Synthesis of a stable indium complex derived from γ-silyl-α, α-difluorobromopropyne: Evaluation of experimental parameters

Article abstract of DOI:10.1016/j.jfluchem.2003.12.017

Fluoroallenes are potentially useful cyclization precursors because of the synergistic combination of fluorine's unique stereolectronic features and the rich chemistry of allenes. We have optimized the experimental conditions needed for the formation of a

Full text of DOI:10.1016/j.jfluchem.2003.12.017

Journal of Fluorine Chemistry 125 (2004) 641–645
Synthesis of a stable indium complex derived from
g-silyl-a,a-difluorobromopropyne: evaluation of
experimental parameters
Satoru Arimitsu, Bo Xu, Tara L.S. Kishbaugh, Leanne Griffin, Gerald B. Hammond*
Department of Chemistry and Biochemistry, University of Massachusetts, Dartmouth, 285 Old Westport Road, North Dartmouth, MA 02747, USA
Received 4 November 2003; received in revised form 25 November 2003; accepted 1 December 2003
Abstract
Fluoroallenes are potentially useful cyclization precursors because of the synergistic combination of fluorine’s unique stereolectronic
features and the rich chemistry of allenes. We have optimized the experimental conditions needed for the formation of a difluoropropargy-
lindium complex 2. Sonication, lower reaction temperatures (5 8C), and dilute concentration of the starting material (0.15 M) are required to
maximize production of this complex. Although the structure of this complex remains unknown, we have found that the nature of the alkyl
substituents on the silyl group does not influence the formation of 2 but it does affects the allenyl 5 to propargyl 4 ratio.
# 2004 Elsevier B.V. All rights reserved.
Keywords: Fluoroallenes; Indium; Propargyldifluoro; tri-iso-Propylsilyl
1. Introduction
fluorinated allenes have been of interest mainly to theoreti-
cians because of their inaccessibility in terms of synthesis.
We are now well qualified to tackle this investigation. After
our discovery of fluoroallenylphosphonate, [3] and, more
recently, our synthesis of gem-difluoroallenol 5a [4] from
tri-iso-propylsilyldifluorobromopropyne (1a), [5] we have
made possible a number of hitherto unknown synthetic path-
ways such as Mo(CO)6 catalyzed [2 þ 2] difluoro allene–
alkyne intramolecular cyclization, [6] a facile synthesis
of functionalized conjugated dienes, [7] and an extremely
mild ‘Baldwin-disfavored’ intramolecular cyclization [4]
(Scheme 1). As shown in Scheme 1, the cornerstone of our
investigations on fluorinated synthons for cyclic systems
relies on an efficient preparation of 2a, which is hindered
by a competing dimerization leading to 3a.
Despite its distinctive stereoelectronic properties and
widespread use—from polymers to liquid crystals to phar-
maceuticals and agrochemicals—fluorine is not an easy
atom to introduce in an organic compound. In particular,
methods for the incorporation of fluorine in cyclic, non-
aromatic systems are scarce. In this regard, the use of a
fluorine substituent as activator and controller (b-cation-
destabilizing effect, b-anion-stabilizing effect and leaving
group ability as Fꢀ) has only recently been realized for ring
construction [1]. Our research addresses the shortage of
convergent methods to synthesize alicyclic difluorinated
systems by employing fluorinated functionalized allenes
as coupling partners in ring construction methodologies.
By developing the chemistry of fluoroallenes as novel
fluorine-containing building blocks, we seek to merge the
unique stereolectronic features of fluorine with the rich
chemistry of allenes, to create functionalized complex mole-
cules not available through fluorinating agents or other build-
ing block approaches. Fluorinated allenes are practically
unknown. Indeed, except for the pioneering work of Dolbier
on the synthesis of the parent di- and mono-fluoroallene [2],
2. Results and discussion
Our ultimate goal is the synthesis of a fluorinated allene
dianion equivalent (C¼C¼CF2)2ꢀ, a task which we expected
to accomplish by introducing two mutually orthogonal func-
tional groups on the g-carbon. Hence, our initial task was to
optimize the production of 2. With this purpose in mind we
have undertaken an extensive study of the reaction conditions
leadingtothemostefficientproductionofthisintermediate,so
* Corresponding author. Tel.: þ1-508-999-8865; fax: þ1-508-910-6918.
E-mail address: ghammond@umassd.edu (G.B. Hammond).
0022-1139/$ – see front matter # 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.jfluchem.2003.12.017

642
S. Arimitsu et al. / Journal of Fluorine Chemistry 125 (2004) 641–645
TIPS
H
F
TIPS
X
F
F
TIPS
O
F
NHTs
Z
R
F
TIPS
TIPS
X
TIPS
F
X
HCHOaq
F
F
F
TIPS
F
Z
In
OH
5a
R
2
3a
2a
R
F
X
TIPS
R
TIPS
F
TIPS
F
TIPS
F
O
Z
F
F
O
Z=N(CO2Et)2, NHTs
R=alkyne, aryl, P(O)(OEt)2
X=H, F
Scheme 1. Cyclization building blocks derived from a single fluorinated allene.
that we might minimize production of dimer 3 and maximize
the formation of 2, thereby gaining a greater understanding
of the chemistry of this intriguing organometallic complex.
all three cases. After the addition of aqueous formaldehyde,
the regioselectivity of the reaction varied considerably.
Whereas 2a reacted with formaldehyde to produce allenol
5a preferentially, 1b and 1c produced homopropargyl alco-
hol 4b and 4c, respectively, rather than the desired allenol 5b
and 5c. The reaction of 2a with other aldehydes (R ¼ aryl,
alkyl), produced exclusively the thermodynamically more
stable homopropargyl alcohol [4].
2.1. Effect of the silyl group on the formation of 2
The synthesis of tert-butyldimethylsilyldifluorobromo-
propyne (1b) and phenyldimethylsilyldifluorobromopro-
pyne (1c) was carried out according to the methodology
used for tri-iso-propylsilyldifluorobromopropyne (1a) [5] in
58 and 75% unoptimized yields, respectively. The chemical
behavior of silyl propynes 1b and 1c towards indium was
similar to that of 1a. Compounds 1b and 1c furnished the
indium intermediate 2b and 2c, respectively, accompanied
by dimer 3b and 3c, in a ratio comparable to that observed
for 1a (Table 1). The reaction was complete after 2 h at rt in
2.2. Effect of metals on the production of 2a
In the presence of zinc, 1a reacts with aldehydes or
ketones to yield the corresponding homopropargyl alcohols
in very good yields [8]. We examined the behavior of 1a
toward other metals (Table 2). In all cases studied, a discrete
organometallic intermediate, such as the one found with
indium (i.e., 2a), was not observed by 19F NMR; and none of
the reactions conditions produced the desired allenol 5a. The
use of cadmium [9,10] gave rise to the homopropargyl
alcohol 4a (entry 1); when lead was used [11,12], the only
compound observed was the reduced product TIPS-CC-
CF2H (6a) (entry 2). Under a variety of reaction conditions,
gallium–which has a similar first ionization potential and
reduction potential as indium, [13] only furnished dimer 3a.
Neither magnesium nor aluminum [14] were successful,
returning 1a unchanged (Scheme 2).
Table 1
Effect of the trialkylsilyl group on the production of the allenol 5
Entry R3Si
Ratio of the
intermediate to dimera
Ratio of productsb
2 (d ꢀ88) 3 (d ꢀ99) 4 (d ꢀ95) 5 (d ꢀ106)
1
2
3
(i-Pr)3Si
7
6
8
1
1
1
1
5
7
6
1
1
t-BuMe2Si
PhMe2Si
a A 0.32 M mixture of 1 and indium powder (1.2 eq.) in H2O: THF
(4:1) system was stirred at rt for 2 h before an aliquot in CDCl3 was
monitored by 19F NMR.
2.3. Effects of addition
b To the above mixture was added 5 eq. of aqueous HCHO; this mixture
was then sonicated for 17 h prior to workup. The ratio is based on the isolated
product.
We had found earlier that prior formation of the indium
complex 2a is necessary for good conversion to the allenol

S. Arimitsu et al. / Journal of Fluorine Chemistry 125 (2004) 641–645
643
Table 2
Effect of metals on the reactivity of 1a
Entry
Metal
Solvent
Ratio of productsa
1a (d ꢀ32)
4a (d ꢀ95)
3a (d ꢀ99)
5a (d ꢀ106)
6a (d ꢀ105)
1
2
3
4
5
6
Cd
DMF sonication
THF sonication
0
0
1
1
1
1
30
0
1
0
1
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
Pb
Ga
Mg
H2O: THF sonication
THF Rt, reflux, sonication
THF or H2O: THF
THF
0
0
Al
0
Al þ 10% PbCl2
0
a Determined by 19F NMR of the crude sample mixture after workup.
F
F
F
F
In
•
F
R3Si
5e q.
aq. HCHO
R3Si
H2O:THF
(4:1)
+
R3Si
F
In
2
Br
2a (R3=i-Pr3)
2b (R3= t-BuMe2)
2c (R3= PhMe2)
3a (R3=i-Pr3)
3b (R3= t-BuMe2)
3c (R3= PhMe2)
1a (R3=i-Pr3)
1b (R3= t-BuMe2)
1c (R3= PhMe2)
R3Si
HO
5a (R3=i-Pr3)
F
F
F
+
•
R3Si
F
OH
4a (R3=i-Pr3)
4b (R3= t-BuMe2)
4c (R3= PhMe2)
5b (R3= t-BuMe2)
5c (R3= PhMe2)
Scheme 2.
5a [4]. In fact, if all of the reagents, including the aqueous
solution of formaldehyde, are mixed at once (entry 1,
Table 3), only the homopropargyl alcohol 4a is produced,
as a minor product. Dimer 3a was the major component in
the reaction mixture. On the other hand, if 2a is allowed to
form first and none of 1a remains in solution prior to the
addition of formaldehyde, the major component in the
reaction mixture is the desired 5a (entry 2).
Presumably, dimerization could be minimized either by
dilution or by keeping the effective concentration of 1a small
near the source of indium. In order to assess this hypothesis,
a THF solution of 1a was added via a syringe pump to an
aqueous slurry of indium over the course of 3 h. As can be
seen in Table 4, this proved to be a positive development
(compare entry 1 to the other entries). Notably, dilution
increases the ratio of 2a to 3a (compare entries 3 and 5).
Furthermore, sonication is more efficient than stirring (com-
pare entries 2 and 3) (Scheme 3).
2.4. Effect of temperature
Early on, we had observed that the amount of dimer 3a
produced seemed to be contingent upon the temperature of
the reaction. The temperature of the ultrasound bath quickly
reaches 45 8C if nothing is done to cool the bath; reactions
performed at this temperature had, on average, a higher
percentage of 3a (entry 1, Table 5). Although sonication at
25 8C, (entry 2), provided a better result, it was very difficult
to maintain that temperature in the ultrasound bath. At lower
Table 4
The effect of slow addition of 1a on the formation of 2a and 3a
Entry Concentration Mixinga
(M)
Rate of
addition
Ratio of productsb,c
2a (d ꢀ88) 3a (d ꢀ99)
1
2
3
4
5
0.32
0.13
0.13
0.5
Stirring
In situ
7
1
1
1
1
1
Sonication Over 3 h 18
Stirring
Stirring
Stirring
Over 3 h 15
Over 3 h 11
Over 3 h 10
Table 3
Effect of the order of addition on the production of 5a
1.0
a Reactions carried out with stirring were at rt; reactions which were
Entry Order of addition
Ratio of productsa
sonicated quickly reach ꢁ45 8C.
b The progress of the reaction was monitored by analyzing an aliquot in
CDCl3 by 19F NMR.
4a (d ꢀ95)
2
3a (d ꢀ99)
3
5a (d ꢀ106)
1
2
in situ
0
c Reactions in which the propyne was added slowly, were complete 1 h
after addition; reactions in which the propyne was added to the indium in
one aliquot were complete after 2 h.
stepwise
1
1
6
a Determined by 19F NMR of the crude sample mixture after workup.

644
S. Arimitsu et al. / Journal of Fluorine Chemistry 125 (2004) 641–645
F
F
F
F
In
•
TIPS
TIPS
F
TIPS
F
+
In
Br
2
H2O:THF
(4:1)
1a
2a
3a
Scheme 3.
Table 5
material 1a remained in the mixture even after 6 h. The
optimal results obtained in entry 2 have been consistently
reproduced. Replacing water with saturated aqueous NH4Cl
in the reaction solvent at 0.5 M reduced the proportion of
unreacted starting material (entry 4) but it showed no
improvement when compared with water at lower concen-
trations (compare entries 2–3 with entries 5–6).
Effect of temperature on the reaction of 1a with indium
Entry
Temp. (8C)
Ratio of productsa
1a (d ꢀ32)
2a (d ꢀ88)
3a (d ꢀ99)
1
45
0
0
3
29
14
8
2
1
1
1
2
25
3
4e
0–5
5–10
27
0
a A 0.5 M mixture of 1a and indium powder (1.2 eq.) in H2O/ THF
(4:1) was sonicated for 6 h before monitoring by 19F NMR.
3. Conclusion
We have optimized the experimental conditions needed
for the formation of the indium complex 2. Sonication, lower
reaction temperatures (5 8C), and dilute concentration of the
starting material (0.15 M) are required to maximize produc-
tion of 2a. Although the structure of this complex remains
unknown, we have found that the nature of the alkyl sub-
stituents on the silyl group does not influence the formation
of 2; however, the presence of a tri-iso-propylsilyl group
is indispensable for the synthesis of fluorinated allenyl
alcohol 5a.
temperatures (entries 3 and 4), the reaction becomes more
sluggish, requiring between 6-8 h to complete, as compared
with 2 h at room temperature. At 0–5 8C (entry 3), the
predominantly aqueous solution froze partially, causing
inefficient mixing of the sample with the metal, and a large
proportion of unreacted starting material remained. If the
temperature was kept between 5–10 8C, the solution did not
freeze and the reaction was complete in 8 h, yet the ratio
of 2a to dimer 3a was far from optimum (entry 4). We
examined the stability of a mixture of 2a and 3a in solution
(diethyl ether or THF-H2O), and found that in either solu-
tion, at room temperature, 2a was converted into 3a over a
24 h period. If the solution of 2a and 3a was kept between
0 and 10 8C, the mixture showed little change even after
three weeks.
4. Experimental
All moisture sensitive reactions were done using flame-
dried glassware flushed with argon, magnetic stirring, and
dry, freshly distilled solvents. THF was dried using a solvent
purification system (Glass Contour, Laguna Beach, CA,
USA). Other solvents were HPLC grade and were used
without purification. TIPS-acetylene was purchased from
GFS Chemicals Inc. (Columbus, OH, USA) and used with-
out further purification. Other commercial reagents were
purchased from Aldrich and used as received. Standard
work-up involves washing ether extraction layer with sat.
2.5. Effects of concentration
The effect of substrate concentration when the ultrasound
bath was set at a temperature of 5–10 8C was investigated.
We noticed that as the concentration decreased, so did the
rate of reaction, but at 0.15 M, the ratio of 2a to 3a reached
its maximum (entry 2, Table 6). At 0.05 M (entry 3), starting
Table 6
Effect of concentration on the production of 2a and 3a
Entry
Solvent (4:1)
Concentration (M)
Ratio of Productsa
1a (d ꢀ32)
2a (d ꢀ88)
3a (d ꢀ99)
1
2
3
4
5
6
H2O: THF
H2O: THF
0.5
27
0.3
8
14
21
12
18
26
11
1
1
1
1
1
1
0.15
0.05
0.5
H2O: THF
sat. aq. NH4Cl:THF
sat. aq. NH4Cl:THF
sat. aq. NH4Cl:THF
0.5
1
0.15
0.05
3
a A mixture of 1a and indium powder (1.2 eq.) in the indicated solvent system was sonicated at 5 8C for 6 h before monitoring by 19F NMR.

S. Arimitsu et al. / Journal of Fluorine Chemistry 125 (2004) 641–645
645
NH4Cl, then brine; drying over MgSO4; and concentration
in vacuo. All reactions were monitored using one of the
following techniques: TLC, GC–MS, and/or 19F NMR.
Analytical TLC was performed using Macherey-Nagel
Polygram Sil G/UV254 precoated plastic plates and visua-
lized using phosphomolybdic acid (5% in methanol). Flash
chromatography was performed using silica gel 230–
400 mesh, 40–63 microns (Lagand Chemicals). 1H, 19F
and 13C NMR spectra were recorded in CDCl3 at 300,
282, and 75 MHz respectively. 19F NMR spectra was refer-
enced against external CFCl3. 19F NMR spectra was broad-
band decoupled from hydrogen nuclei. Elemental analyses
were performed by Atlantic Microlab Inc., Norcross, GA,
USA.
(rel. int.): 270 (Mþ þ1, 5), 268 (Mþ ꢀ1, 5), 189 (5), 132
(20), 57 (100), 43 (80).
Phenyldimethylsilyldifluorobromopropyne (1c): IR (neat)
n: 3277 (s), 3070 (s), 3023 (m), 2961 (s), 2035 (s), 1429 (s),
1251 (s), 1118 (s), 821 (s) cmꢀ1; 1H NMR (CDCl3) d: 7.57–
7.39 (m, 5H), 0.509 (s, 6H); 19F NMR (CDCl3) d: ꢀ33.3(s,
2F); 13C NMR (CDCl3) d: 133.9, 133.2, 130.4, 129.8, 128.2,
122.8, 100.9 (t, 287 Hz), ꢀ1.60. EIMS (probe) 70 eV, m/z
(rel. int.): 290 (Mþ þ1, 3), 288 (Mþ ꢀ1, 3), 275 (4), 273 (4),
209 (38), 159 (16), 132 (100), 77 (10).
6. Optimized production of 2a and 5a
To a 0.15 M solution of 1 in H2O: THF (4:1) is added
indium (1.2 eq.). This mixture is sonicated at 5 8C for 8 h.
The temperature in the Bransonic1 ultrasound bath was
adjusted every 30–45 min by additional ice. An aliquot was
analyzed in either C6D6 or CDCl3 by 19F NMR to monitor
rate of conversion to 2a. For the formation of 4a and 5a, the
ultrasound bath was charged with water at rt, and 5 eq. of
37% aq. HCHO was added to 2a. This mixture was sonicated
for an additional 17 h. Standard workup produced 4a and 5a,
with some contamination by 3a. The alcohols, 4a and 5a,
can be separated from 3a by column chromatography (20:1;
Hexanes:EtOAc).
5. General procedure for the synthesis of 1-trialkylsilyl-
3-bromo-3,3-difluoropropyne
Under Ar atmosphere, n-butyl lithium (1.1 eq., 1.6 M in
hexane) was added dropwise via syringe to a ꢀ78 8C solu-
tion of the trialkylsilylacetylene in THF (0.6 M). The result-
ing solution was stirred for 30 min to 1 h before CF2Br2
(1.5 eq.) was added slowly via cannula. The solution was
allowed to warm to rt over 9 h before being quenched by the
addition of sat. aq. NH4Cl. It was extracted with ether (4ꢂ).
The combined organic layers were washed with brine, dried
(MgSO4), and concentrated to yield a viscous orange oil
which was purified by distillation (bulb-to-bulb in the case
of 1b and 1c) to yield 1 as a colorless oil.
Acknowledgements
tri-iso-Propylsilyldifluorobromopropyne (1a): IR (neat);
n: 2947 (s), 2869 (s), 2190 (m), 1464 (s), 1187 (s), 1098 (s),
The generous financial support of the Camille and
Henry Dreyfus Foundation (SF-01-001) and the National
Science Foundation (CHE- 0213502) is acknowledged with
gratitude.
947 (s) cmꢀ1 1H NMR (CDCl3) d: 1.11 (s); 19F NMR
;
(CDCl3) d: ꢀ32.6 (s, 2F); 13C NMR (CDCl3) d: 100.61 (t,
290 Hz), 97.16 (t, 36.5 Hz), 95.1 (t, 4.7 Hz),18.01, 11.28;
EIMS (probe) 70 eV, m/z (rel. int.): 312 (Mþ þ1, 4), 310
(Mþ ꢀ1, 4), 269 (14), 267 (14), 143 (27), 77 (100). Anal.
Calcd. for C12H21SiF2Br: C, 46.33; H, 6.75. Found: C,
46.91; H, 6.83. The tri-iso-propylsilyl acetylene starting
material, purchased from GFS Co., contained di-iso-propyl-
propenylsilyl acetylene (11%) and di-iso-propylpropylsilyl
acetylene (14%). Fractional distillation did not remove these
impurities. The GC–MS analysis shows all three compo-
nents were alkylated and their ratios perfectly match that of
the starting material. The calculated C% and H% was based
on the formula of TIPS-difluorobromopropyne. If the two
impurities are removed from the calculation, the values
found for C% and H% are 46.84 and 6.67%, respectively.
tert-Butyldimethylsilyldifluorobromopropyne (1b): IR
(neat); n: 2932 (s), 2887 (m), 2861 (s), 2192 (m), 1472
References
[1] J. Ichikawa, Pure Appl. Chem. 72 (2000) 1685–1689.
[2] W.R. Dolbier Jr., Acc. Chem. Res. 24 (1991) 63–69.
[3] A.J. Zapata, Y. Gu, G.B. Hammond, J. Org. Chem. 65 (2000)
227–234.
[4] Z. Wang, G.B. Hammond, J. Org. Chem. 65 (2000) 6547–6552.
[5] Z. Wang, G.B. Hammond, Chem. Commun. (1999) 2545–2546.
[6] Q. Shen, G.B. Hammond, J. Am. Chem. Soc. 124 (2002) 6534–6535.
[7] Q. Shen, G.B. Hammond, Org. Lett. 3 (2001) 2213–2215.
[8] Z. Wang, G.B. Hammond, Tetrahedron Lett. 41 (2000) 2339–2342.
[9] B. Sain, D. Prajapati, J.S. Sandhu, Tetrahedron Lett. 33 (1992) 4795–
4798.
[10] S. Araki, H. Ito, Y. Butsugan, J. Organomet. Chem. 347 (1988) 5–9.
[11] H. Tanaka, S. Yamashita, Y. Ikemoto, S. Torii, Tetrahedron Lett. 29
(1988) 1721–1724.
(s), 1183 (s), 1096 (s), 946 (s) cmꢀ1; H NMR (CDCl3) d:
[12] J.-Y. Zhou, G.-F. Sun, M.-F. Zhang, Y. Jia, S.-H. Wu, Chin. J. Chem.
15 (1997) 361–365.
1
0.19 (s, 6 H) 0.97 (s, 9H); 19F NMR (CDCl3) d: ꢀ32.8 (s,
2F); 13C NMR (CDCl3) d: 104.7 (t, 290 Hz), 96.4 (t, 36 Hz),
95.4 (t, 3 Hz), 26.0,16.8, ꢀ5.26. EIMS (probe) 70 eV, m/z
[13] Z. Wang, S. Yuan, C.-J. Li, Tetrahedron Lett. 43 (2002) 5097–5099.
[14] H. Tanaka, S. Nakahata, H. Watanabe, J. Zhao, M. Kuroboshi, S.
Torii, Inorg. Chim. Acta 296 (1999) 204–207.