a transient species and is
then followed by intramolecular CꢀH bond activation to form
I
+
+
+
the corresponding Cu complexes 3/Cu , 4/Cu , and 5/Cu ,
respectively.
3
2
As mentioned above, also the small t value of the ion with
m/z 373 implies a difference in shape, which is ascribed to
show degradation to the corresponding methyl esters, thereby
demonstrating the facile occurrence of transesterification in
solution.
a
a coordination of copper to the olefinic moiety in the side
+
+
chains of 6/Cu and 7/Cu , respectively (Scheme 3). This
In the present gas-phase experiments, we are not able to
+
[24]
speculation is confirmed by Cu complexes of authentic
close the catalytic cycle.
However, re-oxidation of the
samples of 6 and 7, whose arrival times match that of the
copper species and ester exchange are feasible concepts since
+
+
[
XꢀH O] ion generated from 2/Cu(NO ) .
it is now established that intramolecular CꢀH bond activation
2
3
In summary, our experiments suggest the following
can occur. This gives additional credence to our proposal that
sequence of events (Scheme 4). Initial homolysis of the Nꢀ similar reactivity can be identified in solution in which that
+
O bond in 2/Cu(NO ) leads to the high-valent copper–oxo
the oxidant can be tuned appropriately to create reactive
metal–oxo species under ambient conditions.
3
+
species 2/CuO , which then rapidly undergoes CꢀH bond
+
activation to give the corresponding alcohol complexes 3/Cu ,
+
+
4
/Cu , and 5/Cu , respectively (only the former is shown in
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
by application of an excitation AC voltage to the end caps of the
trap to induce collisions of the kinetically excited ions with the
helium buffer gas. For a CID excitation period of 20 ms and
Received: April 24, 2012
Published online: && &&, &&&&
a trapping parameter of q = 0.25, we have recently introduced
z
Keywords: CꢀH activation · copper · gas phase · ion mobility ·
an empirical calibration scheme that allows a conversion of the
.
[
15]
oxidation
experimental collision energies to an absolute scale.
[
12] A. Tintaru, J. Roithovꢁ, D. Schrçder, L. Charles, I. Ju sˇ inski, Z.
Glasovac, M. Eckert-Maksi c´ , J. Phys. Chem. A 2008, 112, 12097 –
12103.
[13] D. Schrçder, T. Weiske, H. Schwarz, Int. J. Mass Spectrom. 2002,
219, 729 – 738.
[
1] J. Roithovꢁ, D. Schrçder, Chem. Rev. 2010, 110, 1170 – 1211.
2] H. Schwarz, Angew. Chem. 2011, 123, 10276 – 10297; Angew.
Chem. Int. Ed. 2011, 50, 10096 – 10115.
[
[
3] N. Dietl, C. van der Linde, M. Schlangen, M. K. Beyer, H.
Schwarz, Angew. Chem. 2011, 123, 5068 – 5072; Angew. Chem.
Int. Ed. 2011, 50, 4966 – 4969.
4] The dixoide cations CrO2 and PtO2 , for example, are among
the most reactive metal-oxide cations in the gas phase, but they
already show low selectivity in product formation even with
methane as the most simple hydrocarbon substrate, see: a) A.
Fiedler, I. Kretzschmar, D. Schrçder, H. Schwarz, J. Am. Chem.
Soc. 1996, 118, 9941 – 9952; b) M. Brçnstrup, D. Schrçder, I.
Kretzschmar, H. Schwarz, J. N. Harvey, J. Am. Chem. Soc. 2001,
[14] R. A. J. OꢃHair, Chem. Commun. 2006, 1469 – 1481.
[15] a) ꢄ. Rꢅvꢅsz, P. Milko, J. Zabka, D. Schrçder, J. Roithovꢁ, J.
Mass Spectrom. 2010, 45, 1246 – 1252; b) E. L. Zins, C. Pepe, D.
Schrçder, J. Mass Spectrom. 2010, 45, 1253 – 1260.
[16] The general sequence for such hydroxylation reactions of
ˇ
+
+
[
+
alkanes by gaseous metal-oxide ions is: R-H + MO ! R-H/
+
+
+
+
+
MO ! RC/MOH or R-MOH ! R-OH/M ! R-OH + M ;
see Refs. [1] and [2].
[17] a) A. B. Kanu, P. Dwivedi, M. Tam, L. Matz, H. H. Hill, J. Mass
Spectrom. 2008, 43, 1 – 22; b) B. C. Bohrer, S. I. Merenbloom,
S. L. Koeniger, A. E. Hilderbrand, D. E. Clemmer, Annu. Rev.
Anal. Chem. 2008, 1, 293 – 327; c) J. Puton, M. Nousiainen, M.
Sillanpꢆꢆ, Talanta 2008, 76, 978 – 987.
[18] In brief, ion-mobility separation can be compared to chroma-
tography in the condensed phase. An electric field gradient
provides a driving force for the mobile phase, which consists of
ions of interest. Collisional interactions with the inert gas (here
1
23, 142 – 147.
5] D. Schrçder, M. C. Holthausen, H. Schwarz, J. Phys. Chem. B
004, 108, 14407 – 14416.
6] R. Breslow, S. Baldwin, T. Flechtner, P. Kalicky, S. Liu, W.
Washburn, J. Am. Chem. Soc. 1973, 95, 3251 – 3262.
[
[
[
2
7] a) D. Yang, M.-K. Wong, X.-C. Wang, Y.-C. Tang, J. Am. Chem.
Soc. 1998, 120, 6611 – 6612; b) S. Das, G. W. Brudvig, R. H.
Crabtree, J. Am. Chem. Soc. 2008, 130, 1628 – 1637.
N at 2 mbar) serve as the stationary phase. Much like in
2
[
8] For gas-phase variants of remote functionalization, see: H.
Schwarz, Acc. Chem. Res. 1989, 22, 282 – 287.
electrophoresis, two isobaric ions of different shape (“potato and
cigar”), the more compact one interacts less with the stationary
phase and thus reaches the detector before the more extended
(larger surface area) ion of identical mass.
[
9] For remote functionalization of carbonyl compounds in the gas
phase, see: a) G. Czekay, K. Eller, D. Schrçder, H. Schwarz,
Angew. Chem. 1989, 101, 1306 – 1308; Angew. Chem. Int. Ed.
Engl. 1989, 28, 1277 – 1278; b) D. Schrçder, H. Schwarz, J. Am.
Chem. Soc. 1990, 112, 5947 – 5953; c) D. Schrçder, H. Schwarz, J.
Am. Chem. Soc. 1993, 115, 8818 – 8820; d) D. Schrçder, W.
Zummack, H. Schwarz, J. Am. Chem. Soc. 1994,116 , 5857 – 5864.
[19] Ion-mobility mass spectrometry experiments were performed
with a SYNAPT G2 ion-mobility mass spectrometer (Waters,
Manchester, UK). For details on the experimental setup and its
operation, see Refs. [17a] and [20].
[20] a) ꢄ. Rꢅvꢅsz, D. Schrçder, T. A. Rokob, M. Havlꢇk, B. Dolensk y´ ,
Angew. Chem. 2011, 123, 2449 – 2452; Angew. Chem. Int. Ed.
2011, 50, 2401 – 2404; b) D. Schrçder, Collect. Czech. Chem.
Commun. 2011, 76, 351 – 369.
[21] M. J. Frisch et al. Gaussian09; RevisionA.02, 2009.
[22] For the method used to calculate the cross sections, see: ꢄ.
Rꢅvꢅsz, D. Schrçder, T. A. Rokob, M. Havlꢇk, B. Dolensk y´ , Phys.
Chem. Chem. Phys. 2012, 14, 6987 – 6995.
[
[
10] U. Kiehne, A. Lꢂtzen, Org. Lett. 2007, 9, 5333 – 5336.
11] The experiments were performed using a Finnigan LCQ Classic
[
12]
ion-trap mass spectrometer which bears a conventional ESI
source consisting of the spray unit (typical flow rates ranged
ꢀ1
between 5 and 30 mLmin ; typical spray voltage was 5 kV) with
nitrogen as a sheath gas, followed by a heated transfer capillary
(
kept at 2008C), a first set of lenses which determined the
softness or hardness of ionization by variation of the degree of
[23] See also: L. Ja sˇ ꢇkovꢁ, E. Hanik y´ rˇ ovꢁ, D. Schrçder, J. Roithovꢁ, J.
[
13]
collisional activation in the medium-pressure regime,
transfer octopoles, and a Paul ion-trap with ca. 10 mbar helium
two
Mass Spectrom. 2012, 47, 460 – 465.
ꢀ3
[24] For a review on gas-phase catalysis, see: D. K. Bçhme, H.
Schwarz, Angew. Chem. 2005, 117, 2388 – 2406; Angew. Chem.
Int. Ed. 2005, 44, 2336 – 2354.
n
for ion storage and manipulation, including a variety of MS
[
14]
experiments. For detection, the ions were ejected from the
trap to an electron multiplier. Low-energy CID was performed
4
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Communications
CꢀH Activation
Replacing the director: A bipyridinyl
ligand with an aliphatic side chain deter-
mines the regioselectivity of copper-cat-
alyzed CꢀH oxidation by intramolecular
C. J. Shaffer, D. Schrçder,* C. Gꢁtz,
A. Lꢁtzen*
&&&& — &&&&
effects. Because the aliphatic chain is
attached through an ester linkage, the
catalytic cycle can in principle be closed
by transesterification. Ion-mobility mass
spectrometry and isotopic labeling pro-
vide mechanistic insight not available
from direct mass spectrometry experi-
ments.
Intramolecular CꢀH Bond Activation
through a Flexible Ester Linkage
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!