Remote carboxylation of halogenated aliphatic hydrocarbons with carbon dioxide

Article abstract of DOI:10.1038/nature22316

Catalytic carbon-carbon bond formation has enabled the streamlining of synthetic routes when assembling complex molecules. It is particularly important when incorporating saturated hydrocarbons, which are common motifs in petrochemicals and biologically relevant molecules. However, cross-coupling methods that involve alkyl electrophiles result in catalytic bond formation only at specific and previously functionalized sites. Here we describe a catalytic method that is capable of promoting carboxylation reactions at remote and unfunctionalized aliphatic sites with carbon dioxide at atmospheric pressure. The reaction occurs via selective migration of the catalyst along the hydrocarbon side-chain with excellent regio- and chemoselectivity, representing a remarkable reactivity relay when compared with classical cross-coupling reactions. Our results demonstrate that site-selectivity can be switched and controlled, enabling the functionalization of less-reactive positions in the presence of a priori more reactive ones. Furthermore, we show that raw materials obtained in bulk from petroleum processing, such as alkanes and unrefined mixtures of olefins, can be used as substrates. This offers an opportunity to integrate a catalytic platform en route to valuable fatty acids by transforming petroleum-derived feedstocks directly.

Full text of DOI:10.1038/nature22316

LETTER
doi:10.1038/nature22316
Remote carboxylation of halogenated aliphatic
hydrocarbons with carbon dioxide
Francisco Juliá-Hernández1, Toni moragas1, Josep Cornella1 & Ruben martin1,2
Catalytic carbon–carbon bond formation has enabled the streamlining and chemoselectivity, representing a remarkable reactivity relay
of synthetic routes when assembling complex molecules1. It is when compared with classical cross-coupling reactions. Our results
particularly important when incorporating saturated hydrocarbons, demonstrate that site-selectivity can be switched and controlled,
which are common motifs in petrochemicals and biologically enabling the functionalization of less-reactive positions in the
relevant molecules. However, cross-coupling methods that involve presence of a priori more reactive ones. Furthermore, we show that raw
alkyl electrophiles result in catalytic bond formation only at specific materials obtained in bulk from petroleum processing, such as alkanes
and previously functionalized sites2. Here we describe a catalytic and unrefined mixtures of olefins, can be used as substrates. This offers
method that is capable of promoting carboxylation reactions at an opportunity to integrate a catalytic platform en route to valuable
remote and unfunctionalized aliphatic sites with carbon dioxide at fatty acids by transforming petroleum-derived feedstocks directly4.
atmospheric pressure. The reaction occurs via selective migration of
Methods that incorporate saturated hydrocarbon chains have tradi-
the catalyst along the hydrocarbon side-chain3 with excellent regio- tionally been problematic using palladium catalysts2. Pioneering work5,6
Figure 1 | Switchable site-selective catalytic
CO2H
a
Ha
carboxylation at remote sp3 C–H sites.
Path a
a, Classical cross-coupling reactions involve
the use of alkyl electrophiles with carbon-based
counterparts (coloured circles labelled ‘C’)
and occur at the original reaction site (left).
Reactivity relay (unconventional coupling)
in carbon–carbon bond formation of alkyl
electrophiles with switchable site-selectivity
gives rise to two divergent products from a
common precursor (right, paths a and b).
b, Mechanistic rationale for the switchable site-
selective carboxylation at remote sp3 C–H sites.
Tunable and controllable displacement of the Ni
catalyst through a saturated hydrocarbon side-
chain (nickel catalytic cycles A and B) is shown.
c, Application to the direct catalytic conversion
of biomass-derived feedstocks into single fatty
acids via a tandem bromination/carboxylation
process. Coloured circles represent substituents
at the aliphatic side-chain.
Remote
site
Reaction at
remote site (Hb)
CO2
feedstock
C
C
Hb
X
Ha
• Switchable site-selectivity
• No organometallic reagents
‘Classical’
coupling
Unconventional
coupling
• Mild and chemoselective
Reaction at
original site
X = I, Br, Cl, OR
X = Br, OR
Alkyl electrophile
Hb
CO2H
Path b
Reaction at
remote site (Ha)
b
CO2H
CO2H
Br
Fatty acid
Fatty acid
Reductant
Easily accessible,
inexpensive
LnNi0
LnNi0
Reductant
CO2
CO2
feedstock
feedstock
Ni
Ni
Ni
Nickel catalytic
cycle (B)
Nickel catalytic
cycle (A)
III
I
V
Slow
Fast
Fast
Fast
Fast
CO2H
Ni
H
H
Ni
II
IV
Not observed
c
Unnecesary reꢀnement
Ha
Single regioisomer
Alkanes
CO2
Bromination
feedstock
Chemical
feedstock
Ha
CO2H
Br
Ni catalyst
Added-value fatty acids
Inconsequential mixtures
of bromoalkanes
Ha
Unreꢀned oleꢀn
mixtures
1Institute of Chemical Research of Catalonia (ICIQ), The Barcelona Institute of Science and Technology, Avinguda Països Catalans 16, 43007 Tarragona, Spain. 2ICREA, Passeig Lluïs Companys 23,
08010 Barcelona, Spain.
8 4
| N A T U R E | V O L 5 4 5 | 4 m A y 2 0 1 7
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

letter reSeArCH
Single regioisomer
2.5 mol% NiI2
4.4 mol% L1
Ph
Ph
Br
H
H
H
H
+
CO2
R
Ha
R
CO2H
n
n
Mn, DMF (0.30–1.0 M)
N
N
25 ºC, 20 h
n-Hex
n-Hex
Alkyl bromide
Carboxylic acid
L1
• Secondary alkyl bromides
OPiv
Me
Me
Me
CO2H
CO2H
CO2H
CO2H
CO2H
CO2H
1
1
81%
1
2
3
77%
4
92% (56%)* (86%)‡
72% (50%)*
58%#
61%#
O
Ph
O
O
O
TsN
CN
OTIPS
Me
CO2Me
CO2H
O
CO2H
CO2H
CO2H
CO2H
CO2H
5
6
61%
7
8
9
57%
10
51%§
65%#
82%#
44%§
O
O
H
OMe
Cl
SCF3
Me
O
N
OH
TsN
O
CO2H
CO2H
CO2H
CO2H
CO2H
CO2H
Me
11
12
13
14
15
16
53%#§
74%#
59%#
56%#
57%†§
51%§
O
• Tertiary alkyl bromides
• Site-selectivity
Me
H
Me
Cl
CO2H
CO2H
CO2H
CO2H
H
H
CO2H
Me
Me
H
17
18
50%¶
19
20
21
54%§
50%#
38%¶
42%§
Figure 2 | Catalytic carboxylation of discrete alkyl halides at remote
sp3 C–H sites. Top, reaction studied, with structure of L1 at right;
bottom, products 1–21. All yields are isolated yields, the average of
at least two independent runs; the variance is estimated to be within 5%.
Ts, p-toluensulfonate; TIPS, triisopropylsilyl; Piv, pivaloyl. Reaction
conditions: NiI2 (2.5 mol%), L1 (4.4 mol%), Mn (3.0 equiv.), CO2 (1 bar),
DMF, 25 °C. *Alkyl tosylate as substrate. ‡10 mmol scale, NiI2 (1 mol%),
L1 (1.80 mol%). §NiI2 (7.5 mol %), L1 (13.2 mol%). ¶NiI2 (10 mol%),
L1 (17.6 mol%). #10 °C. †From the corresponding acetal after hydrolytic
workup.
demonstrated that nickel catalysts markedly improved the efficiency of basis for a chain-walking via iterative β-hydride elimination/migratory
alkyl cross-coupling reactions by minimizing the rate of unproductive insertion sequences3. The resulting nickel intermediates III and V
β-hydride elimination that leads to alkene by-products. These reports formed via II or IV would then enable a final CO2 insertion while
prompted the design of nickel-catalysed reactions of unactivated alkyl ultimately releasing the targeted carboxylic acid. As chain-walking
electrophiles occurring at the initial reaction site, ranging from classical scenarios for forging carbon–carbon bonds remain currently confined
nucleophilic/electrophilic regimes to the coupling of two distinct to ‘unidirectional’ events that result in the activation of a single reaction
electrophiles7, and culminating in stereoconvergent reactions8 or visible- site11,14,19–21 and/or the use of noble expensive metals19,22, our proposed
light photochemical techniques9 (Fig. 1a, left). An emerging strategy switchable selectivity platform based on abundant nickel catalysts could
has been the design of catalytic bond formation at remote reaction unlock a multifaceted challenge for selectively activating less-reactive
sites3,10. However, carbon–carbon bond formation in hydrocarbons positions in the presence of a priori more reactive ones. If such a strategy
is problematic, owing to the presence of multiple, yet similar, sp3 C–H could be implemented, we speculated that valuable fatty acids could
positions. Several methodologies have tackled this issue by activat- be within reach by directly reacting raw materials derived in bulk
ing weak sp3 C–H bonds11 or by using directing groups at a specific from petroleum processing (such as alkanes or unrefined mixtures of
location within the side-chain12–15. As part of our interest in nickel- alkenes) with CO2 without requiring the isolation of the corresponding
catalysed carboxylations16,17, we sought to develop a protocol for reaction intermediates (Fig. 1c). Such a scenario would constitute
incorporating carbon dioxide (CO2) at remote sp3 C–H sites en route a unique platform for converting simple chemical feedstocks into
to fatty acids (Fig. 1a, right), which are relevant in the manufacture valuable compounds.
of soaps, detergents, rubber, plastics and dyes4,18. Indeed, the global
We began our investigations by evaluating a proof-of-principle of
market for carboxylic acids is anticipated to reach approximately our Ni-catalysed remote carboxylation with a discrete alkyl halide
$20 billion by 2023, expanding at an annual growth rate of 5% from (2-bromoheptane) and CO2 (1 bar) at ambient temperature. After
2017 to 20234.
systematically evaluating the reaction parameters, we found that a
A detailed description of our design principle is outlined in Fig. 1b. combination of NiI2 (2.5mol%) and bench-stable L1 (4.4mol%; Fig. 2)
Although retarding β-hydride elimination has long been the goal afforded octanoic acid (1; Fig. 2) in 92% isolated yield as a single regio-
of organometallic chemists when using alkyl (pseudo)halides as isomer using Mn as reductant in DMF (dimethylformamide) at 25°C.
coupling partners (Fig. 1a, left), we questioned whether we could turn a 1,10-Phenanthroline ligands other than L1 possessing less-sterically
to-be-avoided event into a desirable process. Specifically, we envisioned encumbered substituents at C2 or C9, or the absence of aromatic groups
that fine-tuning of the ligand on the nickel catalyst could accelerate the at C4 or C7, resulted in diminished reactivity (see Supplementary
rate of β-hydride elimination from I before CO2 insertion, providing a Information). Control experiments revealed that all of the reaction
4
m A y 2 0 1 7 | V O L 5 4 5 | N A T U R E | 8 5
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

reSeArCH letter
a
Me
Me
Multiple sp3
C–H bonds
CO2H
CO2H
No isolation required
25
59% (from 22)
1
5 mol% NiI2
8.8 mol% L1
Me
65% (from 23)
Me
Br
Br2 (1 equiv.)
MnO2
Single regioisomers
CO2 (1 atm)
Mn, DMF
25 ºC
H
n
H
n
Me
n = 1, 22
n = 3, 23
n = 6, 24
Isomeric mixtures
of bromoalkanes
CO2H
26
38% (from 24)
b
Single regioisomer
Me
Me
Me
No isolation required
Me
Me
27
28
5 mol% NiI2
8.8 mol% L1
Multiple sp3
HBr
C–H bonds
AcOH
CO2 (1 atm)
Mn, DMF
25 ºC
Br
CO2H
H
Me
1
76%
(from 27–29)
Isomeric mixtures
of bromoalkanes
Me
27:28:29 = 1:1:1
Isomeric mixtures of oleꢀns
29
Single regioisomer
No isolation required
5 mol% NiI2
HBr
AcOH
8.8 mol% L1
CO2Et
CO2H
CO2Et
CO2Et
Mn, DMF
10 ºC
Br
H
30
Isomeric mixtures
of bromoalkanes
31
71% (from 30)
Figure 3 | Catalytic carboxylation of feedstock materials by
b, Regioconvergent carboxylation of mixtures of alkenes with carbon
dioxide via a hydrobromination/carboxylation process (top) and use of
pure α-olefin 30 (bottom). Reaction conditions: alkene (1 equiv.), HBr in
AcOH (1 equiv.), then as in Fig. 2.
regioconvergent events. a, Regioconvergent carboxylation of alkanes
(left) with CO2 via a radical bromination/carboxylation process, to obtain
single-regioisomer carboxylic acids (right, boxed). Reaction conditions:
Br2 (1 equiv.), MnO2 (2 equiv.) and alkane (0.20 M), then as in Fig. 2.
parameters were critical for success. With optimal conditions in hand, (see Supplementary Information). This finding provided the basis for
we examined the generality of our transformation by exploring a wide unravelling the preparative potential of this method by designing a
range of discrete secondary alkyl electrophiles (Fig. 2). The reaction unified catalytic strategy by which bulk raw materials derived from
turned out to be widely applicable regardless of the constitutional petroleum processing (such as alkanes or alkenes) could be used as
isomer of bromoheptane used, yielding exclusively 1 in comparable substrates. The collective one-step synthesis of 1, 25 and 26 from their
yields. Notably, the preparation of 1 could be scaled up in 86% yield alkane congeners (Fig. 3a) or from unrefined mixtures of alkenes
on a gram scale at 1mol% Ni. Remarkably, alkyl tosylates could also (Fig. 3b) showcases the potential of a catalytic platform that combines
serve as electrophilic partners. Particularly informative was the chemical feedstocks, demonstrating the synthetic streamlining and
chemoselectivity profile, as esters (4, 7, 12), ketones (5, 10, 18), acetals the rapid production of added-value compounds from inexpensive
(6, 15), sulfonamides (7, 14), nitriles (8), silyl ethers (9), trifluorometh- raw materials. Neither purification nor isolation of the intermediate
ylthiols (11), free hydroxyl groups (16) or heterocycles (13) could all halogenated compounds was necessary, showing the robustness of our
be tolerated, obtaining in all cases linear carboxylic acids. Similarly, protocol. As pure α-olefins are not generally available at an economically
aryl/alkyl chlorides or pivalates do not interfere, leaving ample room viable price, the possibility of using mixtures of olefins from petroleum
for further functionalization via cross-coupling methodologies (4, 14 processing to prepare fatty acids constitutes a powerful alternative to
and 17). No ketone arising from a chain-walking en route to enol-type classical Reppe-type carbonylation techniques with toxic and hazardous
intermediates in 16 was detected14, and perfect linear selectivity was carbon monoxide23. A similar reaction could also be integrated en route
observed in the presence of weak and a priori more reactive benzylic sp3 to 31 (Fig. 3b) that can simply be converted into 1,12-dodecanedioic
C–H bonds within the side-chain (3–6)11. Variable amounts of internal acid (DDA), which is an important component of the synthesis of nylon
olefins and reduced by-products were observed in the carboxylations 12 (ref. 24).
listed in Fig. 2.
The use of 2-bromoheptane-1,1,1-d3 resulted in 1 with substantial
As this protocol is conducted in the absence of base, substrates deuterium incorporation at C2 and C8, suggesting the viability of tar-
possessing relatively acidic protons in the α position relative to geting differently substituted remote sp3 C–H sites (see Supplementary
carbonyl functional groups can be tolerated (5, 7, 8, 18). Although information). Specifically, we found an excellent preference for second-
substrates bearing different primary C(sp3)–H bonds might lead to site- ary sp3 C–H sites at 42°C (32, linear:branched (l:b)=8:92), whereas a
selectivity issues, exclusive carboxylation took place at the less-hindered selectivity switch occurred at 10°C (33, l:b=85:15) (Fig. 4). As shown
primary sp3 C–H site (11). Even single regioisomers could be obtained for 32 and 33, lower linear selectivities were found for substrates
by using sterically hindered tertiary alkyl bromides 19 and 20. Excellent possessing C–Br bonds proximal to the ester motif. These results suggest
site-selectivity could also be accomplished with multiple primary sp3 that regiodivergency arises from a subtle kinetic and thermody-
C–H sites (21).
namic control, forming preferentially an intermediate α-olefin or an
We note that 1 could be selectively obtained in 83% yield from an α,β-unsaturated compound that can be thermally modulated. The
equimolecular mixture of regioisomeric bromoheptanes, showing the generality of this finding could be extended to amides on the side-
viability of implementing regioconvergent carboxylation processes chain, delivering either branched (34 and 36) or linear carboxylic
8 6
| N A T U R E | V O L 5 4 5 | 4 m A y 2 0 1 7
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

letter reSeArCH
a
Remote
sites
H
H
Me
2.5 mol% NiI2
4.4 mol% L1
2.5 mol% NiI2
4.4 mol% L1
m
CO2H
n
m
n
H
Br
H
O
O
O
H
H
CO2H
m
n
CO2 (1 atm)
Mn, DMF
T = 42 ºC
CO2 (1 atm)
Mn, DMF
R
R
R
T = 10 ºC
Branched acid
Linear acid
• Secondary alkyl bromides
H
H
Me
Me
CO2H
T = 42 ºC
T = 10 ºC
R
R
Br
R
CO2H
CO2 (1 atm)
CO2 (1 atm)
O
O
O
R = OMe, 32, 81% (l:b = 8:92)*
R = NPh2, 34, 78% (l:b = 1:99)*
R = NPhMe, 36, 77% (l:b = 1:99)*
R = OMe, 33, 86% (l:b = 85:15)‡
R = NPh2, 35, 84% (l:b = 99:1)‡
R = NPhMe, 37, 74% (l:b = 99:1)‡
H
H
Br
Me
T = 42 ºC
CO2H
Me
T = 10 ºC
CO2 (1 atm)
CO2 (1 atm)
MeO2C
CO2H
32
MeO2C
33
81% yield (l:b = 55:45)‡
MeO2C
78% yield (l:b = 7:93)*
• Long-range remote carboxylation
H
H
Me
Me
T = 10 ºC
T = 42 ºC
CO2H
CO2H
CO2Et
Br
CO2Et
CO2 (1 atm)
CO2 (1 atm)
CO2Et
31
38
80% yield (l:b = 92:8)‡
78% yield (l:b = 6:94)*
• Quaternary centers
H
H
Me
Me
T = 42 ºC
T = 10 ºC
CO2H
CO2H
Me
CO2 (1 atm)
Br
Me
CO2 (1 atm)
MeO2C
MeO2C
Me
MeO2C
dr = 1:1
39
40
63% yield (l:b = 85:15)‡
51% yield (l:b = 1:99)*
• Primary alkyl bromides
H
H
H
H
H
H
T = 50 ºC
T = 25 ºC
CO2H
H
Br
R
R
R
CO2 (1 atm)
CO2 (1 atm)
CO2H
O
O
O
R = OMe, 32, 75% (l:b = 16:84)#
R = NPhMe, 36, 69% (l:b = 1:99)#
R = OMe, 33, 78% (l:b = 99:1)
R = NPhMe, 37, 61% (l:b = 99:1)
b
2.5 mol% NiI2
4.4 mol% L1
CO2 (1 atm)
2
MeO2C
2
MeO2C
CO2Me
T (ºC) C2:C7:C4′ Yield (%) e.r. (42)
7
4
Me
Me
then
35:40:25
100:0:0
70
62
89:11
80:20
25
42
4′ Me Br
Me
TMSCHN2
41
42
d.r. = 1:1 (e.r. = 98:2)
Figure 4 | Switchable site-selective carboxylation of unactivated
C–H remote sites is shown. Reaction conditions as in Fig. 2 at the selected
temperature. b, Retention of configuration of pre-existing stereocentres
on the side-chain. Reaction conditions as in Fig. 2 followed by TMSCH2
treatment. Table at right shows effect of temperature. d.r., diastereomeric
ratio; e.r., enantiomeric ratio. *42 °C. ‡10 °C. #50 °C.
alkyl bromides at remote sp3 C–H sites. All yields are isolated yields,
the average of at least two independent runs; the variance is estimated
to be within 5%. a, Top, reactions studied; bottom, scope of reactants.
Temperature-dependent switchable site-selective carboxylation at sp3
acids (35 and 37). The observed 99:1 site-selectivity for amides at activating remote sp3 C–H sites, resulting in either 32 (l:b=16:84) or
either 10°C or 42°C cannot be simply attributed to electronic effects, 36 (l:b=1:99).
as the pKa of the α-protons of both amides and esters have similar
With an efficient protocol for effecting nickel-catalysed chain-walking
values (pKa =30–32 in DMSO, dimethylsulfoxide). Regiodivergency carboxylation reactions, we wondered whether the inclusion of pre-
could be even accomplished at long-range, a testament to the effi- existing stereogenic centres on the side-chain would be tolerated. To
ciency of our carboxylation event (38, l:b=6:94). A substrate bearing a this end, we conducted the reaction of 41 containing a pre-existing
tertiary sp3 C–H bond could also participate in the reaction, leading stereogenic centre at C4 (enantiomeric ratio e.r.=98:2). Although we
to either 40 (l:b=85:15) or quaternary carbon centres (39, l:b=1:99). found considerable erosion in enantioselectivity at 42°C (e.r.=80:20),
The limits of our regiodivergent strategy were explored with primary substantial preservation of the chiral integrity at C4 was observed
alkyl bromides, and showed that even substrates prone to carbon– at 25°C, leading to 42 after subsequent treatment with TMSCHN2
carbon bond formation before β-hydride elimination25 can be used for (e.r.=89:11). These results indicate that our nickel catalyst remains
4
m A y 2 0 1 7 | V O L 5 4 5 | N A T U R E | 8 7
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

reSeArCH letter
17. Moragas, T., Cornella, J. & Martin, R. Ligand-controlled regiodivergent
Ni-catalyzed reductive carboxylation of allyl esters with CO2. J. Am. Chem. Soc.
136, 17702–17705 (2014).
18. Liu, Q., Wu, L., Jackstell, R. & Beller, M. Using carbon dioxide as a building block
in organic synthesis. Nat. Commun. 6, 5933 (2015).
19. Dupuy, S., Zhang, K.-F., Goutierre, A.-S. & Baudoin, O. Terminal-selective
functionalization of alkyl chains by regioconvergent cross-coupling. Angew.
Chem. Int. Ed. 55, 14793–14797 (2016).
20. Johnson, L. K., Killian, C. M. & Brookhart, M. New Pd(II)- and Ni(II)-based
catalysts for polymerization of ethylene and α-olefins. J. Am. Chem. Soc. 117,
6414–6415 (1995).
21. Bair, J. S. et al. Linear-selective hydroarylation of unactivated terminal and
internal olefins with trifluoromethyl-substituted arenes. J. Am. Chem. Soc. 136,
13098–13101 (2014).
22. Behr, A. et al. Highly selective tandem-isomerization hydroformylation reaction
of trans-4-octene to n-nonanal with rhodium-BIPHEPHOS catalysis. J. Mol.
Catal. Chem. 206, 179–184 (2003).
ligated to the substrate throughout the chain-walking, constitut-
ing a rare example in which pre-existing stereogenic centres are not
substantially altered when displacing a catalyst through an alkyl
chain14,26
.
Online Content Methods, along with any additional Extended Data display items and
Source Data, are available in the online version of the paper; references unique to
these sections appear only in the online paper.
received 27 January; accepted 30 March 2017.
1. Colacot, T. New Trends in Cross-coupling: Theory and Applications (RSC Catalysis
Series, 2015).
2. Kambe, N., Iwasaki, T. & Terao, J. Pd-catalyzed cross-coupling reactions of alkyl
halides. Chem. Soc. Rev. 40, 4937–4947 (2011).
3. Vasseur, A., Bruffaerts, J. & Marek, I. Remote functionalization through alkene
isomerization. Nat. Chem. 8, 209–219 (2016).
4. Szilagyi, M. in Patty’s Toxicology Vol. 3 (eds Bingham, E. & Cohrssen, B.) Ch. 48
(Wiley, 2012).
23. Wu, X.-F. et al. Transition metal-catalyzed carbonylation reactions of olefins and
alkynes: a personal account. Acc. Chem. Res. 47, 1041–1053 (2014).
24. Chou, R. T. & Bendler, H. V. Polyamide composition containing ionomer.
WO patent WO/2013/101891 (2013).
5. Netherton, M. R., Dai, C., Neuschutz, K. & Fu, G. C. Room-temperature
alkyl-alkyl Suzuki cross-coupling of alkyl bromides that possess β hydrogens.
J. Am. Chem. Soc. 123, 10099–10100 (2001).
6. Devasagayaraj, A., Studemann, T. & Knochel, P. A new nickel-catalyzed
cross-coupling reaction between sp3 carbon centers. Angew. Chem. Int. Ed.
Engl. 34, 23–24 (1995).
25. Liu, Y., Cornella, J. & Martin, R. Ni-catalyzed carboxylation of unactivated
primary alkyl bromides and sulfonates with CO2. J. Am. Chem. Soc. 136,
11212–11215 (2014).
26. Singh, S., Bruffaerts, J., Vasseur, A. & Marek, I. A unique Pd-catalysed Heck
arylation as a remote trigger for cyclopropane selective ring-opening.
Nat. Commun. 8, 14200 (2017).
7. Tasker, S. Z., Standley, E. A. & Jamison, T. F. Recent advances in homogeneous
nickel catalysis. Nature 509, 299–309 (2014).
Supplementary Information is available in the online version of the paper.
8. Owston, N. A. & Fu, G. C. Asymmetric alkyl-alkyl cross-couplings of unactivated
secondary alkyl electrophiles: stereoconvergent Suzuki reactions of racemic
acylated halohydrins. J. Am. Chem. Soc. 132, 11908–11909 (2010).
9. Zuo, Z. et al. Merging photoredox with nickel catalysis: coupling of α-carboxyl
sp3-carbons with aryl halides. Science 345, 437–440 (2014).
10. White, M. C. Adding aliphatic C–H bond oxidations to synthesis. Science 335,
807–809 (2012).
Acknowledgements We are grateful for financial support provided by ICIQ, the
European Research Council (ERC-StG-277883 and ERC-2015-PoC-713577),
MINECO (CTQ2015-65496-R and Severo Ochoa Excellence Accreditation
2014-2018, SEV-2013-0319) and the Cellex Foundation. F.J.-H. and J.C.
thank COFUND and Marie Curie Actions for an Intra-European Fellowship
(FP7-PEOPLE-2012-IEF-328381).
11. He, Y., Cai, Y. & Zhu, S. Mild and regioselective benzylic C–H functionalization:
Ni-catalyzed reductive arylation of remote and proximal olefins. J. Am. Chem.
Soc. 139, 1061–1064 (2017).
12. Choi, G. J., Zhu, Q., Miller, D. C., Gu, C. J. & Knowles, R. R. Catalytic alkylation of
remote C–H bonds enabled by proton-coupled electron transfer. Nature 539,
268–271 (2016).
13. Chu, J. C. K. & Rovis, T. Amide-directed photoredox-catalysed C–C bond
formation at unactivated sp3 C–H bonds. Nature 539, 272–275 (2016).
14. Mei, T.-S., Patel, H. H. & Sigman, M. S. Enantioselective construction of remote
quaternary stereocentres. Nature 508, 340–344 (2014).
Author Contributions F.J.-H., T.M. and J.C. performed and analysed the
experiments. R.M. wrote the manuscript. All authors commented on the final
manuscript and contributed to the analysis and interpretation of the results.
Author Information Reprints and permissions information is available at
www.nature.com/reprints. The authors declare competing financial interests:
details are available in the online version of the paper. Readers are welcome to
comment on the online version of the paper. Publisher’s note: Springer Nature
remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations. Correspondence and requests for materials should be
addressed to R.M. (rmartinromo@iciq.es).
15. Vasseur, A., Perrin, L., Eisenstein, O. & Marek, I. Remote functionalization of
hydrocarbons with reversibility enhanced stereocontrol. Chem. Sci. 6,
2770–2776 (2015).
16. Correa, A., León, T. & Martin, R. Ni-catalyzed carboxylation of C(sp2)– and
C(sp3)–O bonds with CO2. J. Am. Chem. Soc. 136, 1062–1069 (2014).
reviewer Information Nature thanks I. Marek and the other anonymous
reviewer(s) for their contribution to the peer review of this work.
8 8
| N A T U R E | V O L 5 4 5 | 4 m A y 2 0 1 7
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.

letter reSeArCH
MethOdS
temperature for 20–48h. The mixture was then carefully quenched with 2M HCl
General carboxylation procedure. An oven-dried Schlenk tube containing a to hydrolyse the resulting carboxylate and extracted with EtOAc (three times). The
stirring bar was charged with NiI2 (2.5–10.0mol%), L1 (4.4–17.6mol%) and Mn combined organic layer was washed with brine (three times), dried over anhydrous
powder (3 equiv.). The Schlenk tube was then evacuated and back-filled under MgSO4, filtrated and evaporated. The resulting crude carboxylic acid was purified
a CO2 flow (this sequence was repeated three times) and finally an atmospheric by conventional flash chromatography in silica gel using hexanes/EtOAc 3:1 with
pressure of CO2 was established. The corresponding alkyl bromide (0.5mmol) 1% formic acid.
and DMF (1M) were added under a CO2 flow. Once added, the Schlenk tube Data availability. The data supporting the findings of this study are available
was closed at atmospheric pressure of CO2 (1 atm) and stirred at the desired within the paper and its Supplementary Information.
© 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.