Enantioselective Synthesis of 6,6-Disubstituted Pentafulvenes Containing a Chiral Pendant Hydroxy Group

Article abstract of DOI:10.1002/chem.201704247

Simple enantioselective synthesis of 6,6-disubstituted pentafulvenes bearing chiral pendant hydroxy groups are attained by cascade reactivity using commercially available proline-based organocatalysts. Condensation of cyclopentadiene with the acetyl function of a 1,2-formylacetophenone, followed by cyclization of a resulting fulvene-stabilized carbanion with the formyl group, generates bicyclic chiral alcohols with initial er values up to 94:6. Exceptional enantio-enrichment of the resultant alcohols results upon crystallization—even near racemic samples spontaneously de-racemize. This enables new families of substituted cyclopentadienes that are both enantiomerically and diastereomerically pure to be rapidly attained.

Full text of DOI:10.1002/chem.201704247

DOI: 10.1002/chem.201704247
Communication
&
Synthetic Methods
Enantioselective Synthesis of 6,6-Disubstituted Pentafulvenes
Containing a Chiral Pendant Hydroxy Group
Ryan Nouch,[a] Melchior Cini,[a] Marc Magre,[b] Mohammed Abid,[a] Montserrat Diꢀguez,*[b]
Oscar Pꢁmies,[b] Simon Woodward,*[a] and William Lewis[a]
Abstract: Simple enantioselective synthesis of 6,6-disub-
stituted pentafulvenes bearing chiral pendant hydroxy
groups are attained by cascade reactivity using commer-
cially available proline-based organocatalysts. Condensa-
tion of cyclopentadiene with the acetyl function of a 1,2-
formylacetophenone, followed by cyclization of a resulting
fulvene-stabilized carbanion with the formyl group, gener-
ates bicyclic chiral alcohols with initial er values up to
94:6. Exceptional enantio-enrichment of the resultant alco-
hols results upon crystallization—even near racemic sam-
ples spontaneously de-racemize. This enables new families
of substituted cyclopentadienes that are both enantiomer-
ically and diastereomerically pure to be rapidly attained.
Scheme 1. Traditional approaches to (chiral) 6,6- and 6-substituted pentaful-
venes.
compound class. Frequent applications include: cycloadditions
to generate complex, polycyclic scaffolds[5,6] and their use as
intermediates in the synthesis of substituted (sometimes chiral)
cyclopentadienyl derivatives by nucleophilic addition to the
exocyclic C=C bond as a route to (asymmetric) cyclopenta-
diene units.[7–9] Herein we describe a simple approach to fami-
lies of asymmetric 6,6-disubstituted pentafulvenes bearing
chiral pendant hydroxy groups (for further functionalization)
by straightforward organocatalytic methodology. The pentaful-
venes are useful as intermediates in the synthesis of substitut-
ed cyclopentadienes as single enantiomers and diastereomers.
To develop a route to new chiral pentafulvenes bearing
pendant hydroxy groups, we investigated the reaction be-
tween 2-acetyl-benzaldehyde and cyclopentadiene in the pres-
ence of organocatalysts (Table 1). Reference samples of (ꢀ)-2a
were prepared via pyrrolidine catalysed reactions, although
these were slower than the later enantioselective reactions
(see Table S1 in the Supporting Information). Interestingly, we
noted that (ꢀ)-2a de-racemizes exceptionally readily, with
each crystal being a single enantiomer, in the same manner as
the classical tartrate crystals of Pasteur.[10,11] Individual crystals
of (R) or (S)-2a are readily attained stochastically from initially
racemic (ꢀ)-2a. This occurs predictably for scalemic 2a making
exceptional enantio-enrichment possible. Conglomerate crystal
formation in 2a is driven by a strongly stereodirecting helical
hydrogen bonding array in its packing (see Figure S5 in the
Supporting Information). Simple (l)-proline gave only low
yields of 2a, but greater success was had with derivatives of
(S)-2-pyrrolidinemethanol (LA). Smaller amounts of achiral 3
and aldol product 4 could also be isolated from the reaction.
Compound 4 is a known product of 1a (formed in low er in
asymmetric aldol chemistry)[12] but the preparation of pentaful-
Synthetic methodology for pentafulvene formation has not al-
tered significantly since these were first prepared by Thiele in
1900 by sodium ethoxide-facilitated condensation of cyclopen-
tadiene with ketones (Scheme 1).[1] Although improved by
Little[2] and Ottosson,[3] among others, none of these allows
access to chiral fulvenes. Little’s method uses pyrrolidine catal-
ysis to increase the reactivity of the ketone, while Ottosson’s
method uses sodium cyclopentadienide as a more reactive
source of the cyclopentadiene nucleophile. Both approaches
allow for the reaction of more hindered or less activated car-
bonyls. Across the board, examples of syntheses of pentaful-
venes bearing chiral pendant functional groups are almost un-
known, one rare example being Togni’s condensation of
sodium cyclopentadienide with
a
homochiral amide
(Scheme 1).[4] Unfortunately, this method was limited in scope
as only two chiral examples of singly substituted 6-derivatives
could be accessed. Despite a complete lack of effective stereo-
selective syntheses, pentafulvenes remain a commonly used
[a] R. Nouch, Dr. M. Cini, M. Abid, Prof. Dr. S. Woodward, Dr. W. Lewis
GlaxoSmithKline Carbon Neutral Laboratories for Sustainable Chemistry
University of Nottingham
6 Triumph Road, Nottingham, NG7 2GA (UK)
E-mail: simon.woodward@nottingham.ac.uk
[b] Dr. M. Magre, Prof. Dr. M. Diꢀguez, Prof. Dr. O. Pꢁmies
Departament de Quꢂmica Fꢂsica i Inorgꢁnica
Universꢃtat Rovira i Virgili
Campus Sescelades, Marcel, lꢂ Domingo 1-43007, Tarragona (Spain)
E-mail: montserrat.dieguez@urv.cat
Supporting information and the ORCID identification number(s) for the au-
thor(s) of this article can be found under https://doi.org/10.1002/
chem.201704247.
Chem. Eur. J. 2017, 23, 1 – 5
1
ꢂ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Communication
Table 1. Summary of optimization for enantioselective preparation of penta-
fulvene 2a.[a]
Table 2. Substrate scope for chiral pentafulvene (2) formation.[a]
Cmpd.
R1
R2
T
[8C]
t
[h]
Yield
[%][b]
Initial
er
Recrystallized
er
2a
H
H
H
H
H
H
H
H
H
Me
F
15
15
15
22
15
15
25
15
25
25
25
25
6
6
4.5
2.5
6
5
4.5
6
3
3
78
65
53
91
72
54
39
67
52
39
38
30
94:6
94:6
88:12
82:18
91:9
91:9
92:8
89:11
88:12
90:10
91:9
>99:1
98:2
(S)-2a[d]
2b
Ph
tBu
Me
OMe
F
H
H
H
H
>99:1
[c]
2c
2d
2e
2 f
2g
2h
(S)-2h[d]
2i
(S)-2i[d]
–
–
[c]
Catalyst
Acid
[equiv.]
t
[h]
T
[8C]
Yield
[%][b]
er
>99:1
>99:1
LA
LB
LB
LB
LC
LD
LE
AcOH (0.38)
AcOH (0.38)
AcOH (0.26)
AcOH (0.13)
AcOH (0.13)
AcOH (0.13)
AcOH (0.13)
24
2
2
6
24
2
25
25
22
15
22
22
22
90[c]
71
76
78
69
71
<5
79:21
86:14
89:11
94:6
80:20
90:10
n.d.[d]
[c]
–
>99:1
>99:1
99:1
F
Cl
Cl
3
2.5
H
89:11
99:1
4
[a] Carried out on
1
(0.6–2.1 mmol) with AcOH (0.13 equiv) and LB
(0.38 equiv) in DMF (1–5 mL); for full details see Supporting Information.
[b] Isolated yield. [c] Could not be enriched by recrystallization. [d] Syn-
thesized under identical conditions but providing the (S)-pentafulvene
using (R)-LB.
[a] Carried out on 1a (0.39 mmol) in DMF (1 mL); for full details of optimiza-
tion see Supporting Information. [b] Isolated yield unless otherwise stated.
[c] Conversion determined by 1H NMR spectroscopy with no internal stan-
dard present (mass balance of >90% confirmed independently). [d] Not de-
termined.
Fulvene formation is tolerant of a range of electronic sub-
stituent effects in the 5-position (2d–f) but is more sensitive to
steric factors (runs 2b,c). The opposite situation applies to sub-
stitution in the 6-position (runs 2g–i). The poorer yields ob-
tained for halogen containing 2 f and 2i appear to be due to
the decreased stability of the product rather than lower con-
version based on control experiments. Modification of the re-
action conditions allowed isolation of the two by-products 3
and 4 in useful quantities. Reduction of the reaction tempera-
ture to 08C provided condensation by-product 3 in 45% yield
after 48 hours. We propose that this is due to the lower tem-
perature strongly disfavoring the less reactive keto functionali-
ty, allowing for an increase in condensation between the more
reactive aldehyde and cyclopentadiene at longer reaction
times. Running the reaction under its optimized conditions
(Table 1, Run 4) but without added cyclopentadiene resulted in
a 55% isolated yield of aldol by-product (S)-4 (er<85:15). Isola-
tion of these by-products enabled us to unambiguously define
the reaction mechanism for pentafulvene formation
(Scheme 2). Pathway A can be discounted as intermediate 5
would produce aldol by-product 4, however, resubmission of
isolated 4 to the reaction conditions produces no fulvene
products discounting cyclopentadiene condensation at the
keto group of 4. In addition, HPLC confirmed the chiral centre
in (S)-4 is of opposite configuration to the (R)-2 provided by
LB, eliminating 4 as a simple precursor to 2.[18,19] Similarly, path-
way C can be discounted on the fact that resubmission of iso-
lated 3 to the reaction conditions also produces no 2, disavow-
ing potential intermediates in the H-shift pathway 7a,b. In ad-
dition, pathway C would provide regioisomeric 2’ (Scheme 2) if
appropriately substituted (regardless of the sense of asymmet-
vene products of type 2 is unprecedented, as far as we are
aware. Catalysts bearing too much steric bulk, such as LE or
the bulky diaryl derivatives developed by Jørgenson[12] and
Hayashi,[13] proved ineffective as did the imidazolidinone deriv-
atives of Macmillan.[14] Optimal results were attained with (2S)-
1-(pyrrolidin-2-ylmethyl)pyrrolidine (LB),[15,16] with a reduced
number of equivalents of acetic acid (0.13 vs. 0.38 equiv).
Below 0.38 equivalents of catalyst LB the reaction conversion
suffered but the er remained high. Alternative acids were also
trialed but all performed worse than acetic acid in the reaction
(see Table S1 in the Supporting Information). The reaction
could be scaled to gram quantities without any significant
negative yield or er effects (see Experimental section). Excesses
of cyclopentadiene were employed, as it is cheap and easily re-
moved during purification, to ensure reliability of the reaction
and to hinder the production of by-product 4 as much as pos-
sible.
The precursor 2-acetyl-benzaldehydes (1a–1i) needed for
generalization of the reaction are easily prepared on multi-
gram scales via two routes. Directed lithiation of a range of
benzyl alcohols followed by reaction with acetaldehyde and
subsequent oxidation provides 1a–c in two steps. Alternatively,
Phan’s phenol formylation, followed by acyl hydrazide forma-
tion and subsequent acyl transfer was used for 1d–i.[17] The op-
timal catalyst LB is commercially available, but also readily and
efficiently prepared on multi-gram scales, starting from low
cost (l)-proline using our optimized method (see Supporting
Information). The generality of the reaction was thus explored
(Table 2).
&
&
Chem. Eur. J. 2017, 23, 1 – 5
www.chemeurj.org
2
ꢂ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Communication
Scheme 2. Mechanistic possibilities for the formation of pentafulvene 2a.
ric induction). Only pathway B, in which condensation of cyclo-
pentadiene with the acetyl ketone occurs first, correctly ac-
counts for both the regiochemistry and sense of stereochemis-
try observed in the reaction. The acidity of the a-methyl ful-
vene group (pKa ca. 22.1 in DMSO)[20] and the subsequent in-
version in the SN1-like cleavage of the amine both have prece-
dent.[21–23] Intermediates 6a,b could not be isolated or
detected but remain the only viable option consistent with the
experimental data. Large leaving groups, as in 6b, strongly
favour inversion outcomes in SN1 hydrolysis reactions and an
ability for additional protonation of the pyrrolidine nitrogen
may account for the improved performance of LB.
diene 9 (see Supporting Information, Figure S2, for an explana-
tion of the structural assignment of 9).
To conclude, we have presented an efficient synthesis of
chiral 6,6-disubstituted pentafulvenes bearing a functionaliza-
ble chiral pendant hydroxy group in moderate to good yields
and enantiomeric ratios. These pentafulvenes possess the in-
teresting and useful characteristic of crystallizing as conglom-
erates, often giving the products in >99:1 er at gram scales.
Once functionalized, these chiral pentafulvene derivatives can
then be converted into substituted cyclopentadienes essential-
ly as both single enantiomers and diastereomers.
Preliminary studies show that, following functionalization of
the pendant hydroxy group, pentafulvenes of type 2 can be
reduced, in the manner of Tacke,[24] using LiBEt3H. One example
of this is shown in Scheme 3. The reduction proceeds with
very high diastereoselectivity (>25:1 as the other diastereomer
Experimental Section
Representative procedure for synthesis of pentafulvene (R)-
2a
1
Commercially available 2-acetyl-benzaldehyde (1.29 g, 78% purity,
6.79 mmol) and cyclopentadiene (3.6 mL, 42.8 mmol) were dis-
solved in DMF (16.0 mL) before the addition of acetic acid (47 mL,
0.82 mmol) and dropwise addition of commercially available (2S)-1-
(pyrrolidin-2-ylmethyl)pyrrolidine (411 mL, 2.52 mmol). This was
then stirred at 158C for 6 hours, after which the reaction was dilut-
ed into ethyl acetate (250 mL) and washed with pH 7.4 phosphate
buffer (3ꢃ100 mL). The solvent was removed in vacuo before pu-
rification via flash column chromatography (eluent: dichlorome-
thane) to yield the crude product as a bright orange solid in 64%
yield (850 mg, 4.33 mmol); yield range on 0.1-1 g scales: 64–78%.
Purification from CH2Cl2/pentane, or dimethoxyethane readily af-
forded (R)-2a as red needles (590 mg, 3.01 mmol, 70% recovery)
with >99:1 er. M.p.: 130–1408C (darkens from this temperature); Rf
is not visible in the H NMR spectrum) following functionaliza-
tion of pentafulvene 2a with triethylsilyl chloride (TESCl). The
silyl ether moiety acts as a blocking group, forcing the hydride
to add anti to it, resulting in the synthesis of syn-cyclopenta-
1
(dichloromethane): 0.30; H NMR (400.2 MHz, CDCl3): dH =7.96 (dd,
Scheme 3. Synthesis of protected pentafulvene 8, via the reaction of penta-
fulvene 2a with TESCl, which is then reduced to syn-cyclopentadiene 9. The
double bond tautomers of 9 are removed upon metal complexation.
J=6.6, 1.9 Hz, 1H, ArH), 7.58–7.55 (m, 1H, ArH), 7.48–7.38 (m, 2H,
ArH), 6.92 (app ddd, J=5.3, 1.7, 1.7 Hz, 1H, CpH), 6.59–6.55 (m, 1H,
Chem. Eur. J. 2017, 23, 1 – 5
www.chemeurj.org
3
ꢂ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Communication
[3] K. Chajara, H. Ottosson, Tetrahedron Lett. 2004, 45, 6741–6744.
[4] H. C. L. Abbenhuis, U. Burckhardt, V. Gramlich, A. Togni, A. Albinati, B.
Mueller, Organometallics 1994, 13, 4481–4493.
[5] Z. Zhou, Z.-X. Wang, Y.-C. Zhou, W. Xiao, Q. Ouyang, W. Du, Y.-C. Chen,
Nat. Chem. 2017, 1–5.
[6] Z. J. Jia, C. Merten, R. Gontla, C. G. Daniliuc, A. P. Antonchick, H. Wald-
mann, Angew. Chem. Int. Ed. 2017, 56, 2429–2434; Angew. Chem. 2017,
129, 2469–2474.
[7] M. Cini, T. D. Bradshaw, W. Lewis, S. Woodward, Eur. J. Org. Chem. 2013,
3997–4007.
[8] M. Cini, T. D. Bradshaw, S. Woodward, W. Lewis, Angew. Chem. Int. Ed.
2015, 54, 14179–14182; Angew. Chem. 2015, 127, 14385–14388.
[9] N. J. Sweeney, J. Claffey, H. Mꢄller-Bunz, C. Pampillꢅn, K. Strohfeldt, M.
Tacke, Appl. Organomet. Chem. 2007, 21, 57–65.
[10] L. Pasteur, Ann. Chim. Phys. 1850, 28, 56–99.
[11] Y. Tobe, Mendeleev Commun. 2003, 13, 93–94.
[12] M. Marigo, T. C. Wabnitz, D. Fielenbach, K. A. Jørgensen, Angew. Chem.
Int. Ed. 2005, 44, 794–797; Angew. Chem. 2005, 117, 804–807.
[13] Y. Hayashi, H. Gotoh, T. Hayashi, M. Shoji, Angew. Chem. Int. Ed. 2005,
44, 4212–4215; Angew. Chem. 2005, 117, 4284–4287.
[14] R. J. Comito, F. G. Finelli, D. W. C. Macmillan, J. Am. Chem. Soc. 2013,
135, 9358–9361.
[15] M. Asami, Bull. Chem. Soc. Jpn. 1990, 63, 721–727.
[16] J. Liu, L. Wang, Synthesis 2016, 960–972.
[17] D. H. T. Phan, B. Kim, V. M. Dong, J. Am. Chem. Soc. 2009, 131, 15608–
15609.
[18] J. Moritani, Y. Hasegawa, Y. Kayaki, T. Ikariya, Tetrahedron Lett. 2014, 55,
1188–1191.
[19] T. Suzuki, K. Ghozati, T. Katoh, H. Sasai, Org. Lett. 2009, 11, 4286–4288.
[20] D. B. Knight, J. A. Harrelson, J. Chem. Soc. Chem. Commun. 1987, 116–
117.
CpH), 6.53–6.49 (m, 2H, CpH), 5.35 (ddd, J=7.0, 7.0, 3.7 Hz, 1H,
CHOH), 3.72 (dd, J=17.5, 7.0 Hz, 1H, CHaHb anti to OH), 3.09 (dd,
J=17.5, 7.0 Hz, 1H, CHaHb syn to OH), 2.01 ppm (d, J=3.7 Hz, 1H,
OH); 13C NMR (100.05 MHz, CDCl3): dC =151.2 (C), 149.5 (C), 139.4
(C), 138.7 (C), 132.8 (CH), 131.1 (CH), 131.0 (CH), 129.4 (CH), 126.5
(CH), 125.4 (CH), 123.4 (CH), 120.0 (CH), 72.9 (CH), 44.0 ppm (CH2);
(CHCl3): n˜max =3614, 3590, 3070, 3045, 3008, 2960, 2927, 2873,
1630, 1476, 1458, 1389, 1368, 1239, 1050, 1021, 997 cmꢁ1; HRMS
found 197.0959 C14H13O+ requires 197.0961 (jsj =1.0 ppm); HPLC
(before crystallization): Chiralpak AD-H; mobile phase, hexane:2-
propanol (4:1 v/v); flow rate, 0.5 mLminꢁ1; retention times (S)-
enantiomer: 11.8 min (6.2%), (R)-enantiomer: 14.2 min (93.8%), er
94:6; [a]D23: +68.0 (er>99:1, c=0.50 in CHCl3); Elemental analysis
(%): Calcd. for C14H12O C, 85.68; H, 6.16; found C, 85.20; H, 6.47.
Acknowledgements
We thank the Edith Johnson Bequest Fund and Aesica Pharma-
ceuticals for partial support of the University of Nottingham
studentships of M.C. and R.N. respectively. M.A. is grateful to
the University of Anbar (Iraq) for PhD study leave (award
IDB600032695). MM, MD and OP thank the Spanish Ministry of
Economy and Competitiveness (CTQ2016-74878-P) and Euro-
pean Regional Development Fund (AEI/FEDER, UE), the Catalan
Government (2014SGR670), and the ICREA Foundation (ICREA
Academia award to MD) for financial support.
[21] R. M. Pagni, Found. Chem. 2011, 13, 131–143.
[22] P. Brewster, E. D. Hughes, C. K. Ingold, P. A. D. S. Rao, Nature 1950, 166,
179–180.
[23] S. M. E. D. Hughes, C. K. Ingold, J. Chem. Soc. 1935, 1196–1200.
[24] N. J. Sweeney, O. Mendoza, H. Mꢄller-Bunz, C. Pampillꢅn, F. J. K. Re-
hmann, K. Strohfeldt, M. Tacke, J. Organomet. Chem. 2005, 690, 4537–
4544.
Conflict of interest
The authors declare no conflict of interest.
Keywords: carbanions · asymmetric catalysis · fulvenes ·
organocatalysis · synthetic methods
Manuscript received: September 10, 2017
Accepted manuscript online: && &&, 0000
Version of record online: && &&, 0000
[1] J. Thiele, Ber. Dtsch. Chem. Ges. 1900, 33, 666–673.
[2] K. J. Stone, R. D. Little, J. Org. Chem. 1984, 49, 1849–1853.
&
&
Chem. Eur. J. 2017, 23, 1 – 5
www.chemeurj.org
4
ꢂ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Communication
COMMUNICATION
&
Synthetic Methods
R. Nouch, M. Cini, M. Magre, M. Abid,
M. Diꢀguez,* O. Pꢁmies, S. Woodward,*
W. Lewis
&& – &&
Organocatalytic reversal of reactivity:
the acyl group in 1,2-C6H4(Ac)(CHO)
reacts with cyclopentadiene over the
formyl group – leading to pentafulvenes
with pendant chiral sec-alcohols
Enantioselective Synthesis of 6,6-
Disubstituted Pentafulvenes
Containing a Chiral Pendant Hydroxy
Group
Chem. Eur. J. 2017, 23, 1 – 5
www.chemeurj.org
5
ꢂ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ