The synthesis of silicon containing cage compounds

Material Information

The synthesis of silicon containing cage compounds
Hankin, Joseph A
Publication Date:
Physical Description:
xi, 57 leaves : illustrations ; 29 cm


Subjects / Keywords:
Silicones ( lcsh )
Cage hydrocarbons ( lcsh )
Cage hydrocarbons ( fast )
Silicones ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Includes bibliographical references.
General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Science, Department of Chemistry
Statement of Responsibility:
by Joseph A. Hankin.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
26922839 ( OCLC )
LD1190.L46 1992m .H36 ( lcc )

Full Text
The Synthesis of Silicon-Containing
Cage Compounds
Joseph A. Hankin
B.A., Northeastern Illinois University, 1982
B.A., University of Colorado at Denver, 1988
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Science

This thesis for the Master of Science
degree by
Joseph A. Hankin
has been approved for the
Department of

Hankin, Joseph A. (M.S., Chemistry)
The Synthesis of Silicon-Containing Cage Compounds
Thesis directed by Professor Robert Damrauer
The synthesis of two silicon-containing cage compounds has been
achieved. The compounds are isomers and have the molecular formula
CigH3o03Si. These two compounds have been fully characterized by 1H
and 13C NMR spectroscopy, elemental analysis, mass spectrometry, and
single crystal X-ray studies. They are the first of a series of silicon-
containing cage compounds (cyclophanes) that have been proposed as
part of an ongoing study of the physical and chemical properties of the
silicon atom in a unique, strained, bridgehead environment.
The work involved in the synthesis, purification, and characterization
of these two isomeric compounds is described in this thesis. The method-
ology developed in this work has been applied to the synthesis of other
silicon-containing cyclophanes with work ongoing to study the physical
properties of these compounds.
This abstract accurately represents the content of the candidate's thesis.
I recommend its publication.

There are a number of people who have helped me as I have
worked towards this degree. Professor Damrauer has encouraged me from
the start both professionally and personally. I owe him many thanks. Jim
Crofter taught me to not be afraid of complex instruments. Mike Milash
taught me something about organization and working hard. Roger Simon
taught me a lot about doing lab work. The Faculty at the UCD Chemistry
Department have been consistently patient and helpful to me, and I am
grateful to them. Curt Haltiwanger and Martin Ashley were generous with
their time and knowledge. Finally, I'd like to thank my family and friends
who have made up a large part of my life.

"I never felt easy till the raft was two mile below there and out in the
middle of the Mississippi. Then we hung up our signal lantern, and judged
that we was free and safe once more........We said there warn't no home
like a raft, after all. Other places do seem so cramped up and smothery,
but a raft don't. You feel mighty free and easy and comfortable on a raft."
-Mark Twain (from The Adventures of
Huckleberry Finn)
"Nothing is real to us except hunger"
- Kakuzo Okakura
"Hebee hebee hebee
hebee hebee hebee
- Billy (a 12 year old, "cortically blind,"
profoundly mentally retarded boy on seeing
headlights of cars zoom past him while walking
along the road.)

1. Introduction and Purpose ....................................1
2. The Synthesis of Silicon-Containing Cage
2.1. Synthetic Strategy.....................................9
2.2. Early Synthetic Results...............................14
2.3. Model System Studies .................................19
2.4. Preparation and Cyclotrimerization of
t-Butyl-tripod .......................................25
2.4.1. Discussion of NMR Data of t-Butyl-
siloxycyclophanes .............................26
2.4.2. Discussion of X-ray Crystal Structures of
t-Butyl-siloxycyclophanes .....................28
3. Summary and Future Work......................................30
3.1. Summary...............................................30
3.2. The Purpose of this Project Retrospect and
4. Experimental Section.........................................36
4.1. General Procedures, and Instruments
4.2. Preparation of 5-Chloro-1 -trimethylsilyl-1 -
4.3. Preparation of 5-lodo-
4.4. Preparation of the Grignard Reagent of
and Reaction with Trichlorosilanes....................39

4.5. General Procedure for the Preparation
of Tri-alkoxysilanes...................................41
4.5.1. Preparation of Tris(4-pentyn-1 -oxy)silane
4.5.2. Preparation of Methyl-
tris(4-pentyn-1 -oxy)silane
4.5.3. Preparation of t-Butyl-
tris(4-pentyn-1 -oxy)silane
4.6. General Notes for Model System Reactions...............45
4.6.1. Preparation of TiCl4/iBu3AI Catalyst Stock
4.6.2. General Procedure for Model System
Reactions with TiCU/iBu3AI......................46
4.6.3. General Procedure for Model System
Reactions with CpCo(CO)2 ....................46
4.6.4. General Procedure for Model System
Reactions with Pd/C, MesSiCI..................47
4.7. Preparation of t-Butyl-siloxycyclophanes...............47
List of References..........................................................51
Appendix I NMR Spectra....................................................53
l.a 1H NMR (300 MHz, CDCI3) 1,2,4 t-Butyl-
l.b 1H NMR (300 MHz, CDCI3) 1,3,5 t-Butyl-

c 13C NMR (75 MHz, CDCI3) 1,2,4 t-Butyl-
d 13C NMR (75 MHz, CDCI3) 1,3,5 t-Butyl-

1.1. Covalent radii for silicon and carbon.......................2
1.2. Flusilazole.................................................3
1.3. Tetrahedral geometries: sp3 hybridization...................4
1.4. Silicon with pentacoordinate (dsp3) hybridization...........5
1.5. Silatranes..................................................5
1.6. Nucleophilic substitutions of carbon and silicon
1.7. Pyridinoparacyclophane-1,9-diene............................7
2.1. Anisotropic ellipsoid plots of 1,2,4 and 1,3,5 t-butyl-
3.1. Series of different sized 1,3,5 t-butyl-silocycyclophanes.34

I. Retrosynthetic Analysis of Target Compound (1)........10
II. Pascal's Synthesis of Carbon Analog (2)...............11
III. Detailed Synthetic Plan for Target Compound (1).......13

Results from Model Study Comparing Catalysts

1. Introduction and Purpose
The original synthetic target for this project was a silicon-containing
cage compound with a basal aromatic ring and three hydrocarbon bridges
to a silicon-hydride (1). The term "cyclophane" is used to describe such
cage compounds with at least one aromatic ring in the structure.
The carbon analog (2) had been reported and was characterized by the
unusually large "upfield" chemical shift for the hydrogen directed inwards
toward the aromatic ring.1

When this project began, silicon-hydride cyclophanes with this type of
proximal relationship between the silicon atom and the aromatic ring had
not been reported in the chemical literature.2 We believed that the target
silicon-hydride cyclophane (1) would be an interesting molecule to study
for several reasons outlined below.
One interesting aspect of silicon-containing compounds is the
contrast of their physical properties and behavior with analogous carbon-
containing compounds. Fundamentally, silicon has a larger covalent
radius (1.17A) than carbon (0.77A) (Figure 1.1).
Figure 1.1. Covalent radii for silicon and carbon.

A listing of select comparative bond lengths reflects the larger size of the
silicon atom.3-4
Bond Bond Lengths (A)
Si-C 1.90
C-C 1.54
Si-H 1.47
C-H 1.09
Si-CI 2.01
C-CI 1.77
Thus, silicon analogs of carbon compounds are different by the fact of the
larger size and longer bond lengths about the silicon atom.
A commercially exploited example of this is "Flusilazole" (Figure
1.2), an agricultural fungicide developed and produced by the Dupont
Chemical Company.5
Figure 1.2. Flusilazole.

Flusilazole, when compared to its carbon analog, is reported to have
improved biological activity and superior physical properties (ie. increased
lipophilicity, and increased volatility). The enhanced properties of
flusilazole are attributed to the longer Si-C bonds in the molecule which
are thought to "fine-tune the molecular size for optimal interaction with the
target enzyme, while at the same time, improving physical properties
governing penetration and distribution in plant tissues."5
Silicon is in the third period of the periodic table with a principle
quantum number n=3. Both carbon and silicon generally adopt tetrahedral
geometry, characteristic of hybrid orbitals forming from atomic s and p
orbitals (Figure 1.3).
Figure 1.3. Tetrahedral geometries: sp3 hybridization.

Additionally, silicon has access to the 3d orbitals. Silicon is known to form
stable five and six coordinate species with molecular shapes characteristic
of hybridization involving the 3d orbitals (Figure 1.4).
Figure 1.4. Silicon with pentacoordinate (dsp3) hybridization.
A series of molecules called "silatranes" (Figure 1.5) exhibit
geometry approaching pentacoordination, with five "bonds" to silicon, and
a rehybridization of the silicon valence orbitals to dsp3 hybrid orbitals.6
(R= methyl, ethyl, phenyl)
Figure 1.5. Silatranes.

Furthermore, 1-chloronorbornane will not undergo nucleophilic
substitution reactions either by Sn1 or Sn2 mechanisms because its rigid
structure prohibits inversion or formation of a planar intermediate. The
analogous 1-chlorosilanorbornane will undergo nucleophilic substitution
reactions with a variety of nucleophiles, presumably because its valance
shell can expand using d-orbital participation (Figure 1.6).7
Figure 1.6. Nucleophilic substitutions of carbon and silicon compounds.
The 1H NMR of Pascal's "in" cyclophane (2) shows an unusually
large upfield chemical shift for the hydrogen directed inwards towards the
aromatic ring. This is due to the anisotropic effect of the aromatic ring
current which causes net shielding of the central "in" hydrogen, and a
consequent upfield shift. Other "in" cyclophanes isolated by Pascal and
co-workers also show this upfield shift.8 The IR spectrum of 2 is also

unusual with a relatively high energy (3325 cm*1) for the C-H stretching
frequency attributed to compression of the "in" methine.8 We expect that
our target compound (1) will show interesting and unusual spectroscopic
behavior if the hydrogen is directed inwards toward the center of the
aromatic ring.
It is not clear, however, whether or not our target silicon-containing
cage compound will exhibit the same thermodynamic preference for the
"in" isomer as demonstrated by Pascal and co-workers for several carbon-
containing cage compounds.8 Preliminary calculations using semi-
empirical methods (MOPAC) show a preference for the "out" isomer in the
silicon-containing cage by 41 kcal/mol.9
There are numerous cyclophane compounds where a bridgehead
heteroatom is forced into proximity with an aromatic ring. Generally, the
reactivity of the heteroatom is affected by the interaction of its valence
electrons with the electrons of the k cloud of the aromatic ring. An
example of this is the diminished basicity of the nitrogen in a pyridinium-
cyclophane (Figure 1.7) as contrasted to pyridine.10
Figure 1.7. Pyridinoparacyclophane-1,9-diene.

In summary, the proposed synthesis of our silicon-hydride
cyclophane (1) is the first step towards a number of interesting and
potentially informative chemical studies in the area of organosilicon
chemistry. These studies will allow a direct contrast of a silicon-containing
molecule with its carbon isostere. Also, we plan to look at the orbital
hybridization of silicon as a function of strain imposed by being at the
bridgehead position of a cage molecule as well as the thermodynamic
stability of the "in" vs the "out" isomer for the silicon analog. Finally, we plan
to study interactions between silicon and the aromatic ring.

2. The Synthesis of Silicon Containing Cage Compounds
The first step in this project was to attempt to synthesize our target
compound (1). Further studies of the physical properties of 1 could
subsequently be undertaken.
2.1. Synthetic Strategy
Retrosynthetic analysis of our target compound (1) suggested
three basic synthetic pathways: assembly of the cage at the top, middle, or
bottom positions (Scheme I).

Scheme I Retrosynthetic Analysis of Target Compound (1).
(Top Assembly)
R=organometallic species:
(0CH2 M(+)
(Middle Assembly)
(Bottom, Aromatic Assembly)

Pascal's synthesis of the carbon analog (2) formed bonds in
middle positions (Scheme II). Although successful, his overall yield was
low (0.59%).1
Scheme II Pascal's Synthesis of Carbon Analog (2).
Br^\ '>S^Br
KOH, Ethanol
78C, 15hrs
5.5% yield
For our target compound (1), it was decided to attempt a synthesis
requiring assembly of the aromatic ring through cyclotrimerization of three
acetylene groups (Scheme I Bottom, Aromatic Assembly). This plan
seemed simplest in terms of preparation of precursors and number of steps

involved, and thereby held the greatest potential to give good yields of
Numerous transition metal catalysts have been reported to
cyclotrimerize both internal and terminal acetylene groups.11'17
Vollhardt's catalyst, CpCo(CO)2, has been documented to be a very
effective cyclotrimerization catalyst. For example, the reaction combining
1,7-octadiyne and bis(trimethylsilyl)acetylene in the presence of catalytic
amounts of CpCo(CO)2 produced 6,7-bis(trimethylsilyl)tetraiin in 83%
yield (Equation 2.1 ).11
In order to synthesize 1, a more detailed and complete synthetic
scheme was devised (Scheme III).

Scheme III Detailed Synthetic Plan for Target Compound (1)
1) nBuLi
2) Me3SiCI Cl ^
1) Mg, THF
2) 1/3 HSiCI3


The proposed synthesis begins with 5-chloro-1-pentyne which is
commercially available from the Aldrich Chemical Company. The terminal
acetylene of 5-chloro-1-pentyne would require protection before making
the Grignard reagent of the compound. Reaction with n-butyllithium and
chlorotrimethylsilane could readily achieve this. Subsequent preparation of
the Grignard reagent of the protected chloropentyne and reaction with
trichlorosilane would be expected to yield a silicon-carbon tripod.
Deprotection of the acetylene groups with silver nitrate would give us the
reagent we would then try to cyclotrimerize.
2.2. Early Synthetic Results
Reaction of n-butyllithium, and chlorotrimethylsilane gave the
protected haloalkyne in 74% yield (Equation 2.2, Experiment 4.2).
Cl ^ 1) nBuLi/THF, hexanes, -78C
H _____________________________________ '-/'
2) Me3SiCI/THF, 0C
3) NHiCI/HaO
:= SiMe3
However, preparation of the Grignard reagent from this protected
chloroalkyne did not occur readily (Equation 2.3).
n Mg/Et20 Cl Mq
C SiMe3 -1-i----------- 9^-=-SiMe3

Addition of several drops of dibromoethane served to initiate the Grignard
reaction, but the reaction did not sustain itself. These difficulties appeared
using both diethyl ether and THF as solvents.
We attempted to increase reactivity by converting the protected
chloroalkyne to the more reactive iodoalkyne (Equation 2.4,
Experiment 4.3).
Improved yields of the organometallic species were achieved with the
protected 5-iodo-1-pentyne, however, a new problem was created on
preparation of the Grignard. Approximately 50% conversion to the
coupled side product occurred along with the desired organometallic
reagent (Equation 2.5).

1 SiMe,
Along with difficulties in preparing the Grignard reagent, the reagent
did not readily yield trisubstitution of trichlorosilanes (Equation 2.6,
Experiment 4.4).

1) Mg,THF
2) 1/3 RSiCI3
\ SiMe3
R = H, Me
Reactions in both diethyl ether and THF produced mixtures of mono, di,
and ^substituted alkylsilanes. Use of catalytic amounts of CuCN served to
affect rapid and more complete trisubstitutions on trichlorosilanes,18
however, mixtures of mono, di, and tri substituted products were still
obtained. The reaction could effectively be pushed to trialkylation using
twice the molar equivalent of haloalkyne in the reaction procedure. With
the overall loss to "coupling," and the need for excess haloalkyne to push
the equilibrium to completion, the process was inefficient (-23% yield).
Because of these initial difficulties of making three carbon-silicon
bonds on a chlorosilane and because we still wanted to investigate the
suitability of cyclotrimerization for silicon cage formation, we decided to shift
our synthesis to a more readily accessible system, one containing three
oxygen-silicon bonds. Classical silicon chemistry describes facile
synthesis of trisubstituted alkoxysilanes in high yields by reactions of
trichlorosilanes with a variety of alcohols in the presence of equimolar
amounts of pyridine (Equation 2.7).19'20

Cl A Cl
benzene (reflux)
R=methyl,vinyl HO-R' = MeOH,
These reactions occur rapidly, in good yields, and under milder conditions
than trisubstituted carbon species are produced. The difference in
reactivity might be due to the stronger silicon-oxygen bond or to mechan-
istic differences in the reaction.
Reaction of 4-pentyn-1-ol with trichlorosilane and methyltrichloro-
silane produced a trialkoxysilane and a methyltrialkoxysilane in 56% and
77% yield, respectively (Equation 2.8, Experiments 4.5 4.5.2).
(Note The abbreviated names "methyl-tripod," "hydrogen-tripod," etc. will
be used to designate compounds of the type R-Si-(OCH2CH2CH2C=CH)3,
R = methyl, hydrogen, etc.).

benzene (reflux),
Cu powder
(-70% yield)
Conditions were mild enough that the terminal hydrogen of the alkyne did
not need to be protected. These two "tripod" compounds were used in the
next step of our synthesis: cyclotrimerization of the three terminal acetylene
groups to a benzene ring.
Several cyclotrimerization reactions were conducted with the
hydrogen-tripod and methyl-tripod. The reagents were added via syringe
pump to a flask containing catalyst in refluxing solvent. Three different
catalyst systems were evaluated in these early experiments: CpCo(CO)2
(Vollhardt's catalyst), Pd/C with MeaSiCI, and TiCIViBuaAI.12-13-15 None
of these early reactions produced the expected cage compounds although
they did give a variety of nonvolatile, polymeric solids which could not be
analyzed by GC or GC/MS.
At this point we decided to study the catalyst systems in greater detail
by using a model system of simpler molecules that modeled the functional
groups and steric variables presented by our tripod precursors.

2.3. Model System Studies
We studied three catalyst systems in greater depth to determine which
one might work best with our tripod precursors. The three catalysts chosen
were CpCo(CO)2 (Vollhardt's catalyst), Pd/C with TMSCI, and
TiCl4/iBu3AI.12-13'15 Our selection of these three catalysts was based on
ease of reported experimental procedures, cost, and availability of the
A series of trial cyclotrimerization reactions with 3-hexyne, and 1-
hexyne was executed with each catalyst giving the results tabulated in
Table I (Experiments 4.6 4.6.4). Two isomers were formed in
cyclotrimerization reactions of unsymmetrical alkynes. The isomers form as
a function of the directionality of the unsymmetrical alkynes as they align
with each other. When the three groups align in the same direction, the
symmetrically substituted 1,3,5 aromatic ring is formed. When they align in
different directions, the unsymmetrical 1,2,4 isomer forms. Statistically,
there are three times as many ways that the unsymmetrical isomer can form
over the symmetrical isomer, so expected isomer product ratios (1,2,4 :
1,3,5) would be 3:1.

Table I Results from Model Study Comparing Catalysts
-Beageut Catalyst Solvent Reaction conditions Product / Results
3-Hexyne CpCo(CO)2 Octane, 24 hr reflux 8% conversion to hexaethylbenzene
3-Hexyne CpCo(CO)2 Octane, 48 hr reflux 23% conversion hexaethylbenzene
3-Hexyne TiCl4/iBU3AI Benzene 12 hr reflux 100% conversion to hexaethylbenzene
3-Hexyne Pd/C (10%) THF, 12 hr 80% conversion to
MesSiCI reflux hexaethylbenzene
1-Hexyne TiCl4/iBU3AI Benzene 5 days reflux 68% conversion to combined 1.2.4 tri-nbutyl benzene and 1.3.5 tri-nbutyl benzene
1 -Hexyne CpCo(CO)2 Octane 2 days reflux 25% conversion to combined 1.2.4 tri-nbutyl benzene and 1.3.5 tri-nbutyl benzene
1-Hexyne Pd/C, Me3SiCI THF 3 days reflux No expected product
* Reported yields are based crudely on GC determination of relative peak areas without
an internal standard.

Results from these preliminary studies suggested that the
TiCMBuaAl catalyst was the best catalyst of the three in terms of yields,
and ability to cyclotrimerize terminal alkynes. Other catalysts were not
studied nor were results carefully quantified at this early stage of study.
A precedent for the synthesis of aromatic cage compounds using the
cyclotrimerization of three alkyne groups can be found in studies by
A.J. Hubert and coworkers. They successfully synthesized a series of
cyclophanes using TiCl^iBusAI as the cyclotrimerization catalyst
(Equation 2.9).15
n = 3-7
1,3,5 isomer
1,2,4 isomer
These workers used slow addition of trisubstituted aromatic tripods to
a catalyst solution via syringe pump under dilute conditions to minimize
intermolecular reactions between acetylene groups. Hubert's reported
yields of 16-53% of crude isomeric mixtures suggested the potential for

reasonable yields in our synthesis. We chose to use the TiCU/iBu3AI
catalyst system for our ensuing work.
A series of reactions was designed and conducted to study the effects
of functional groups and steric hindrance on cyclotrimerization reactions.
Simple alkynes were found to be successfully cyclotrimerized in good yield
with the TiCMBuaAl catalyst (Table I). 5-Chloro-1-pentyne was
cyclotrimerized readily to the two expected isomers in 60% combined yield
(Equation 2.10).
Benzene fl, 24hrs 60% yield of mixed isomers
1 :1.5
However, 4-pentyne-1-ol failed to cyclotrimerize under these conditions. A
polymeric residue was produced in the reaction flask and no starting
material remained (Equation 2.11).
HC=CCH2CH2CH2OH ---------------- Decrease in starting material;
TiCi4/iBu3Ai polymerization?
Benzene No expected product. (2.11)
Addition of a bulky trimethylsilane group on the terminal acetylene of 5-chloro-

1-pentyne appeared to hinder its reactivity (Equation 2.12). None of the
expected cyclotrimerization products was produced and starting material
Me3SiC=CCH2CH2CH2CI -------------------- No reaction
Benzene U, 24hrs (2.12)
The analogously protected 4-pentyne-1 -ol also failed to cyclotrimerize, yet
starting material noticeably diminished (Equation 2.13). This result suggested
that the unprotected alcohol functionality might be undergoing a side reaction
with the catalyst system.
------------> Decrease of starting material;
TiCi4/iBu3Ai no expected product.
Benzene U,24hrs (2.13)
Protection of the oxygen functionality as well as the acetylene group with
trimethylchlorosilane and subsequent reaction with catalyst gave virtually
no reaction. Starting material diminished only slightly (Equation 2.14).
Me3SiC=CCH2CH2CH20SiMe3 --------------- No reaction
Benzene N, 24hrs (2.14)

A reaction with the acetylene group unprotected and the alcohol group
protected resulted in a 10% yield of cyclotrimerization products (Equation
TiCl4/iBu3AI +1,2,4 isomer
Benzene tl, 24hrs 10% yield (total)
some polymerization,
some hydrolysis
Protection of the alcohol functionality with a larger protecting group
(t-butyldimethylchlorosilane) gave a substrate that successfully
cyclotrimerized in 65% yield when introduced to the catalyst solution
(Equation 2.16).
TiCl4/iBu3AI +1,2,4 isomer
Benzene W, 24hrs 65% yield (total)

The results of these model studies showed that the TiCLj/iBusAI
catalyst system was sensitive to the presence of oxygen and to the pre-
sence of bulky groups on the alkyne. These results suggested that a tripod
with sufficiently protected oxygen groups might successfully cyclotrimerize.
2.4. Preparation and Cyciotrimerization of t-Butyl-tripod
A siloxy-tripod was prepared with a t-butyl group on silicon to protect
the oxygen functionalities. This t-butyl-tripod was prepared in the same
fashion as other silicon-oxygen tripods using t-butyltrichlorosilane, 4-
pentyn-1-ol, and pyridine (Equation 2.17, Experiment 4.5.3).
benzene (reflux),
Cu powder
(-70% yield)
-/1 -
\ ^
Reaction of t-butyl-tripod with the TiCU/iBusAI catalyst system re-
sulted in successful cyciotrimerization, as assessed by initial GC/MS anal-
ysis (Equation 2.18, Experiment 4.7). The starting material had disap-
peared, and two new peaks with similar retention times and nearly identical

mass spectra appeared. The base peak, m/z=277 (100%) in each
spectrum represented a diagnostic fragment from the loss of a t-butyl group.
The two isomer GC peaks were observed in a ratio of of 7:1 in favor of the
1,2,4 isomer (3) (as determined by later characterization). Purification
and separation of the two cage products was effected by flash chromato-
graphy. The combined yield of the two products was 30%. Both com-
pounds were white crystalline solids which were characterized by 1H and
13C NMR, elemental analysis, and single crystal X-ray structure analysis.
The 1,2,4 isomer (3) was recognized to be a chiral molecule, and
was assumed to be a racemic mixture of the two enantiomers. Separation
of the two enantiomers was not attempted.
2.4.1. Discussion of NMR Spectra of t-Butyl-siloxycyclophanes
The 1H NMR spectrum of the 1,2,4 isomer (3) (300 MHz, CDCI3,
Appendix I.a) was very complex. Some recognizable features are a singlet
at 0.79 ppm for the nine equivalent hydrogens of the t-butyl group, a diag-
1,2,4 cage
1,3,5 cage

nostic abx pattern for the three nonequivalent protons in the aromatic
region, and some regional grouping of the complex multiplets according to
chemical shift values of the 18 magnetically nonequivalent hydrogens on
the three bridging arms. The 13C NMR spectrum of the 1,2,4 isomer (75
MHz, CDCI3, broadband decoupling, Appendix l.c) shows 17 peaks
representing the 17 nonequivalent carbons of the molecule.
The 1,3,5 isomer (4) showed a simpler 1H NMR spectra (300 MHz,
CDCI3, Appendix l.b) as would be expected for a symmetrically substituted
molecule.* A large singlet appears for the nine equivalent protons of the
t-butyl group. A singlet appears at 6.88 ppm representing the three
equivalent aromatic protons. Each bridging arm of the symmetrical
molecule is equivalent, so the 18 hydrogens appear cleanly as three
groupings at 3.70-3.76 ppm (m) 2.68 ppm (t), and 1.87-1.96 ppm (m).
The "multiplet" at 3.70-3.76 ppm appears as a distorted triplet (the
expected multiplicity if the molecule was fully symmetrical), suggesting a
degree of nonequivalency of the bridging arms as they twist slightly to the
most stable molecular conformations. The 13C NMR spectrum of the 1,3,5
isomer (75 MHz, CDCI3, broadband decoupled, Appendix l.d) shows
seven singlets representing the seven magnetically nonequivalent carbons
in the symmetrical structure.
*Note Although the 1,3,5 isomer is symmetrically substituted on the
aromatic ring, the strain and constraints of the cage structure result in a
slight spiralling of the three bridging arms. This twisting, if extreme enough,
would have the effect of making all six of the hydrogens along one bridging

arm diastereotopic, yielding a more complex 1H NMR spectrum. Similarly,
the 13C NMR spectrum would be expected to be more complex if
conformational transitions were slower than the time scale of NMR
relaxation times. Our 1,3,5 isomer showed little nonequivalency.
2.4.2. Discussion of X-ray Crystal Structures of
Single crystals of both isomers were grown in acetone by cooling the
solution slowly from boiling to room temperature. ORTEP projections of the
two X-ray structures are shown as Figure 2.1
1,2,4 isomer
t-butyl- siloxycyclophane
1,3,5 isomer
Figure 2.1. Anisotropic ellipsoid plots of 1,2,4 and 1,3,5 t-butyl-
siloxycyclophanes (with adopted numbering scheme).21

The solution to the X-ray structure of the 1,2,4 isomer (3) revealed
that the compound had crystallized as a single enantiomer. The X-ray
structure of (3) shows distortion of the benzene ring from planarity at the
positions of substitution of the bridging arms. The silicon atom at the
bridgehead shows no evidence of distortion with near tetrahedral bond
angles to each of its substituents. However, the silicon-oxygen-carbon
bonds are apparently distorted with angles between 130 and 137. The
methylene groups in the bridging arms are bent at angles of 115-117. A
relatively large opening into the cage is apparent between bridging arms
substituted at the 1 and 4 carbons of the aromatic ring.
The X-ray structure of the 1,3,5 isomer (4) shows planarity of the
benzene ring, silicon with tetrahedral geometry, distorted Si-O-C bond
angles of 137-140, and distorted C-C-C bond angles of the methylene
bridging arms of 120-127.

3. Summary and Future Work
3.1. Summary
Two novel silicon-containing cage compounds 3 and 4, have
been synthesized and fully characterized by 1H and 13C NMR spectro-
scopy, elemental analysis, mass spectrometry, and single crystal x-ray
The compounds are isomers with the molecular formula C-igH3o03Si and
were produced in the same synthetic procedure. The cage compound
produced in higher yield (3) contains an unsymmetrically substituted
aromatic ring (1,2,4 substitution pattern) while the other isomer (4)
contains a symmetrically substituted aromatic ring (1,3,5 pattern). The
mass spectra of both isomers are similar, showing a diagnostic base peak

of M-57 (loss of t-butyl). The 1,2,4 substituted cage isomer (3) is chiral
by nature of its unsymmetrical substitution pattern, and is assumed to have
formed as a racemic mixture.
Both cage compounds show some unusual structural features in the
larger than expected bond angles of the Si-O-C and C-C-C bonds of the
bridging arms (by single crystal x-ray analysis). Additionally, the 1,2,4
isomer shows a distorted aromatic ring bent slightly into a "boat-like"
The 1H NMR spectrum of the 1,2,4 substituted cage is very complex.
Due to the lack of symmetry in the molecule, all of the 18 methylene
hydrogens of the bridging arms are magnetically nonequivalent. The three
aromatic hydrogens are magnetically nonequivalent and show an "abx"
splitting pattern, diagnostic of their substitution pattern on the aromatic ring.
The nine hydrogens of the t-butyl group on the top of silicon show a narrow
singlet as expected. The 13C NMR spectrum shows 17 unique reso-
nances, corresponding to 17 magnetically nonequivalent carbons on the
The 1H NMR spectrum of the 1,3,5 substituted cage is very simple
by contrast. A singlet appears in the aromatic region representing the three
magnetically equivalent aromatic hydrogens. A large singlet appears for
the nine hydrogens of the t-butyl group. The 18 hydrogens on the bridging
arms form three magnetically different groups due to the high symmetry of
the molecule. The 13C NMR spectrum shows seven unique resonances as
expected for the symmetrical isomer.

3.2. The Purpose of this Project Retrospect and Future
This work represents only an initial phase of a broader project to
study the physical properties of silicon in a unique cage environment.
Successful synthesis of the original synthetic target (1) or other similar
silicon-containing cage compounds is the first necessary step before further
studies of such compounds can take place.
We have not yet successfully synthesized our original synthetic
target (1), and therefore cannot yet make comparisons with its carbon
analog (2). Work is still in progress towards this goal. There presently are
no carbon analogs to the t-butyl-siloxycyclophanes we have synthesized,
so comparisons with carbon isosteres cannot be done at this point of the
The question of orbital hybridization of silicon in a strained cage
environment can be addressed in part with our work up to this point. For
silicon-oxygen containing cages, the distortion appears in other parts of the

molecule in preference to distortion at silicon. Both of our cage compounds
show silicon in an undistorted tetrahedral conformation while showing
distortions of the Si-O-C bond angles, the C-C-C bond angles of the
bridging arms, and in the planarity of the aromatic ring (1,2,4 isomer only).
These characteristics will be studied further as other silicon-oxygen cage
compounds are produced and characterized by this lab.
The synthesis of our silicon-oxygen cage compounds has required
the presence of a sterically bulky t-butyl group on silicon to prevent
interaction of the catalyst with the oxygen groups. The large size of this
group forces the cage into an "out" conformation. 1H NMR studies where
a silicon hydride is placed inwards towards the anisotropic ring current are
not yet available to us.
Similarly, the presence of the large t-butyl group on our cages
precludes the question of the thermodynamic stability of the "in" vs "out"
isomer. With the t-butyl group present, the only steric possibility is for an
"out" isomer. Larger sized cages might allow the possibility for a t-butyl
group to turn inward toward the aromatic ring. This possibility is presently
being explored.
The interaction of silicon at the bridgehead position with the basal
aromatic ring is a quality we can begin to study with our present molecules.
We plan to study these features using 29Si NMR spectroscopy and UV-Vis
spectroscopy. Comparisons with other similar molecules would add
perspective to this data, and to that end, we are presently involved in the
synthesis of a series of t-butyl-siloxycyclophanes of varying sizes (Figure

n = 2 n = 3 n = 4 n = 9
Figure 3.1. Series of different sized 1,3,5 t-butyl-siioxycyclophanes.
(1,2,4 isomers have also been prepared.)
The methodology developed in the synthesis and characterization of 3
and 4 has led to the successful synthesis of all of the molecules in the
above series. Work is currently underway towards their full character-
ization. Additional comparisons of these molecules will be carried out with
single crystal X-ray structures, 1H NMR, and 13C NMR.
Our success in developing silicon-oxygen containing cage com-
pounds over silicon-carbon containing cages has suggested the possibility
of synthesizing a "siloxy-hydride-cyclophane" (5).

Investigations are presently underway in this lab towards finding a
cyclotrimerization catalyst that is less sensitive to the presence of oxygen.
Alternative synthetic routes that will allow placement of a hydride on silicon
by displacement of other groups are also being investigated.
Work also continues towards our original synthetic target (1) using
our original synthetic scheme. Additionally, other synthetic paths towards
this compound are being investigated.

4. Experimental Section
4.1. Geaeral Procedures and Instruments Used
For all procedures, glassware was oven dried for 24 hours at 140C
and set up under a flow of argon. A positive flow of argon was maintained
in the glassware throughout the experiment. Chemicals used for this work
were purchased from Aldrich Chemical Company unless otherwise noted.
Reagents were used as received unless otherwise noted.
Instruments used in this work include:
a Hewlett Packard Gas Chromatograph Series # 5890 with FID
a Hewlett Packard Mass Selective Detector Series # 5970 (paired
with HP GC Series # 5890), El source,
a Bruker NR-80 80 MHz NMR Spectrometer,
a Varian VXR-300 300 MHz NMR Spectrometer (Boulder
X-ray crystal structures were determined by the X-ray Crystallography Lab
at the University of Colorado at Boulder. Elemental analyses were
performed by Huffman Labs, Golden, Colorado. Initial NMR spectra were
taken on the Bruker NR-80. Final NMR spectra of purified compounds were
taken on the Varian VXR-300 in Boulder and are the spectra reported in
this work unless otherwise noted. 1H NMR samples were prepared in

CDCI3 (referenced to the CHCI3 impurity at 5 = 7.24 ppm) or acetone-d6
(referenced to the CHD2COCD3 impurity at 2.05 ppm). 13C NMR samples
were prepared in CDCI3 and referenced to CDCI3 at 5 = 77.0 ppm. All
13C NMR spectra reported in this work were decoupled with broad band
4.2. Preparation of 5-Chioro-1-trimethylsilyl-1-pentyne
1) nBuLi/THF, hexanes, -78C
-------------------------- '^-^-SiMe3
2) Me3SiCI/THF, 0C J
3) NH4CI/H20 fr)
A 500 ml, 3-necked, 24/40 round bottom flask was fitted with a
125 ml pressure equalizing addition funnel and a condenser. A septum
was fit over the addition funnel. 5-Chloro-1-pentyne (25.0 g, 0.24 moles)
and 150 ml THF (freshly distilled from Na metal) was added to the flask
arid 100 ml of 2.5 M n-BuLi in hexanes (0.25 moles) was added to the
addition funnel using double needle transfer. The flask was cooled to
-78C with a dry ice-acetone bath whereupon the n-BuLi was added at a
rate -2 drops/second, stirred for one hour, and warmed to room
temperature. Chlorotrimethylsilane (26.5 g, 0.24 moles), in 20 ml THF
was added at a rate of -1 drop per second at 0C after which the reaction
contents were stirred for one hour at room temperature.

The reaction was worked up with 10% aqueous NH4CI, and
extracted four times with diethylether. Removal of solvent by rotary
evaporation yielded 46.1 g of clear tan liquid. This crude product was
purified by distillation (bp=45C, 2.4 Torr) yielding 31.5 g of 5-chloro-1-
trimethylsiiyl-1-pentyne (74.2% yield).
Characterization of the product was carried out by GC/MS, 1H NMR,
and 13C NMR. The MS showed a base peak at m/z = 93 (100),
corresponding to a rearranged fragment of [(CH3)2SiCI]+. The 1H NMR
spectrum (300 MHz, CDCI3) showed 4 distinct groupings: 5 = 0.15 ppm
(s, 9 H), 1.96 ppm (p, 2 H, J = 6.6 Hz), 2.41 ppm (t, 2 H, J = 6.9 Hz), and
3.64 ppm (t, 2 H, J = 6.3 Hz). The 13C NMR spectrum (75 MHz, CDCI3)
showed 6 unique resonances at 5 = 0.16, 16.99, 31.29, 43.00, 85.18,
104.77 ppm.
4.3. Preparation of 5-lodo-1-trimethylsilyl-1-pentyne
Clv^=- SiMe3 Nal
The Finkelstein Reaction was used to convert 5-chloro-1-
trimethylsilyl-1 -pentyne to the corresponding iodide. Sodium iodide
(21.1g, 0.14 moles, Fisher Scientific) was added to a solution of 5-chloro-
1 -trimethylsilyl-i -pentyne (18.5 g, 0.11 mols) and 2-butanone (200 ml) in
a 500 ml round bottom flask fitted with a condenser and argon inlet. The
contents were set to reflux (71 C) for 15 hours.

Work-up entailed filtration of solid from the liquid portion of reaction
mixture, subsequent washing of filtrate with deionized water, back
extraction with 2-butanone, and removal of solvent by rotory evaporation to
yield -28 g of crude product. Purification by distillation (bp=77C, P=3.1
Torr) through a 5 cm column packed with glass wool yielded 26.1 g of 5-
lodo-1 -trimethylsilyl-1 -pentyne (92% yield).
Characterization of product was carried out by GC/MS and
comparison to data collected for the already characterized 5-chloro-1-
trimethylsilyl-1-pentyne. GC/MS analysis with identical temperature and
mass acquisition parameters as applied to the chloride gave a peak with
retention time of 8.48 minutes as contrasted to 7.30 minutes for the chloride.
The MS showed a base peak at m/z = 185 (100) corresponding to a
rearranged fragment of [(CH3)2Sil]+ and a molecular ion fragment at
m/z = 266 (25).
4.4. Preparation of the Grignard Reagent of 5-Chloro-1-
trimethylsilyl-1-pentyne and Reaction with Trichlorosilanes
1) Mg, THF j
2) 1/3 RSiCI3
R = H, Me

A Grignard reagent was prepared from 5-halo-1 -trimethylsilyl-1 -
pentyne using 1-2 molar equivalents of commercial grade magnesium
turnings, diethylether/THF, and 0.5-1 molar equivalent of dibromoethane as
an entrainer. Copper(l) cyanide (0.5% molar equivalent of silane) was
added to the Grignard reagent followed by 1/3 molar equivalent of
trichlorosilane or methyltrichlorosilane diluted in THF. These reactions
were allowed to stir from 0.5-14 hours adjusting time as a variable in
different trials. Progress of the reactions was followed by GC.
Reactions were quenched with 10% aqueous NH4CI solution.
Extractions with ether, and subsequent removal of solvent by rotory
evaporation yielded crude products which were analyzed by GC and
GC/MS. In one case, further purification of 0.3 g of crude tris(5-trimethyl-
silyl-4-pentynyl)silane by flash chromatography yielded a product that was
roughly 78% pure (estimated by GC relative areas without internal
Characterization of the product was carried out by GC/MS, and 1H
NMR (for the reaction where the product was purified). The MS showed a
base peak at m/z = 73 ([Me3Si]+, 100) with significant peaks m/z = 373
(M-TMS, 0.5), and 307 (M-139, 30). The 1H NMR spectrum (300 MHz,
CDCI3) showed five groupings of peaks: 5 = 0.01 ppm (m, 1 H), 0.10 ppm
(s, 27 H ), 0.64 0.72 ppm (dd, 6 H, J1 = 9 Hz, J2 = 3 Hz), 1.46-1.58 ppm
(m, 6 H), and 2.20 ppm (t, 6 H, J = 8.4 Hz).

4.5. General Procedure for Preparation of Tri-alkoxysilanes*
3 pyridine
benzene (reflux)
Cu powder
R = H, Me, tBu
(2.8) &
Note All alkoxysilanes prepared for this work were made using this same
procedure. The appropriate trichlorosilane was used to prepare the
particular derivative.
A 500 ml, 3 necked flask fitted with a condenser, addition funnel,
magnetic stir bar, and argon inlet was charged with benzene (200 ml),
trichlorosilane (0.08 mol), dried pyridine (18.3 g, 0.23 mol, distilled from
KOH) and copper powder (0.31 g, 4.90x1 O'3 mol, Baker). To the addition
funnel was added 4-pentyne-1-ol (19.4 g, 0.23 mol) dissolved in 30 ml of
benzene. The addition proceeded for -0.5 hour at a rate of 1 drop/sec at
0C with vigorous stirring, during which a white pale green precipitate
appeared. The reaction mixture was refluxed overnight.
The work-up entailed filtration of the precipitate from the liquid portion
of the reaction mixture through a short column of silica gel. Glassware was
rinsed six times with hexanes. All rinsings were forced through the
precipitate on the short silica gel column (using argon flow). Rinsings were

collected, and solvent was removed by rotary evaporation to yield the crude
4.5.1. Preparation of Tris(4-pentyn-1-oxy)silane
* alHcT^V] pyrldine >
benzene (reflux)
Cu powder
The product (see Experimental 4.5) was obtained in -70% purity after
filtration and rotary evaporation of solvents. Experimental yield was -55%.
Characterization was carried out by GC, GC/MS, and 1H NMR. GC anal-
ysis showed loss of starting material, and appearance of a new peak at
higher retention time. The GC/MS spectrum showed no diagnostic masses.
The 1H NMR spectrum (80 MHz, acetone d6) showed three complex
groupings at 8 = 1.6-2.0 ppm (m, 6 H), 2.2 2.5 ppm (m, 9 H), and 3.8 -
4.1 ppm (m, 7 H). The spectrum was consistent with the expected product,*
but was too complex to allow confident identification at the time.
*Note The 1H NMR spectrum of triethoxysilane, HSi(OCH2CH3)3,22 a
similar compound to the reaction product, shows a chemical shift value for

the silicon hydride at 5 = 4.23 ppm. This is consistent with the overlapping
multiplet observed in the range of 3.8 4.1 ppm (7 H) for our product.
4.5.2. Preparation of Methyl-tris(4-pentyn-1-oxy)silane
The product (see Experimental 4.5) was purified by filtration and
removal of solvents under reduced pressure. Experimental yield was 77%,
Characterization was effected by GC/MS, and 1HNMR. TheGC/MS
spectrum was complex with no diagnostic masses at high abundance. The
1H NMR spectrum (80 MHz, acetone d6), showed four distinct groupings:
5 = 0.10 ppm (s, 3 H), 1.55-1.95 ppm (m, 6 H), 2.16-2.45 ppm (m, 9 H),
3.85 ppm (t, 6 H, J = 6.8 Hz).
benzene (reflux)
Cu powder

4.5.3. Preparation of t-Butyl-tris(4-pentyn-1-oxy)silane
Filtration of reaction mixture (see Experimental 4.5) yielded -25 g of
crude liquid. Purification by flash chromatography yielded 18.1 g of clear,
oily liquid (71% yield). Characterization of the product was carried out by
GC/MS, 1H NMR, and elemental analysis. The GC/MS spectrum was
complex with no diagnostic masses at high abundance. The 1H NMR
spectrum (80 MHz, acetone-dg) showed four distinct groupings: 5 = .98
ppm (s, 9 H), 1.6-2.0 (m, 6 H), 2.2-2.5 (m, 9 H, overlap with alkynyl H),
and 3.94 (t, 6 H, J=6.0 Hz). Results of the elemental analysis; calc'd for
CigHaoOgSi: C, 68.22; H, 9.04. Found: C, 68.08; H, 9.12 (V2O5 used as
oxidation catalyst).
(-70% yield)
benzene (reflux),
Cu powder

4.6. General Notes for Model System Reactions
The model system reactions were done in a non-quantitative fashion
with the aim of studying relative reactivities and observing how these
reactions worked. Yields were estimated by GC analysis, comparing peak
size of starting material at the start of the reaction with peak size after
various times. Internal standards were not used for yield calculations nor
were isolated yields determined for these reactions.
These reactions were all done on a small scale. For all of these
reactions, a 25 ml 3-necked round bottom flask was fitted with a condenser,
argon inlet, a septum covered glass sleeve, and a magnetic stirring bar.
Catalyst and solvent were added to the flask under a flow of argon. The
alkyne (0.5 g) was added rapidly by Pasteur pipet through the condenser
(-60 seconds).
The progress of the reactions was monitored by the loss of starting
material and appearance of new products. New peaks appearing on the
GC were analyzed by mass spectrometry. The appearance of two, closely
spaced isomer product peaks on both the GC and GC/MS was taken as
evidence of cyclotrimerization.
In reactions where further confirmation of product production was
needed, reaction mixtures were filtered through 1 cm of silica gel on a glass
frit, solvent was removed by evaporation under reduced pressure, and the
remaining product was analyzed by 1H NMR. The product at this point
contained both cyclotrimerization isomers, some solvent, some starting
material, and other reaction impurities. Relatively large, new resonances

in the aromatic region of 1H NMR spectrum were taken as evidence of
successful cyclotrimerization to an aromatic ring.
4.6.1. Preparation of TiCU/iBusAI Catalyst Stock Solution
Benzene or hexanes (Aldrich # 17,891-8,100 ml) were added to a
clean, oven-dried bottle which was then covered with a rubber septum.
Titanium(IV) chloride (1.0 ml, 9.12 x 10*3 moles) was added to the solvent
via syringe through the septum. Triisobutylaluminum (5.0 ml, 1.98 x 10*2
moles) was transferred by syringe to the solution. The contents were
mixed by successive drawings and ejections in the syringe used to transfer
the iBusAI.
4.6.2. General Procedure for Model System Reactions with
The stock solution of TiCU/iBusAI (12 ml), prepared as described in
Experiment 4.6.1, was added to the reaction flask. After addition of the
alkyne (0.5 g) the reaction was refluxed overnight.
4.6.3. General Procedure for Model System Reactions with
Six microliters of CpCo(CO)2 and 5 ml of octane were added to the
flask under a flow of argon. After addition of the alkyne (0.5 g), the reaction
flask was refluxed and stirred overnight. Some reactions, were illuminated
[GE ELH 300 Watt light bulb, d= 4 cm from flask center, power regulated by
variable transformer to 50V] in addition to reflux.

4.6.4. General Procedure for Model System Reactions with
Pd/C, Me3SiCI
THF (5 ml, distilled from sodium, stored over 4A molecular sieves)
and 10% Pd/C (0.05 g) were added to the reaction flask followed by
trimethylchlorosilane (0.54 g, 5x1 O'3 moles, Petrarch). After the alkyne was

added (-0.5 g equimolar to MeaSiCI), the mixture was refluxed and
stirred overnight.
4.7. Preparation of t-Butyl-siloxycyclophanes
Hexane (reflux)
Syringe Pump, 36hrs
(-30% yield)
1,2,4 cage
A 1 L, 3 necked, 24/40 round bottom flask was charged with hexane
(600 ml), TiCU (1.5 ml, 0.014 mole), and iBuaAl (7.5 ml, 0.030 mole). The
flask was fitted with a 50 ml Soxhlet extractor and a condenser on top of the
extractor. Reagent (t-butyl-tripod, 15.0 g, .045 mol) diluted with hexane
(25.0 ml) was placed into a 50 ml Luer lock syringe set up on a syringe
pump (36 hr addition). A long 1 mm diameter plastic tube led from the

syringe into the Soxhlet extractor.* The flask was stirred with a magnetic
stir bar and heated to reflux, whereupon the syringe pump was turned on.
*Note Addition of reagent into the fill/empty cycle of theSoxhlet extractor
further diluted the reagent before it contacted the catalyst. Use of the
syringe pump along with dilute addition conditions was used to increase
the chances of intra-molecular cyclotrimerization as opposed to
intermolecular cyclotrimerization.
After addition was complete, the flask was cooled, and the contents
were filtered through a 2" column of silica gel to remove polymeric and
inorganic residues. Solvent was removed by rotary evaporation yielding
12.5 g of tan colored, clear liquid, that later partially solidified. The isomers
were separated by flash chromatography (1.5" column; 9" silica gel, 95%
heptane 5% ethyl acetate) yielding pure 1,2,4 isomer (1.30 g), pure 1,3,5
isomer (0.56 g), and a fraction that was a purified mixture of the two
isomers (2.50 g). The overall combined yield of both isomers was 30%.
The ratio of 1,2,4 :1,3,5 isomers (by GC analysis) was 7:1.
The 1,2,4 cage compound was a crystalline white solid with
m.p. = 95-96C. The compound was recognized to be chiral, however,
separation of the enantiomers was not attempted. The assumedly racemic
mixture was further characterized by GC/MS, 1H NMR, 13C NMR,
elemental analysis, and single X-ray crystal structure analysis. The MS
showed a base peak at m/z=277 (100%) representing the loss of the t-
butyl group. The 1H NMR spectrum (300 MHz, CDCI3, Appendix I.a)

was very complex, but showed 5 groupings: 5 = 0.79 ppm (s, 9 H), 1.6-2.1
ppm (m, 6 H), 2.6-3.2 ppm (m, 6 H), 3.7-4.2 ppm (m, 6 H), and 7.0-7.2
ppm (abx pattern, 3 H). The 13C NMR spectrum (75 MHz, CDCI3i
Appendix l.c) showed 17 unique peaks: 5= 19.40,26.29,28.57,30.09,
30.33, 32.30, 34.11,35.46, 63.65, 63.76, 65.73, 125.27, 129.69, 129.98,
136.89, 137.46, and 138.62 ppm. (See section 2.4.1 for a discussion of the
1H and 13C NMR spectra.) Elemental analysis gave the following results;
calcd for Ci9H3o03Si: C, 68.22; H, 9.04. Found: C, 67.87; H, 8.90. The
solid was recrystallized in acetone and submitted for single crystal X-ray
structure analysis. The compound recrystallized as a single enantiomer.
(See section 2.4.2 for a discussion of the crystal structure.) A full data set
for the crystal structure is referenced by Damrauer and coworkers.21
The 1,3,5 isomer was a white crystalline solid with m.p. = 112-115.
The compound was further characterized by GC/MS, 1H NMR, 13C NMR,
elemental analysis, and single X-ray crystal structure analysis. The MS
showed a base peak at m/z=277 (100%) representing the loss of the
t-butyl group. The 1H NMR spectrum (300 MHz, CDCI3i Appendix I.b)
showed 5 groupings: 8 = 0.73 ppm (s, 9 H), 1.897 1.96 (multiplet, 6 H),
2.68 (t, J=6.6 Hz, 6 H), 3.70 3.76 (m, 6 H), 6.88 (s, 3 H). The peak at
3.73 ppm appears as a deformed triplet (J=4.4 Hz.). The 13C NMR
spectrum (75 MHz, CDCI3i Appendix l.d) showed seven unique carbons:
5= 19.70, 26.21,32.46, 34.66, 62.91, 128.30, and 139.85 ppm. (See
section 2.4.1 for a discussion of the 1H and 13C NMR spectra.) Elemental
analysis gave the following results; calc'd for CigH3o03Si: C, 68.22; H,
9.04. Found: C, 67.19; H, 8.83. The solid was recrystallized in acetone.

Single crystals were submitted to the X-ray crystallography lab at the
Boulder campus for structure analysis. (See section 2.4.2 for a discussion
of the crystal structure.) A full data set for the crystal structure is referenced
by Damrauer and coworkers.21

(1) Pascal, R. A., Jr.; Grossman, R. B.; Van Engen, D. J. Am. Chem. Soc.
1987, 109, 6878-6880.
(2) LEsperance, R. P.; West, A.P., Jr.; Van Engen, D.; Pascal, R. A., Jr. J.
Am. Chem. Soc. 1991,773, 2672-2676. This reference reports the
successful synthesis and characterization of a silicon-containing
cyclophane prepared by base-promoted condensation of tris (2-
mercaptophenyl)silane with 1,3,5-tris(bromomethylbenzene). In this
molecule, the hydrogen on the silicon atom is directed "inward" towards
the aromatic ring.
(3) Eaborn, C. Qraanosilicon Compounds: Academic: New York, 1960;
p 490.
(4) Lide, David R. CRC Handbook of Chemistry and Physics 71st ed.:
CRC: Boston, 1990; pp 9-1 9-5.
(5) Moberg, W.K.; Basarab, G.S.; Cuomo, J.; Liang, P. H. Synthesis and
Chemistry of Agrochemicals ; American Chemical Society, Symposium
Series No. 355, 1987, pp 288-301; U.S. Patent 4 510 136, 1985.
(6) Boer, P. F.; Turley, J. W. J. Am. Chem. Soc. 1969,97, 4134-4139.
(7) Homer, G. D.; Sommer, L. H. J. Am. Chem. Soc. 1973, 95, 7700-7707.
(8) Pascal, R. A., Jr.; Winans, C. G.; Van Engen, D. J. Am. Chem. Soc.
1989, 111, 3007-3010.
(9) Damrauer, R., University of Colorado at Denver, unpublished results.
(10) Boekelheide, V.; Galuszko, K.; Szeto, K.S. J. Am. Chem. Soc. 1974,
96, 1578-1581.
(11) Hillard, R.L., III; Vollhardt, K. P. C. J. Am. Chem. Soc. 1977,
99, 4058-4069.
(12) Berris, B. C.; Hovakeemian, G. H.; Lai, Y. H.; Mestdagh, H.; Vollhardt,
K. P. C. J. Am. Chem. Soc. 1985, 107, 5670-5687.

(13) Jhingan, A. K.; Maier, W. F. J. Org. Chem. 1987, 52, 1161-1165.
(14) Williams, A. C.; Sheffels, P.; Sheehan, D.; Livinghouse, T.
Organometallics 1989, 8, 1566-1567.
(15) Hubert, A.J.; J. Chem. Soc. 1967,6-10.
(16) Meriwether, L.S.; Colthup, E.C.; Kennerly, G.W.; Reusch, R.N. J. Org.
Chem. 1961,26, 5155-5163.
(17) Franzius.B.; Canterino, P.J.; Wickliffe, R.A. J. Am. Chem. Soc. 1959,
81, 1514.
(18) Lennon, P.J.; Mack, D.P.; Thompson, Q.E. Organometallics 1989,8,
(19) Tamborski, C.; Post, H. W.; J. Org. Chem. 1952, 17, 1400-1405.
(20) Nagel, R.; Tamborski, C.; Post, H. W.; J. Org. Chem. 1951,16,
(21) Damrauer, R.; Hankin, J.A.; Haltiwanger, R.C. Organometallics
1991,10, 3962-3964.
(22) Pouchert, C. J., Ed.; The Aldrich Library of NMR Spectra: Aldrich
Chemical: Milwaukee, 1983; ed II, spectrum 2-987.

Appendix I. NMR Spectra

Appendix l.a. 1H NMR (300 MHz, CDCI3) 1,2,4 t-Butyl-

Appendix l.b. 1H NMR (300 MHz, CDCI3) 1,3,5 t-Butyl-

Appendix l.c. 13C NMR (75 MHz, CDCI3) 1,2,4 t-Butyl-

Appendix l.d. NMR (75 MHz, CDCI3) 1,3,5 t-Butyl-