Unprecedented selective homogeneous cobalt-catalysed reductive alkoxylation of cyclic imides under mild conditions
Abstract
Peer reviewed
Full text
Unp eceden ed selec i e homogeneous
cobal -ca alysed educ i e alkoxyla ion
o cyclic imides unde mild condi ions†
Jose R. Cab e o-An onino, Rosa Adam, Ve onica Papa, Ma es Hols en,
Ka h in Junge and Ma hias Belle *
The fi s gene al and efficien non-noble me al-ca alysed educ i e C2-alkoxyla ion o cyclic imides
(ph halimides and succinimides) is p esen ed. C ucial o he success is he use o [Co(BF
4
)
2
$6H
2
O/
iphos (L1)] combina ion and no ex e nal addi i es a e equi ed. Using he op imal cobal -sys em, he
hyd ogena ion o he a oma ic ing o he pa en ph halimide is a oided and only one o he ca bonyl
g oups is selec i ely unc ionalized. The esul ing p oduc s, N- and a yl- ing subs i u ed 3-alkoxy-2,3-
dihyd o-1H-isoindolin-1-one and N-subs i u ed 3-alkoxy-py olidin-2-one de i a i es, a e p epa ed
unde mild condi ions in good o excellen isola ed yields. In amolecula educ i e couplings can also
be pe o med affo ding icyclic compounds in a one-s ep p ocess. The p esen p o ocol opens he way
o he de elopmen o new base-me al p ocesses o he s aigh o wa d syn hesis o unc ionalized N-
he e ocyclic compounds o pha maceu ical and biological in e es .
In oduc ion
Ca aly ic educ i e ans o ma ions o ca boxylic acid de i a-
i es a e o impo ance and a cu en ho opic in ca alysis.
1
These eac ions a e al eady applied in indus y o he o ma-
ion o bulk p oduc s and in e media es. Mo eo e , hey offe
in e es ing possibili ies o alo iza ion o biomass-de i ed
building blocks and o apply new s a egies in o ganic
syn hesis. Hence, he design o imp o ed ca alys s, and also he
de elopmen o new me hodologies o he selec i e educ ion
o his class o compounds, con inues o a ac he in e es o
academic and indus ial esea che s.
Cyclic imides, and in pa icula ph halimides,
2
a e an
impo an ype o ca boxylic acid de i a i e and se e al o hese
compounds show in e es ing biological ac i i ies. Among he
possible p oduc s ob ained om he educ ion o ph halimides,
isoindolinones and subs i u ed de i a i es a e he mos desi ed
as hey a e aluable scaffolds in pha maceu icals and ag o-
chemicals, as well as ele an building blocks o o ganic
syn hesis (Fig. 1).
3
As a consequence, in he las yea s se e al –oen mul i-s ep
–o ganic me hodologies ha e been epo ed o hei syn he-
sis.
3q,4
Clea ly, he selec i e mono- educ ion o eadily a ailable
ph halimides ep esen s he mos sui able and di ec app oxi-
ma ion o hese compounds. T adi ionally, p ocedu es o his
educ ion equi ed he use o o e -s oichiome ic amoun s o
Zn o Sn in he p esence o s ong acids o o ganome allic
hyd ides (NaBH
4
,B
2
H
6
and LiAlH
4
). Despi e he use ulness o
hese me hods on labo a o y scale, hey ha e d awbacks due o
hei limi ed unc ional g oup ole ance, he gene a ion o o e -
educ ion p oduc s and signican amoun s o was e.
3 , ,4g,k
As a g eene app oach o he educ ion o ph halimides,
hyd ogena ions using he e ogeneous ca alys s ha e been applied
(i.e. RANEY® nickel), al hough hey equi e ha sh eac ion con-
di ions.
1d,
To o e come hese limi a ions, in he las decade also
me hodologies based on molecula ly-dened complexes ha e
been de eloped p oceeding a milde eac ion condi ions (Fig. 2).
Fig. 1 Examples o ele an isoindolinone de i a i es.
Leibniz-Ins i u ¨
u Ka alyse e.V. an de Uni e si ¨
a Ros ock, Albe -Eins ein-S aße
29a, 18059 Ros ock, Ge many. E-mail: [email p o ec ed]e
†Elec onic supplemen a y in o ma ion (ESI) a ailable: Gene al in o ma ion
conce ning expe imen al p ocedu es, addi ional ables, gu es, schemes,
cha ac e iza ion da a and NMR spec a o he isola ed compounds a e a ailable.
See DOI: 10.1039/c7sc01175j
Ci e his: Chem. Sci.,2017,8,5536
Recei ed 15 h Ma ch 2017
Accep ed 5 h June 2017
DOI: 10.1039/c7sc01175j
sc.li/chemical-science
5536 |Chem. Sci.,2017,8, 5536–5546 This jou nal is © The Royal Socie y o Chemis y 2017
Chemical
Science
EDGE ARTICLE
Open Access A icle. Published on 12 June 2017. Downloaded on 1/7/2025 4:22:20 PM.
This a icle is licensed unde a
C ea i e Commons A ibu ion 3.0 Unpo ed Licence.
View A icle Online
View Jou nal
| View Issue
Ae he o iginal epo by Pa on and D ago dealing wi h he
hyd ogena ion o N-me hylsuccinimide wi h a u henium ca a-
lys ,
5
al e na i e p ocedu es we e epo ed by he g oups o B u-
neau,
6
Ika iya,
7
Be gens,
8
Ga c´
ıa,
9
Agbossou-Niede co n,
10
Xie,
11
Zhang
12
and ou g oup.
13
These p o ocols affo ded aluable
p oduc s such as alipha ic lac ams, 2-hyd oxyme hylbenzamides,
u-hyd oxylac ams, benzamides, alipha ic cyclic amines, 1,4-diols
and isoindolinone de i a i es di ec ly om ph halimides.
Despi e all hese ad ancemen s, s ill signican limi a ions exis ,
especially ela ed wi h na ow subs a e scope, he use o
p ecious ca alys s and/o he employmen o hyd osilanes as
educing agen s. Mo eo e , om he poin o iew o ob aining
he desi ed isoindolinone de i a i es, some o hese p o ocols
p esen d awbacks as he concomi an hyd ogena ion o he
a oma ic ing, he lack o selec i i y in he educ ion o one o he
ca bonyl g oups o he occu ence o C–Nbondclea age.
Recen ly, ou g oup de eloped a di ec p o ocol o he
selec i e one-po C2-alkoxyla ion and amina ion o cyclic imides
o gi e 3-subs i u ed-2,3-dihyd o-1H-isoindolinones.
14
In his Ru-ca alysed me hodology, a oma ic ing hyd oge-
na ions we e comple ely a oided and a yl ing-subs i u ed
ph halimides showed selec i e monoalkoxyla ion o one o he
ca bonyl g oups (Fig. 2). C ucial o he ca aly ic ac i i y was he
p esence o me hanosul onic acid (MSA) as an addi i e.
In he las yea s, (1,1,1- is(diphenylphosphinome hyl)-
e hane), so-called iphos, became a p i ileged ligand o he
hyd ogena ion o ca boxylic acids and ela ed de i a i es.
14,15
Basically, in all hese cases, ac i e Ru ca alys s a e gene a ed. In
he las decade, he eplacemen o p ecious me als by inex-
pensi e and widely abundan s - ow base me als such as Fe,
16
Co
16h,i
and Mn
17
has gained inc easing impo ance in hyd oge-
na ion chemis y. Fo example, se e al cobal -based sys ems
ha e shown in e es ing ac i i y o educ ions o C–O,
18
C–
N,
18b,19
C–C
18b,c,20
mul iple bonds and N-he e ocycles.
21
No ably
in 2015, he g oups o de B uin and Else ie
18
achie ed o he
s ime he hyd ogena ion o ca boxylic acids and es e s using
[Co(BF
4
)
2
$6H
2
O/ iphos (L1)]. In addi ion, ou g oup epo ed
he CO
2
hyd ogena ion o me hanol using a modied ela ed
ca alys [Co(acac)
3
/ iphos (L1)/HNT
2
].
18i
Inspi ed by hese wo ks, we en isaged he possibili y o
pe o m he selec i e educ ion o imides using a cobal -based
ca alys sys em. He e, we show o he s ime, a gene al
and efficien me hodology o he non-noble me al-ca alysed
educ i e C2- unc ionaliza ion o cyclic imides (ph halimides
and succinimides).
Resul s and discussion
A he s a o his p ojec he educ i e me hoxyla ion o N-
me hylph halimide 1a using me hanol as sol en was selec ed
as benchma k eac ion (Table 1). Ini ially, he eac ion was
pe o med wi h Co(BF
4
)
2
$6H
2
O using simila condi ions
(150 C, 60 ba H
2
, 18 h) known o Ru ca alys s. Howe e , no
ac i i y was obse ed (Table 1, en y 1). When he same eac ion
was conduc ed in he p esence o 5 mol% o iphos L1,
a quan i a i e yield o 3-me hoxy-2-me hylisoindolin-1-one 2a
was ob ained wi h excellen selec i i y (Table 1, en y 2). G a i-
yingly, no aces o p oduc s coming om educ ion o bo h
ca bonyl g oups o a oma ic ing hyd ogena ion we e obse ed.
Then, he effec o p essu e and empe a u e was e alua ed
in mo e de ail (Table 1, en ies 3–13). To ou deligh , he eac-
ion p oceeds efficien ly a much milde condi ions, and
excellen yields o me hoxyla ed p oduc 2a we e ob ained a
90 C and 20 ba o hyd ogen (Table 1, en y 9). In addi ion, he
ca aly ic sys em also showed high ac i i ies a 70 C, albei
highe p essu es o hyd ogen o ca alys loadings we e equi ed
in hese cases (Table 1, en ies 11–13). To demons a e he need
o hyd ogen, we pe o med he eac ion a 90 C in a p essu e
ube and no con e sion was de ec ed (Table 1, en y 14).
A his poin , he effec o he ela i e amoun s o ligand L1
wi h espec o he cobal p ecu so was in es iga ed in mo e
de ail (Table 1, en ies 16–18). While o 2.5 mol% o cobal p e-
ca alys , wo equi alen s o ligand L1 we e equi ed o pe o m
Fig. 2 (Up) Desc ibed examples o homogeneous ca aly ic educ ion
o ph halimides using hyd ogen o silanes as educ o . (Bo om)
Gene al [Co/ iphos]-ca alysed in e - and in amolecula selec i e
educ i e alkoxyla ion o cyclic imides ( his wo k). (Bn ¼benzyl).
This jou nal is © The Royal Socie y o Chemis y 2017 Chem. Sci.,2017,8, 5536–5546 | 5537
Edge A icle Chemical Science
Open Access A icle. Published on 12 June 2017. Downloaded on 1/7/2025 4:22:20 PM.
This a icle is licensed unde a
C ea i e Commons A ibu ion 3.0 Unpo ed Licence.
View A icle Online
he eac ion efficien ly (Table 1, en ies 9, 16 and 17), a highe
cobal ca alys loadings (5 mol%) only one equi alen o L1 was
enough o an equally efficien me hoxyla ion o N-me hyl-
ph halimide 1a (Table 1, en y 18). Finally, he ca aly ic sys em
also affo ded high yields o p oduc 2a a 1.5 mol%, bu using 30
ba o hyd ogen (Table 1, en y 21).
Nex , he ca aly ic ac i i y o diffe en me al p e-ca alys s was
e alua ed (Table S1†). Among all he diffe en cobal p ecu so s
es ed (Table S1,†en ies 1–12), [Co(BF
4
)
2
$6H
2
O] and
[Co(ClO
4
)
2
$6H
2
O] (Table S1,†en ies 1 and 10) affo ded he
highes ac i i y. In con as , Co(acac)
3
, Co(acac)
2
and also
Ru(acac)
3
in combina ion wi h ligand L1 ( iphos) we e no
ac i e unde he op imal eac ion condi ions (Table S1,†en y 2,
3 and 13, espec i ely). The non-ac i i y o he u henium p e-
ca alys is no wo hy, since [Ru(acac)
3
/L1/MSA] has been
ecen ly desc ibed as an ac i e ca aly ic sys em o he same
eac ion. Appa en ly, o he o ma ion o ou ac i e non-noble
ca alys no ex a acid addi i e is necessa y. Taking in o accoun
he impo an ole o he e auo obo a e anion (
BF
4
)in
p esen sys em, e auo obo a e sal s o coppe (II), i on(II) and
zinc(II) we e also es ed in he benchma k eac ion (Table S1,†
en ies 14–16). Howe e , no ac i i y was de ec ed o any o
hem, indica ing ha cobal is unique o his eac ion.
Rega ding he ligand, we es ed, besides L1, se e al i-
den a e (L2–L5), e aden a e (L6), biden a e (L7–L11) and
monoden a e (L12) ligands (Scheme 1) o he educ i e
me hoxyla ion o N-me hylph halimide (1a). Apa om iphos
(L1), which exhibi ed he bes yield (>99%), only he iden a e
ligand L2 affo ded 2a, albei in low yields (11%).
Ha ing es ablished he [Co(BF
4
)
2
$6H
2
O/ iphos (L1)] sys em
as he bes ca alys , we decided o explo e i s ac i i y o he
educ i e me hoxyla ion o mo e han 20 symme ical
subs i u ed cyclic imides (Table 2). In gene al, all he eac ions
we e conduc ed using 2.5 mol% o cobal p e-ca alys , 5 mol%
o ligand L1 unde 90 C and 20 ba o hyd ogen.
G a i yingly, excellen selec i i y o he monoalkoxyla ion
p oduc was obse ed o all he s udied subs a es. N-Alkyl
subs i u ed ph halimides (1a–1e)affo ded 3-me hoxyla ed iso-
indolinones 2a–2e in e y good isola ed yields (89–97%, Table 2,
en ies 1–5). To s udy he inuence on he ca aly ic ac i i y o
he elec onic cha ac e o he N-subs i uen , diffe en N-a yl o
N-benzyl subs i u ed ph halimides (1 –n) we e es ed wi h he
Table 1 [Co/ iphos (L1)]-ca alysed educ i e me hoxyla ion o N-
me hylph halimide 1a: op imiza ion o he eac ion condi ions
En y
a
T(C) H
2
(ba ) [Co] [L1] Con .
b
(%) 2a
b
(%)
1 150 60 2.5 —— —
2 150 60 2.5 5 >99 >99
3 130 60 2.5 5 >99 >99
4 130 30 2.5 5 >99 >99
5 110 60 2.5 5 >99 >99
6 110 30 2.5 5 >99 >99
7 90 50 2.5 5 >99 >99
8 90 30 2.5 5 >99 >99
9 90 20 2.5 5 >99 >99
10 90 10 2.5 5 86 84
11 70 60 2.5 5 91 90
12 70 40 2.5 5 87 84
13 70 20 5 10 90 89
14
c
90 —2.5 5 ——
15
d
90 20 2.5 5 53 52
16 90 20 2.5 3.75 77 75
17 90 20 2.5 2.5 60 58
18 90 20 5 5 >99 99
19 90 20 1.5 3 93 93
20 90 20 0.5 1 40 38
21 90 30 1.5 3 >99 96
a
S anda d eac ion condi ions: N-me hylph halimide 1a (82.2 mg, 0.5
mmol), Co(BF
4
)
2
$6H
2
O (0.5 o 5 mol%), iphos L1 (1 o 10 mol%), H
2
(10–60 ba ), MeOH (2 mL), 90–150 C and 3–18 h. [Co] ¼
[Co(BF
4
)
2
$6H
2
O] and [L1] in mol% wi h espec o 1a.
b
Con e sion o
1a and yields o 2a we e calcula ed by GC using hexadecane as
in e nal s anda d.
c
The eac ion was ca ied ou wi hou hyd ogen
using a p essu e ube.
d
Run a 3 h.
Scheme 1 Cobal -ca alysed educ i e me hoxyla ion o N-me hyl-
ph halimide (1a): influence o he ligand. S anda d eac ion condi ions:
N-me hylph halimide 1a (82.2 mg, 0.5 mmol), Co(BF
4
)
2
$6H
2
O (4.3 mg,
0.0125 mmol, 2.5 mol%), ligand (0.025 mmol, 5 mol%, 2.5 eq. o Co),
H
2
(20 ba ), MeOH (2 mL), 90 C and 18 h. Yields o p oduc 2a we e
calcula ed by GC using hexadecane as in e nal s anda d.
5538 |Chem. Sci.,2017,8, 5536–5546 This jou nal is © The Royal Socie y o Chemis y 2017
Chemical Science Edge A icle
Open Access A icle. Published on 12 June 2017. Downloaded on 1/7/2025 4:22:20 PM.
This a icle is licensed unde a
C ea i e Commons A ibu ion 3.0 Unpo ed Licence.
View A icle Online
Table 2 [Co/ iphos (L1)]-ca alysed educ i e me hoxyla ion o
diffe en subs i u ed cyclic imides
En y
a
Cyclic imide 1[Co] (mol%) 2
b
[%]
12.5 2a [95]
242b [89]
342c [90]
442d [96]
52.5 2e [97]
62.5 2 [93]
72.5 2g [88]
862h [91]
942i [98]
10
c
2.5 2j [95]
Table 2 (Con d. )
En y
a
Cyclic imide 1[Co] (mol%) 2
b
[%]
11 2.5 2k [96]
12 2.5 2l [94]
13 42m [99]
14 52n [99]
15
c
2.5 2o [86]
16 2.5 2p [89]
17 2.5 2q [81]
18 2.5 2 [89]
19 2.5 2s [94]
20 2.5 2 [85]
This jou nal is © The Royal Socie y o Chemis y 2017 Chem. Sci.,2017,8, 5536–5546 | 5539
Edge A icle Chemical Science
Open Access A icle. Published on 12 June 2017. Downloaded on 1/7/2025 4:22:20 PM.
This a icle is licensed unde a
C ea i e Commons A ibu ion 3.0 Unpo ed Licence.
View A icle Online
cobal sys em. Ph halimides con aining uo o-, i-
uo ome hyl-, me hoxy-, me hyl hio-, hyd oxy- and chlo ide-
subs i u ed a yl ings in o ho-, me a- and pa a-posi ion we e
success ully con e ed affo ding he co esponding me hoxy-
la ed p oduc s 2 –nin high isola ed yields (86–99%, Table 2,
en ies 6–14). No ably, his unusual educ i e ans o ma ion
wo ked well o he uo ina ed ph halimide de i a i e 1o
ob aining he me hoxy compound 2o in high isola ed yield
(86%, Table 2, en y 15). Fo he s ime, he [Co/ iphos]
allowed o his educ i e ans o ma ion o NH-ph halimides.
In ac , he sys em showed an excellen ac i i y o he
me hoxyla ion o 1p affo ding he me hoxy p oduc 2p in
excellen yield (89%, Table 2, en y 16). Finally, diffe en N-
subs i u ed succinimides (1q–u) we e es ed, oo. N-Me hyl, N-
phenyl, and N-benzyl succinimides, we e smoo hly me hoxy-
la ed gi ing he desi ed 3-alkoxy-py olidin-2-one de i a i es
2q–uin good o e y good isola ed yields (80–94%, Table 2,
en ies 17–21). Un o una ely, when N-benzyl-2,3-
py idinedica boximide and N-anisoyl-2-py olidinone, as
examples o he e ocyclic and linea imides espec i ely, we e
subjec ed o he op imized eac ion condi ions no desi ed
p oduc could be ob ained.
Once we had shown he gene ali y o ou p o ocol, we
became in e es ed o s udy he selec i e educ ion o non-
symme ical ph halimides (Scheme 2). These a e mo e chal-
lenging subs a es as he eac i i y o he wo ca bonyl unc ions
migh be simila . Up o da e, he e is only one u henium
ca alys desc ibed,
14
which showed mode a e o good egiose-
lec i i ies in such ans o ma ions (see Fig. 2). The e o e, he
de elopmen o new non-p ecious me al-based s a egies o
selec i ely unc ionalize one o he ca bonyl g oups s ill emains
a challenging ask.
As shown in Scheme 2 he egioselec i e monome hoxyla ion
o unsymme ical a yl ing-subs i u ed ph halimides using he
cobal / iphos sys em p oceeded selec i ely. Mos o he
ph halimides used o his s udy a e no comme cially a ailable
and had o be syn he ized (see ESI o Expe imen al de ails†). To
ou deligh , C4-subs i u ed ph halimides wi h ni ogen-based
elec on-dona ing g oups (3a–c), exhibi ed excellen egiose-
lec i i ies (>33 : 1) o he mono unc ionaliza ion on he C2
ca bonyl g oup (isome A), affo ding he co esponding iso-
indolinones in good yields. On he o he hand, C4-subs i u ed
ph halimides wi h an elec on-wi hd awing g oup such as B
(3d), o an oxygen-based elec on-dona ing g oup like me hoxy
(3e), ga e lowe egioselec i i ies (2.3 : 1 and 2.4 : 1) o he same
ca bonyl g oup han ni ogen-based subs i uen s. Fu he mo e,
excellen isola ed yields we e achie ed o he mix u e o
egioisome p oduc s (4dA +4dB) (89%) as well as o he
egioisome s 4eA and 4eB (70 and 29% yield, espec i ely). The
diffe ence be ween he obse ed egioselec i i ies o an amino
and a me hoxy C4-subs i u ed ph halimide can be explained by
he mo e impo an coo dina ing cha ac e o he ni ogen, ha
can di ec he cobal complex o unc ionalize he C2 ca bon-
yl.
11a
In e es ingly, a C3-uo ine subs i u ed N-phenyl ph hali-
mide (3 )affo ded a good egioselec i i y o he
unc ionaliza ion in he ca bonyl g oup bu a posi ion C7,
hence gi ing isome B. This swi ch in he egioselec i i y could
be exploi ed as a syn he ic ool. Bo h egioisome s (4 A and 4 B)
we e isola ed sepa a ely in 6 and 80% yield, espec i ely. Nex ,
a small amily o C4-a yl subs i u ed N-me hyl ph halimides 3g–
mwas syn he ized (see Scheme S2†) and hei educ i e alkox-
yla ion was s udied. Diffe en subs i uen s such as o-Cl (4h), m-
and p-F (4i–j), p-OMe (4k), p-CF
3
(4l) and p-C(O)OMe (4m) a yl
g oups affo ded mode a e o good egioselec i i ies (>2.2 : 1) o
he ca bonyl Aposi ion, wi h no inuence o hei elec onic
cha ac e . Fo all o hese examples, alkoxyla ed p oduc s 4g–m
we e success ully isola ed in up o 96% yield as a mix u e o
egioisome s (Aand B). Func ional g oups like halogen, e he ,
iuo ome hyl and es e g oups we e ole a ed in he p esence
o his cobal -based sys em.
Fu he mo e, we en isaged he possibili y o pe o m selec-
i e in amolecula educ i e alkoxyla ions. This ou e gi es
s aigh o wa d access o in e es ing building blocks o he
syn hesis o alkaloids and in e media es o he p oduc ion o
a s e eogenic ca bon on he a-posi on o he ni ogen lac-
am.
3c,d,g
Using se e al N-(3-hyd oxyp opyl)ph halimides (5a–e),
i was possible o efficien ly syn hesize hese icyclic
compounds
22
in one-s ep (Scheme 3). In o de o achie e ull
con e sions, he eac ions we e conduc ed unde 20 ba o
hyd ogen in me hanol a 90 o 110 C in he p esence o 2.5–6
mol% ca alys . N-(3-Hyd oxyp opyl)ph halimide 5a wi h no
subs i u ion in he a oma ic ing affo ded cyclic compound 6a
in an excellen isola ed yield (94%). Encou aged by his esul ,
diffe en C4-subs i u ed N-(3-hyd oxyp opyl)ph halimides we e
s udied in o de o explo e he egioselec i i y o he p ocess.
Subs a es wi h an elec on-dona ing g oup in C4 posi ion such
as NH-Ph (5b) o OMe (5c), showed good o excellen egiose-
lec i i ies (8 : 1 and 3.5 : 1, espec i ely) o he a ack o ca bonyl
g oup A. The be e egioselec i i y o he amino subs i u ed
ph halimide 5b in compa ison wi h he me hoxy one 5c can be
also explained by he di ec ing effec o he ni ogen.
11a
Table 2 (Con d. )
En y
a
Cyclic imide 1[Co] (mol%) 2
b
[%]
21 42u [80]
a
S anda d eac ion condi ions: cyclic imide (0.5 mmol), Co(BF
4
)
2
$6H
2
O
(4.25 mg, 0.0125 mmol, 2.5 mol%), iphos L1 (15.6 mg, 0.025 mmol, 5
mol%, 2 eq. o Co), H
2
(20 ba ), MeOH (2 mL), 90 C and 18 h. When he
eac ion was ca ied ou using 4 o 6 mol% o cobal p eca alys , 1.5 eq.
o L1 espec o he me al was added.
b
Isola ed yield o he p oduc ae
pu ica ion by column ch oma og aphy on silica a e gi en be ween
b acke s.
c
Run a 110 C.
5540 |Chem. Sci.,2017,8, 5536–5546 This jou nal is © The Royal Socie y o Chemis y 2017
Chemical Science Edge A icle
Open Access A icle. Published on 12 June 2017. Downloaded on 1/7/2025 4:22:20 PM.
This a icle is licensed unde a
C ea i e Commons A ibu ion 3.0 Unpo ed Licence.
View A icle Online
Regioselec i i y was mode a e owa ds isome Ain he case o
a ph halimide subs i u ed wi h an elec on-wi hd awing g oup
such as F (5d), gi ing p oduc s 6dA and 6dB in 86% yield.
Finally, no egioselec i i y was de ec ed in he in amolecula
alkoxyla ion o he ph halimide 5e, con aining an alkyl
subs i uen in he a oma ic ing. Two egioisome ic posi ions A
and Bwe e eac ed wi h he same selec i i y, affo ding he
mix u e o egioisome s 6eA and 6eB in 87% isola ed yield.
Finally, we decided o in es iga e he gene al applicabili y o
diffe en alcohols in his educ i e unc ionaliza ion, eac ing
ph halimide 1a wi h a wide ange o alcohols unde nea
condi ions (Scheme 4). All he eac ions we e conduc ed unde
he p e iously op imized condi ions o me hanol (2.5 mol%
Co, 5 mol% L1, 20 ba o hyd ogen, 90 C, 18 h) and in specic
cases, highe ca alys loadings we e equi ed o ob ain ull
con e sions o 1a. Bo h alipha ic p ima y alcohols (e hanol, 2-
me hoxye hanol, pen anol, cyclopen aneme hanol) and
seconda y ones (isop opanol, 3-pen anol) affo ded he co e-
sponding 3-alkoxyla ed isoindolinones 7a– wi h excellen iso-
la ed yields (83–95%). In addi ion, benzyl and phene hyl
alcohols also eac ed success ully o gi e he co esponding C3
unc ionalized isoindolinones in e y good yields (89 and 91%,
espec i ely). In conclusion, his cobal -ca alysed ans-
o ma ion allows he s aigh o wa d syn hesis o a a ie y o
unc ionalized isoindolinone de i a i es.
Scheme 2 Regioselec i e [Co/ iphos (L1)]-ca alysed educ i e me hoxyla ion o se e al asymme ical ing-subs i u ed ph halimides. S anda d
eac ion condi ions: ph halimide (0.5 mmol), Co(BF
4
)
2
$6H
2
O(4–6 mol%), iphos L1 (6–9 mol%, 1.5 eq. o Co), H
2
(20 ba ), MeOH (2 mL), 90–
110 C and 18 h. Specific eac ion condi ions o cyclic imides 3a,3e and 3g: Co(BF
4
)
2
$6H
2
O (4 mol%) a 110 C, o cyclic imides 3b–dand 3 :
Co(BF
4
)
2
$6H
2
O (4 mol%) a 90 C and o cyclic imides 3h–m: Co(BF
4
)
2
$6H
2
O (6 mol%) a 110 C. Isola ed yields o he p oduc s a e gi en
be ween b acke s. In pa en heses is shown he ela i e selec i i y o each egioisome Aand B, calcula ed by GC-MS and
1
H-NMR analysis (see
ESI†).
a
The isola ed yield co esponds o he unsepa able mix u e o egioisome s Aand B.
b
Small amoun s (<5%) o p oduc s Aand Bcon aining
es e g oup hyd ogena ed o alcohol we e de ec ed.
This jou nal is © The Royal Socie y o Chemis y 2017 Chem. Sci.,2017,8, 5536–5546 | 5541
Edge A icle Chemical Science
Open Access A icle. Published on 12 June 2017. Downloaded on 1/7/2025 4:22:20 PM.
This a icle is licensed unde a
C ea i e Commons A ibu ion 3.0 Unpo ed Licence.
View A icle Online
To gain insigh in he mechanism o he cobal -ca alysed
eac ion, some kine ic s udies (see Fig. 3 and S1–S4†) we e
pe o med. Fig. 3 (up) shows he yield/ ime kine ic p oles o
he o ma ion o 3-me hoxy-2-me hylisoindolin-1-one 2a in he
educ i e me hoxyla ion o 1a a diffe en hyd ogen p essu es:
(A) 30 ba , (B) 20 ba and (C) 10 ba . No induc ion pe iod was
de ec ed in any expe imen , indica ing ha he ca aly ically
ac i e species can be o med easily. The compa ison o he
diffe en kine ic p oles e eals ha he ini ial a es (
0
),
exp essed as [yield (%) o 2a (min)
1
], dec ease no ably om
30 o 10 ba o hyd ogen (0.7325, 0.4379 and 0.149, espec i ely).
The e o e, he eac ion exhibi s a s ong dependence on he
hyd ogen p essu e indica ing ha he ini ial hyd ogena ion o
he ph halimide 1a o he in e media e hemiaminal 1aI is he
a e limi ing s ep o he o e all p ocess. Thus, he subsequen
me hoxyla ion o 1aI is expec ed o be he as s ep. In o de o
con m hese assump ions, addi ional kine ic expe imen s
using he hemiaminal 1aI as subs a e we e pe o med. Fig. 3
(bo om) shows he yield/ ime kine ic p ole o he educ i e
me hoxyla ion o 1aI unde 20 ba o hyd ogen (see also
Fig. S4†). The ini ial a e o he o ma ion o 2a in his case (
0
¼2) is almos e imes la ge han he one ob ained using
ph halimide 1a as s a ing ma e ial (
0
¼0.4379).
This obse a ion suppo s he me hoxyla ion o 1aI o 2a as
he as s ep, and he hemiaminal 1a as a eal in e media e o
his ans o ma ion.
23
Indeed, addi ional con ol expe imen s
s a ing om 1aI co obo a e his obse a ion (see Scheme S5†).
The eac ion o he hemiaminal 1aI in he p esence o lowe
ca alys loadings (0.5 mol% Co) affo ded good yields o he
me hoxyla ed p oduc 2a. Mo eo e , 2a can be p oduced in
quan i a i e yields (98%) om 1aI wi h ligand- ee [Co(BF
4
)
2
-
$6H
2
O] as ca alys .
4m
Appa en ly, his simple cobal sal is able
o ca alyze he alkoxyla ion p ocess. In e es ingly, when he
same eac ion is pe o med adding ligand L1 and in he absence
o hyd ogen, N-me hylph halimide (1a) was de ec ed in 19%
yield as a by-p oduc coming om he de-hyd ogena ion eac-
ion o 1aI media ed by [Co/L1].
In addi ion, poisoning s udies wi h TEMPO (2,2,6,6-
e ame hylpipe idine-1-oxyl), a adical inhibi o , and TMTU
( e ame hylu ea), a binding poison, we e pe o med (Table
Scheme 3 Syn hesis o icyclic compounds by one-s ep egioselec i e [Co/ iphos (L1)]-ca alysed in amolecula educ i e cycliza ion o
diffe en C4-subs i u ed N-hyd oxyp opyl ph halimides. S anda d eac ion condi ions: ph halimide (0.5 mmol), Co(BF
4
)
2
$6H
2
O (2.5–6 mol%),
iphos L1 (5–9 mol%, 1.5–2 eq. o Co), H
2
(20 ba ), MeOH (2 mL), 90–110 C and 18 h. When he eac ion was ca ied ou using 6 mol% o cobal
p eca alys , 1.5 eq. o ligand L1 espec o he me al was added. Specific eac ion condi ions o cyclic imide 5a: Co(BF
4
)
2
$6H
2
O (2.5 mol%) a
90 C, o cyclic imides 5b and 5d: Co(BF
4
)
2
$6H
2
O (6 mol%) a 90 C and o cyclic imides 5c and 5e: Co(BF
4
)
2
$6H
2
O (6 mol%) a 110 C.
a
Isola ed
yield o he p oduc s a e gi en.
b
The ela i e selec i i y o each cyclic egioisome Aand Bwas calcula ed by GC-MS and
1
H-NMR analysis (see
ESI†).
c
When (R ¼H) compounds Aand Ba e he same.
d
Isola ed yield o he unsepa able mix u e o p oduc s Aand Bis gi en.
Scheme 4 [Co/ iphos]-ca alysed educ i e alkoxyla ion o N-me h-
ylph halimide 1a wi h diffe en alcohols. S anda d eac ion condi ions:
N-me hylph halimide 1a (82.2 mg, 0.5 mmol), Co(BF
4
)
2
$6H
2
O
(4.25 mg, 0.0125 mmol, 2.5 mol%), iphos L1 (16.7 mg, 0.025 mmol, 5
mol%, 2 eq. o Co), H
2
(20 ba ), alcohol (2 mL), 90 C and 18 h. Isola ed
yields o he p oduc s a e gi en. [a] Run wi h Co(BF
4
)
2
$6H
2
O (4 mol%)
and iphos L1 (6 mol%). [b] Run wi h Co(BF
4
)
2
$6H
2
O (6 mol%) and
iphos L1 (9 mol%).
5542 |Chem. Sci.,2017,8, 5536–5546 This jou nal is © The Royal Socie y o Chemis y 2017
Chemical Science Edge A icle
Open Access A icle. Published on 12 June 2017. Downloaded on 1/7/2025 4:22:20 PM.
This a icle is licensed unde a
C ea i e Commons A ibu ion 3.0 Unpo ed Licence.
View A icle Online
S2†).
18
In he case o TEMPO he educ i e me hoxyla ion p o-
ceeded success ully, indica ing ha no adicals a e in ol ed in
he mechanism. In con as , TMTU caused a comple e inhibi-
ion o he eac ion wi h only hal equi alen wi h espec o he
ca alys . This effec , p e iously obse ed o he ca boxylic acids
hyd ogena ion wi h he same ca alys ,
18
can be explained by
ei he a dinuclea cobal species o TMTU ac ing as a b idging
poison.
Wi h he aim o unde s anding in mo e de ail he ca aly ic
sys em, high- esolu ion elec osp ay ioniza ion mass spec-
ome y (HR ESI-MS) expe imen s we e pe o med (Fig. S5–
S9†). A sho eac ion imes o 0.5 and 2 hou s a signal a m/z
728.156 was de ec ed, consis en wi h [Co(L1)(COO
)]
+
. The
o ma ion o his species is jus ied by he use o MeOH/0.1%
HCOOH mix u e as sol en o he ESI Expe imen . In ac , in
samples con aining a simple mix u e o Co(BF
4
)
2
$6H
2
O/L1 and
Co(BF
4
)
2
$6H
2
O/L1/2a, he same peak was de ec ed. Any o hese
es s showed a cobal species coo dina ed o N-me hyl-
ph halimide 1a. In e es ingly, a species wi h m/z1309.404 co -
esponding o [Co(L1)
2
(H)
2
]
+
could be de ec ed as an impo an
peak in samples a sho eac ion imes. Howe e , in a sample
aken a he end o he eac ion, his species becomes less
impo an and, using ace oni ile as ESI-MS sol en , a new peak
a m/z901.245 appea s. The la e signal is consis en wi h
[Co(L1)(CH
3
CN)(2a)]
+
, a possible es ing s a e con aining he
me hoxyla ed p oduc 2a.
Based on all hese obse a ions, a plausible mechanism o
he [Co/ iphos (L1)]-ca alysed educ i e alkoxyla ion o cyclic
imides is depic ed in Fig. 4. The [Co(L1)
2
(H)
2
]
+
species, de ec ed
by ESI-MS, is p oposed as he ac i e ca alys o he hyd ogena-
ion o 1a o he hemiaminal in e media e 1aI. A he end o he
eac ion, [Co(L1)(CH
3
CN)(2a)]
+
is de ec ed as he es ing s a e.
Conclusions
In conclusion, a gene al and efficien cobal -ca alysed educ i e
alkoxyla ion o cyclic imides was p esen ed o he s ime. This
g een p o ocol a oids he use o s oichiome ic amoun s o silanes
o me al hyd ides. Hyd ogena ion o he a oma ic ing o he
ph halimide co e does no ake place and excellen chemo-
selec i i y o hemono-alkoxyla ionp oduc sisob ained.Awide
ange o ph halimides/succinimides a e selec i ely unc ionalized
unde mild condi ions. No ably, he [Co/ iphos] sys em is ac i e
wi hou he need o any acid addi i e o gi e 3-alkoxy-2,3-dihyd o-
1H-isoindolin-1-one and 3-alkoxy-py olidin-2-one de i a i es in
high isola ed yields. Fu he mo e, his cobal based ca alys allows
he selec i e unc ionaliza ion o one o he ca bonyl g oups in
non-symme ical a yl ing-subs i u ed ph halimides. Addi ionally,
he eac ioncanbepe o medinanin amolecula ashion,
gi ing N,O-ace al icyclic compounds in one-s ep wi h high yields.
Kine ic in es iga ions e ealed ha he ini ial hyd ogena ion o
he ph halimide o he hemiaminal in e media e is he a e
limi ing s ep o he o e all p ocess. This no el base me al p o ocol
opens a doo o he de elopmen o en i onmen ally-benign
p ocesses o he selec i e syn hesis o unc ionalized N-he e o-
cyclic compounds.
Fig. 3 (Up) Yield/ ime kine ic p ofile o he o ma ion o p oduc 2a in
he educ i e me hoxyla ion o N-me hylph halimide 1a using me h-
anol a 90 C unde diffe en p essu es o molecula hyd ogen: (A) 30
ba , (B) 20 ba and (C) 10 ba . (Bo om) Yield/ ime kine ic p ofile o he
o ma ion o p oduc 2a om he in e media e hemiaminal 1aI using
me hanol and molecula hyd ogen (20 ba ) a 90 C. Inse s co espond
o he ini ial a e plo s whe e (
0
) is he slope o he linea equa ion:
[yield (%) ¼
0
ime (min)] defined a ini ial eac ion imes and
exp essed as [yield (%) o p oduc (min)
1
]. S anda d eac ion
condi ions: subs a e 1a o 1aI (3.0 mmol), Co(BF
4
)
2
$6H
2
O (25.5 mg,
0.075 mmol, 2.5 mol%), iphos L1 (100.5 mg, 0.15 mmol, 5 mol%, 2 eq.
o Co), MeOH (12.0 mL) and H
2
(10, 20 o 30 ba ) a 90 C. Yields o
p oduc 2a we e calcula ed by GC using hexadecane as in e nal
s anda d. Ve ical e o ba (5%) o all da a poin s is shown.
Fig. 4 Possible eac ion mechanism o he [Co/L1]-ca alysed
educ i e me hoxyla ion o cyclic imides.
This jou nal is © The Royal Socie y o Chemis y 2017 Chem. Sci.,2017,8, 5536–5546 | 5543
Edge A icle Chemical Science
Open Access A icle. Published on 12 June 2017. Downloaded on 1/7/2025 4:22:20 PM.
This a icle is licensed unde a
C ea i e Commons A ibu ion 3.0 Unpo ed Licence.
View A icle Online
Expe imen al de ails
Gene al p ocedu e o he educ i e me hoxyla ion o N-
me hylph halimide (1a)
A 4 mL glass ial con aining a s i ing ba was sequen ially
cha ged wi h N-me hylph halimide 1a (82.2 mg, 0.5 mmol),
Co(BF
4
)
2
$6H
2
O (4.25 mg, 0.0125 mmol, 2.5 mol%), iphos L1
(16.75 mg, 0.025 mmol, 5 mol%, 2.5 eq. o Co), n-hexadecane
(50.0 mg) as an in e nal s anda d and MeOH (2.0 mL) as
sol en . Ae wa ds, he eac ion ial was capped wi h a sep um
equipped wi h a sy inge and se in he alloy pla e, which was
hen placed in o a 300 mL au ocla e. Once sealed, he au ocla e
was pu ged h ee imes wi h 30 ba o hyd ogen, hen p essu -
ized o 20 ba and placed in o an aluminium block, which was
p ehea ed a 90 C. Ae 18 h, he au ocla e was cooled in an ice
ba h, and he emaining gas was ca e ully eleased. Finally,
he eac ion mix u e was dilu ed wi h e hyl ace a e and analysed
by GC.
Gene al p ocedu e o he educ i e alkoxyla ion o cyclic
imides
A 4 mL glass ial con aining a s i ing ba was sequen ially
cha ged wi h cyclic imide (0.5 mmol), Co(BF
4
)
2
$6H
2
O (2.5–6
mol%), iphos L1 (5–9 mol%, 1.5–2 eq. o Co) and alcohol (2.0
mL) as sol en . Ae wa ds, he eac ion ial was capped wi h
a sep um equipped wi h a sy inge and se in he alloy pla e,
which was hen placed in o a 300 mL au ocla e. Once sealed,
he au ocla e was pu ged h ee imes wi h 30 ba o hyd ogen,
hen p essu ized o 20 ba and placed in o an aluminium block,
which was p ehea ed a 90–130 C. Ae 18 h, he au ocla e was
cooled in an ice ba h, and he emaining gas was ca e ully
eleased. Finally, he eac ion mix u e was dilu ed wi h e hyl
ace a e and pu ied by silica gel column ch oma og aphy (n-
hep ane/e hyl ace a e mix u es) ob aining he desi ed alkoxy-
la ed de i a i es.
Gene al p ocedu e o he kine ic s udies
A 100 mL glass inle con aining a s i ing ba was sequen ially
cha ged wi h he co esponding subs a e 1a o 1aI (3.0 mmol),
Co(BF
4
)
2
$6H
2
O (25.5 mg, 0.075 mmol, 2.5 mol%), iphos L1
(100.5 mg, 0.15 mmol, 5 mol%, 2.5 eq. o Co), n-hexadecane
(250.0 mg) as an in e nal s anda d and MeOH (12.0 mL) as
sol en . Ae wa ds, he eac ion inle was hen placed in o
a 100 mL au ocla e. Once sealed, he au ocla e was pu ged
h ee imes wi h 30 ba o hyd ogen, hen p essu ized o 10, 20
o 30 ba and placed in o an aluminium block, which was
p ehea ed a 90 C. Pe iodically, aliquo s o 200 mL we e aken
a diffe en imes o eac ion, dilu ed wi h e hyl ace a e and
analysed by GC.
Acknowledgemen s
This wo k was suppo ed by he s a e o Mecklenbu g-
Vo pomme n and he BMBF. J. R. C.-A. and R. A. hanks
Ramon A eces Founda ion o a pos doc o al ellowship.
No es and e e ences
1(a) A. Kleemann, J. Engel, B. Ku schne and D. Reiche ,
Pha maceu ical Subs ances: Syn heses, Pa en s, Applica ions,
Thieme, S u ga , New Yo k, 5 h edn, 2009; (b) P. A. Dub
and T. Ika iya, ACS Ca al., 2012, 2, 1718–1741; (c)
H. A. Wi coff, B. G. Reuben and J. S. Plo kin, Indus ial
O ganic Chemicals, Wiley-In e science, New Yo k, 3 d edn,
2013; (d) D. Dodds and D. J. Cole-Hamil on, in Sus ainable
Ca alysis: Challenges and p ac ices o he Pha maceu ical
and Fine Chemical Indus ies, ed. P. J. Dunn, K. K. Hii, M. J.
K ische and M. T. Williams, John Wiley and Sons,
Hoboken, NJ, 2013, pp. 1–36; (e) S. We kmeis e , K. Junge
and M. Belle , O g. P ocess Res. De ., 2014, 18, 289–302; ( )
A. M. Smi h and R. Whyman, Chem. Re ., 2014, 114, 5477–
5510; (g) G. E. A no , in Comp ehensi e O ganic Syn hesis,
ed. P. Knochel and G. A. Molande , Else ie B.V., 2nd edn,
2014, ol. 8, pp. 368–409.
2 Fo selec ed examples, see: (a) H. Miyachi, A. Azuma,
A. Ogasawa a, E. Uchimu a, N. Wa anabe, Y. Kobayashi,
F. Ka o, M. Ka o and Y. Hashimo o, J. Med. Chem., 1997,
40, 2858–2865; (b) K.-E. Kil, P. Pou iainen, Z. Zhang,
A. Zhu, J.-K. Choi, K. Joki a si and A.-L. B ownell, J. Med.
Chem., 2014, 57, 9130–9138; (c) P. M. U. Ung, J. B. Dunba
J , J. E. Ges wicki and H. A. Ca lson, J. Med. Chem., 2014,
57, 6468–6478; (d) M. A. Bha , M. A. Al-Oma , M. A. Ansa i,
K. M. A. Zohei , F. Imam, S. M. A ia, S. A. Bakhee ,
A. Nadeem, H. M. Ko ashy, A. Vo onko , V. Be ish ili and
S. F. Ahmad, J. Med. Chem., 2015, 58, 8850–8867.
3 Fo selec ed examples, see: (a) V. G. S ygles, J. F. New on and
J. B. Hook, Res. Commun. Chem. Pa hol. Pha macol., 1977, 18,
329–340; (b) E. J. Co ey and M. M. Meh o a, Te ahed on
Le ., 1988, 29,57–60; (c) D. Romo and A. I. Meye s,
Te ahed on, 1991, 47, 9503–9569; (d) A. A. Bahajaj,
P. D. Bailey, M. H. Moo e, K. M. Mo gan and J. M. Ve non,
J. Chem. Soc., Chem. Commun., 1994, 2511–2512; (e)
J. T. Link, S. Ragha an and S. J. Danishe sky, J. Am. Chem.
Soc., 1995, 117, 552–553; ( ) M. H. No man, D. J. Minick
and G. C. Rigdon, J. Med. Chem., 1996, 39, 149–157; (g)
J. M. And ´
es, I. He ´
aiz-Sie a, R. Ped osa and A. P´
e ez-
Encabo, Eu . J. O g. Chem., 2000, 1719–1726; (h)
U. C. Yoon, Y. X. Jin, S. W. Oh, C. H. Pa k, J. H. Pa k,
C. F. Campana, X. Cai, E. N. Duesle and P. S. Ma iano, J.
Am. Chem. Soc., 2003, 125, 10664–10671; (i)
B. E. Ma yanoff, H.-C. Zhang, J. H. Cohen, I. J. Tu chi and
C. A. Ma yanoff,Chem. Re ., 2004, 104, 1431–1628; (j)
J. Roye , M. Bonin and L. Micouin, Chem. Re ., 2004, 104,
2311–2352; (k) I. R. Ha dcas le, S. U. Ahmed, H. A kins,
A. H. Cal e , N. J. Cu in, G. Fa nie, B. T. Golding,
R. J. G iffin, S. Guyenne, C. Hu on, P. Kaellblad,
S. J. Kemp, M. S. Ki ching, D. R. Newell, S. No bedo,
J. S. No hen, R. J. Reid, K. Sa a anan, H. M. G. Willems
and J. Lunec, Bioo g. Med. Chem. Le ., 2005, 15, 1515–1520;
(l) I. R. Ha dcas le, S. U. Ahmed, H. A kins, G. Fa nie,
B. T. Golding, R. J. G iffin, S. Guyenne, C. Hu on,
P. Kaellblad, S. J. Kemp, M. S. Ki ching, D. R. Newell,
5544 |Chem. Sci.,2017,8, 5536–5546 This jou nal is © The Royal Socie y o Chemis y 2017
Chemical Science Edge A icle
Open Access A icle. Published on 12 June 2017. Downloaded on 1/7/2025 4:22:20 PM.
This a icle is licensed unde a
C ea i e Commons A ibu ion 3.0 Unpo ed Licence.
View A icle Online