Sterically-controlled regioselective para-substitutions of aniline

Article abstract of DOI:10.1039/b506824j

Introduction of statically demanding 1-isopropyl-2-methyl-propyl or triisopropylsilyl groups at the nitrogen of aniline allows high-yielding regioselective para-substitution to be achieved using a lithiation/substitution sequence. The Royal Society of Che

Full text of DOI:10.1039/b506824j

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Sterically-controlled regioselective para-substitutions of aniline{
Philip W. Dyer,*ab John Fawcett,b Gerald A. Griffith,b Martin J. Hanton,bc Ce´line Olivier,b
Adam R. Pattersona and Samuel Suharda
Received (in Cambridge, UK) 13th May 2005, Accepted 27th May 2005
First published as an Advance Article on the web 24th June 2005
DOI: 10.1039/b506824j
Introduction of sterically demanding 1-isopropyl-2-methyl-
propyl or triisopropylsilyl groups at the nitrogen of aniline
allows high-yielding regioselective para-substitution to be
achieved using a lithiation/substitution sequence.
Regiospecific preparation of polysubstituted aromatics has been a
long-standing challenge in synthetic chemistry. The directed ortho-
lithiation/electrophilic substitution sequence, which relies on
heteroatom (O, N, halogen) directing effects, has become one of
the most versatile methodologies, and has grown to be routine in
many organic syntheses.1–5 In contrast, there are few examples of
directed para-lithiation, this regioisomer usually being favoured by
the prior introduction of substituents in the ortho and meta ring
positions or by employing strongly para-directing CF3 groups.6–8
Here it is reported that the introduction of bulky 1-isopropyl-2-
methylpropyl or triisopropylsilyl (TIPS) substituents at the
nitrogen of aniline results in regioselective electrophilic ring
substitution by soft electrophiles in the para-position, following
Scheme 1 Reaction conditions: i. BunLi, Et2O, 278 uC to RT, 1 h; ii.
initial N-lithiation.
E1Cl, Et2O, 278 uC to RT, 18 h; iii. E2Cl, Et2O, 278 uC to RT, 12 h; iv.
TMEDA.
Treatment of an ethereal solution of (1-isopropyl-2-methyl-
propyl)-phenyl-amine, 1,9 with 1 equiv. of BunLi, followed by the
addition of a range of soft electrophiles E1Cl (E1 5 (Pri2N)2P,
Ph2P, Me3Si, Me3Sn) affords the para-substituted secondary
anilines 2a–d in excellent isolated yields (Scheme 1).{ Subsequent
addition of a second equivalent of BunLi and electrophile leads to
the formation of the corresponding N,N-difunctionalised para-
substituted anilines 3a and 3b in good yields. The symmetrical
derivative, 3a (89%), can also be prepared in a ‘one-pot’ procedure,
by treating 1 with 2 equiv. of BunLi and then 2 equiv. of
chlorophosphine. No reaction was observed between 1 and a range
of soft electrophiles in the presence of pyridine or NEt3.
To probe the course of these reactions, identifying any
intermediate lithiated species involved was of interest. Thus,
lithiation of 1 was undertaken as above, followed by addition of
1 equiv. of N,N,N9,N9-tetramethylethylenediamine (TMEDA).
This afforded TMEDA-bound lithium amide, 4 (Scheme 1), which
was isolated as air, moisture and light sensitive yellow platelets
(73%) and fully characterised (Fig. 1).§ Treating an isolated sample
aDepartment of Chemistry, University of Durham, South Road, Durham,
UK DH1 3LE. E-mail: p.w.dyer@durham.ac.uk;
Fax: +44 (0) 191 384 4737; Tel: +44 (0) 191 334 2150
bDepartment of Chemistry, University of Leicester, University Road,
Leicester, UK LE1 7RH
Fig. 1 Molecular structure of 4 (H-atoms omitted). Selected bond
cSasol Technology UK, Purdie Building, North Haugh, St. Andrews,
lengths (s) and angles (u): Li(1)–N(1) 1.910(6), Li(1)–N(3) 2.033(6), Li(1)–
Fife, UK KY16 9ST
N(2) 2.077(6), N(1)–C(1) 1.344(3), N(1)–C(7) 1.459(3), N(1)–Li(1)–N(3)
{ Electronic Supplementary Information (ESI) available: Synthetic
procedures and characterisation data for all compounds. See http://
dx.doi.org/10.1039/b506824j
127.9(3), N(1)–Li(1)–N(2) 137.1(3), N(3)–Li(1)–N(2) 89.1(2), C(1)–N(1)–
C(7) 119.4(2), C(1)–N(1)–Li(1) 118.4(2), C(7)–N(1)–Li(1) 122.2(3).
This journal is ß The Royal Society of Chemistry 2005
Chem. Commun., 2005, 3835–3837 | 3835

View Article Online
of 4 with E1Cl generates para-substituted aniline 2 in near
quantitative yield.
In contrast to the reactions of lithiated 1 with soft electrophiles,
the reaction of 1 itself with BunLi, followed by treatment with the
hard electrophiles CO2, acetone or benzaldehyde, led to 1 being
recovered in .95% yield in each case, following a H2O quench.
Similarly, attempts to introduce deuterium onto the ring of 1 by
sequential reaction with BunLi and either D2O or CH3CO2D were
unsuccessful, resulting only in N-deuteration. The observed
reactivity of 1 with soft electrophiles is consistent with that
expected of a soft, C-based nucleophile.
In order to make this overall substitution procedure more
attractive from a general synthetic standpoint, it was of interest to
try and incorporate a readily cleavable group at nitrogen in the
place of the 1-isopropyl-2-methylpropyl substituent, and hence
provide access to versatile para-substituted primary anilines.
Although silyl groups are rarely used for the protection of amines
due to the reactivity of the N–Si linkage, it was exactly this feature,
combined with their ease of introduction, significant steric
demands and commercial availability that led us to explore the
use of the TIPS group in this role. Reaction of the sterically-
encumbered N-TIPS aniline, 5,10 with BunLi and (Pri2N)2PCl
under identical conditions to those used for the preparation of 2a,
afforded a mixture of the para-phosphino-, 6, and N-substituted, 7,
derivatives in a 3 : 1 ratio (Scheme 2). Subsequent reaction of this
mixture of 6 and 7 with excess HCl/Et2O achieved simultaneous
P–N and N–Si bond cleavage, generating the ammonium salt 8,
Pri2NH2Cl, TIPS-Cl and anilinium chloride. The disubstituted
compound, 9, could be isolated in good yield (69%) from a
procedure similar to that used to prepare 3a. Although the
regioselectivity of the reactions of 5 is lower than those of 1, these
experiments demonstrate the feasibility of using readily cleavable,
bulky silyl groups to effect para-substitution in anilines.
Scheme 3 Reaction conditions: i. BunLi, Et2O, 278 uC to RT, 1 h; ii.
Ph2PCl, Et2O, 278 uC to RT; iii. (Pri2N)2PCl, Et2O, 278 uC to RT.
Scheme 4 Proposed mechanism for the formation of para-substituted
anilines 2 and 6 (B–H 5 base).
When the para-position of the ring of 1 was blocked, e.g. in 11,12
initial lithiation (BunLi, Et2O) followed by reaction with Ph2PCl
resulted in a complex, largely intractable mixture of products
according to 31P NMR spectroscopy, which contained
N-phosphinoaniline 15 (31P NMR d 5 +54.8 ppm). Treatment
of the crude reaction mixture with excess S8, followed by GC–MS
analysis, revealed the thio derivative of 15 to be present as ca. 30%
of the total products.
To further explore the substitution chemistry of 1, two sets of
reactions were undertaken (Scheme 3). N-methyl aniline, 10, was
treated sequentially with 1 equiv. of BunLi, followed by Ph2PCl,
under conditions analogous to those used for the preparation of
2b. This quantitatively afforded the known aminophosphine, 12,
according to 31P NMR spectroscopy (d 5 56.8 ppm).11
Collectively, these results are consistent with para-substituted
compounds 2 and 6 being formed from secondary anilines 1 and 5
respectively, in a three-step process which may be regarded as an
aza analogue of the dienone–phenol reaction (Scheme 4).13 In this
process initial N-lithiation occurs, rendering the para-position of
the ring more nucleophilic, facilitating the subsequent directed
electrophilic attack of E1–Cl. The resulting cyclohexadienyl-idene-
amines, 16 and 17, re-aromatise following deprotonation, to afford
2 and 6 respectively. Given the significant driving force associated
with re-aromatisation, it is difficult to identify the base involved in
this final step with any certainty. It is possible that residual lithium
amide may be transferring a proton catalytically or indeed that 16
and 17 are acting as internal bases.
The role of the bulky substituents, which appear essential for the
clean para-functionalisation rather than the N-substitution of 1
and 5, is somewhat unclear. However, since the secondary lithium
amides that result from the reaction of 1 and 5 with BunLi are
extremely sterically hindered, it is presumed that their rate of
Scheme 2 Reaction conditions: i. BunLi, Et2O, 278 uC to RT, 1 h; ii.
(Pri2N)2PCl, Et2O, 278 uC to RT, 18 h; iii. HCl, Et2O, RT; iv. 2 equiv.
BunLi, Et2O, 278 uC to RT, 1 h; v. 2 equiv. (Pri2N)2PCl, Et2O, 278 uC to
RT, 18 h.
3836 | Chem. Commun., 2005, 3835–3837
This journal is ß The Royal Society of Chemistry 2005

View Article Online
nucleophilic substitution with E1Cl is, as a result, retarded. Hence,
the rearrangement outlined in Scheme 4 is favoured, giving rise to
the observed reactivity in the para-position. However, when the
ring para-position is blocked, as in 11, a complex mixture of
products results, which includes the N-substituted compound 15.
In contrast, the amide resulting from N-methyl aniline is
considerably less hindered than that from 1 and thus favours
reaction at nitrogen. Consistent with this idea is the reaction of
lithiated 10 with the bulkier electrophile (Pri2N)2PCl, which gives
rise to a mixture of products, including the N- and ring-
functionalised compounds 13 and 14 respectively (identified
by 31P NMR spectroscopy and GC–MS analysis following
thiolation).
Notes and references
{ Representative synthesis of 2a: To a solution of 1 (5.12 g, 2.68 6
1022 mol) in Et2O (40 mL) at 278 uC was added BunLi (1.6 M, hexane,
17.6 mL, 2.81 6 1022 mol) dropwise, the resulting mixture then being
allowed to slowly warm to RT with stirring over 1 h. The re-cooled
(278 uC) solution was added to a suspension of (Pri2N)2PCl (7.15 g, 2.68 6
1022 mol) in Et2O (40 mL) at 278 uC. Following reaction at RT for 18 h,
removal of volatiles in vacuo and extraction with hexane, prolonged cooling
(230 uC) afforded 2a as white crystals (5.61 g, 96%). Found: C, 71.14; H,
11.59; N, 9.95. C25H48N3P requires C, 71.21; H, 11.47; N, 9.97%; dH
(250.13 MHz, CDCl3) 7.98 (2 H, dd, 3JHH 8.7, 3JPH 6.7 Hz, m-C6H4), 6.53
(2 H, dd, 3JHH 8.7, 4JPH 1.9 Hz, o-C6H4), 3.42 (4 H, d sept., 3JHH 6.7, 3JPH
3
2.5 Hz, PNCH), 2.99 (1 H, br d, JHH 10.1 Hz, NH), 2.84 (1 H, m,
CNHCH), 1.58 (2 H, overlapping sept., 3JHH 6.4 Hz, CCH), 1.29 (12 H, d,
3JHH 7.0 Hz, NCH(CH3)2), 1.27 (12 H, d, 3JHH 7.0 Hz, NCH(CH3)2), 0.83
(6 H, d, 3JHH 6.7 Hz, CH(CH3)2), 0.76 (6 H, d, 3JHH 6.7 Hz, CH(CH3)2);
dC{1H} (62.90 MHz, CDCl3) 150.1 (s, ipso-C6H4), 133.1 (d, 1JPC 21.9 Hz,
Additionally, bulky substituents located at nitrogen will
effectively ‘block’ any reaction at the nearby ring ortho-positions.
The origin of the lower regioselectivity engendered by the TIPS
group compared to the 1-isopropyl-2-methylpropyl moiety
remains unresolved. However, the greater length of N–Si and
C–Si bonds relative to those of N–C and C–C may mean that the
steric constraints imposed by the TIPS unit are somewhat lower
than those of the more successful hydrocarbon moiety, although
the fact that electronic factors may also play a role cannot be
ignored.
2
ipso-C6H4), 131.1 (s, C6H4), 112.9 (d, JPC 5.6 Hz, C6H4), 64.4 (s, NCH),
48.2 (d, 2JPC 11.7 Hz, NCH), 31.8 (s, CH), 25.1 (d, 3JPC 7.6 Hz, NCH), 25.0
3
(d, JPC 7.1 Hz, NCH), 21.3 (s, CH(CH3)2), 18.4 (s, CH(CH3)2); dP{1H}
(101.26 MHz, CDCl3) 59.4 (s); m/z (EI): 421 (M+).
§ Crystal data for 4: C19H36LiN3, M 5 313.45, orthorhombic, space group
P2(1)2(1)2(1), a 5 9.1878(18), b 5 14.344(3), c 5 15.575(3) s,
V 5 2052.7(7) s3, Z 5 4, m(Mo-Ka) 5 0.059 mm21, T 5 150(2) K,
crystal size 0.26 6 0.25 6 0.15 mm, 13076 reflections collected, 3619
independent reflections (Rint 5 0.0849), R1 5 0.0563 [I . 2s (I)],
wR2 5 0.1090, R1 5 0.0976, wR2 5 0.1297 for all data. CCDC 236521.
See http://dx.doi.org/10.1039/b506824j for crystallographic files in CIF
format.
Recently, the regioselective para-bromination of certain primary
anilines has been achieved with varying degrees of success by using
a ‘one-pot’ procedure, involving the sequential treatment of the
amine with equimolar quantities of BunLi, Me3SnCl and elemental
bromine.14 It was proposed that this reaction proceeded via
formation of a tin amide (through a salt elimination reaction
between the lithium amide and Me3SnCl), which activated the
para-position of the aromatic ring towards direct electrophilic
bromination. In light of the results presented here, this pathway
was to be expected, rather than direct stannylation as for 2d, since
no sterically demanding groups were present at nitrogen.
Overall, the approach of using bulky N-substituents to direct
regioselective substitution affords a ready means of preparing a
range of variously-functionalised, para-substituted anilines. In
particular, aromatic amines bearing the Me3Sn moiety are
accessible—compounds that are of potential utility in Stille
cross-coupling reactions. Work is ongoing to enhance the
regioselectivity of the reaction by involving a sterically demanding,
yet readily cleavable group at nitrogen.
1 A. E. H. Wheatley, Eur. J. Inorg. Chem., 2003, 3291.
2 G. Mugesh and H. B. Singh, Acc. Chem. Res., 2002, 35, 226.
3 J. Clayden, Organolithiums: Selectivity for Synthesis, Pergamon,
Amsterdam, 2002.
4 P. Beak and V. Snieckus, Acc. Chem. Res., 1985, 15, 306.
5 V. Snieckus, Bull. Soc. Chim. Fr., 1988, 67; V. Snieckus, Chem. Rev.,
1990, 90, 879; N. J. R. van Eikema Hommes and P. von Rague´
Schleyer, Angew. Chem., Int. Ed. Engl., 1992, 31, 755.
6 N. Kuhn, A. Kuhn and M. Schulten, J. Organomet. Chem., 1991, 402,
C41.
7 G. E. Carr, R. D. Chambers, T. F. Holmes and D. G. Parker,
J. Organomet. Chem., 1987, 325, 13.
8 K. C. Grega, M. R. Barbachyn, S. J. Brickner and S. A. Mizsak, J. Org.
Chem., 1995, 60, 5255.
9 1 was prepared by LiAlH4 reduction of (1-isopropyl-2-methylpropyli-
dene)-phenylamine (94% yield) using a modification of the literature
procedure: R. K. Howe, J. Org. Chem., 1968, 33, 2848.
10 M. Ali, C. Eaborn and D. R. M. Watson, J. Organomet. Chem., 1974,
78, 83; A. R. Bassindale, C. Eaborn and D. R. M. Watson,
J. Organomet. Chem., 1970, 25, 57.
11 R. Contreras, J. M. Grevy, Z. Garc´ıa-Herna´ndez, M. Gu¨izado-
Rodriguez and B. Wrackmeyer, Heteroat. Chem., 2001, 12,
542.
The Royal Society (PWD) and Nuffield Foundation (PWD;
ARP {Undergraduate Summer Bursary}) are thanked for grants.
Financial support from the EPSRC (MJH; SS), Lubrizol (CASE
award MJH) and the University of Leicester is gratefully
acknowledged. Prof. J. M. Percy and Drs M. R. Crampton,
D. L. Davies, S. Handa and D. Hodgson are thanked for helpful
discussions.
12 11 was prepared from 1 by reductive methylation using a modified
literature procedure: R. F. Borch and A. I. Hassid, J. Org. Chem., 1973,
37, 1673.
13 (a) V. P. Vitullo, J. Org. Chem., 1969, 34, 224; (b) J. March, Advanced
Organic Chemistry, Wiley, New York, 1992, pp. 1079.
14 M. B. Smith, L. Guo, S. Okeyo, J. Stenzel, J. Yanella and
E. LaChappelle, Org. Lett., 2002, 4, 2321.
This journal is ß The Royal Society of Chemistry 2005
Chem. Commun., 2005, 3835–3837 | 3837