A photo Lewis acid generator (PhLAG): Controlled photorelease of B(C 6F5)3

Article abstract of DOI:10.1021/ja3042977

A molecule that releases the strong organometallic Lewis acid B(C ₆F₅)₃ upon irradiation with 254 nm light has been developed. This photo Lewis acid generator (PhLAG) now enables the photocontrolled initiation of several reactions catalyzed by this important Lewis acid. Herein is described the synthesis of the triphenylsulfonium salt of a carbamato borate based on a carbazole function, its establishment as a PhLAG, and the application of the photorelease of B(C₆F₅) ₃ to the fabrication of thin films of a polysiloxane material.

Full text of DOI:10.1021/ja3042977

Communication
pubs.acs.org/JACS
A Photo Lewis Acid Generator (PhLAG): Controlled Photorelease of
B(C6F5)3
Andrey Y. Khalimon,† Warren E. Piers,*,† James M. Blackwell,§ David J. Michalak,§ and Masood Parvez†
†Department of Chemistry, University of Calgary, 2500 University Drive NW, Calgary, Alberta, T2N 1N4 Canada
§Components Research, Intel Corporation, 2511 NW 229th Avenue, Hillsboro, Oregon 97214, United States
S
* Supporting Information
and polysiloxane scaffolds via the Piers−Rubinsztajn reaction,
eq 1,13,14 a process for which controlled photogeneration of
ABSTRACT: A molecule that releases the strong organo-
metallic Lewis acid B(C6F5)3 upon irradiation with 254 nm
light has been developed. This photo Lewis acid generator
(PhLAG) now enables the photocontrolled initiation of
several reactions catalyzed by this important Lewis acid.
Herein is described the synthesis of the triphenylsulfonium
salt of a carbamato borate based on a carbazole function,
its establishment as a PhLAG, and the application of the
photorelease of B(C6F5)3 to the fabrication of thin films of
a polysiloxane material.
B(C6F5)3 would be particularly attractive in thin film synthesis
and patterning processes since siloxane-based films find many
uses in the semiconductor industry.15
B(C6F )3
5
(1%)
R3SiOR′ + HSiR3″ ⎯⎯⎯⎯⎯⎯⎯→⎯ R3SiOSiR3″ + HR′
(1)
While photochemical unmasking of Lewis acidic coordina-
tion sites in transition-metal complexes is well-known, an
efficient photo Lewis acid generator (PhLAG), in which a well-
defined main group Lewis acid is released photochemically
from a thermally stable and easily handled precursor, is a less
well developed concept. A few reports in the early 1970s
described the photopolymerization of epoxides upon irradiation
of aryl diazonium salts of general formula [ArN2]+[EFn]− (E =
P, n = 6; E = B, n = 4); the reaction was proposed to be
catalyzed by EFn−1 Lewis acids formed upon photodecompo-
sition of the salt.16,17 While reasonably effective, the sensitivity
of these Lewis acids to water raises questions as to the true
nature of the catalyst and the reproducibility of the catalyzed
transformations. With this in mind, we sought a protecting
group for B(C6F5)3 whose removal could be induced with a
photon, releasing the water stable borane. Essentially, the
photochemical activation must convert the strongly Lewis basic
protecting group into one or more weaker Lewis bases whose
presence does not preclude the desired catalysis.
We have previously observed that the “frustrated Lewis
pair”18 combination of 2,2,6,6-tetramethylpiperidine (TMP)
and B(C6F5)3 takes up CO2 to form the ammonium carbamato
borate ion pair 1-TMPH (Scheme 1).19 This process is
reversible as evidence by incorporation of 13C label into 1-
TMPH upon exposure to 13CO2. Based on the X-ray structure,
which shows a hydrogen bond between the ammonium cation
and the carbamato carbonyl oxygen, we hypothesized that the
reverse process is aided by protonation of the carbamato
carbonyl oxygen. We thus set out to prepare compounds similar
to 1-TMPH but with a triaryl sulfonium countercation; to
reduce the basicity of the nitrogen center, we targeted the
carbazole-based system 2-Ph3S20 (Scheme 2). We anticipated
that irradiation of 2-Ph3S at 254 nm would cause the [Ph3S]+
cation to eject a proton (along with several other photo-
products derived from this cation)4 and trigger the release of
CO2 and B(C6F5)3 from the anion.
n homogeneous catalysis, reactions typically commence
when catalyst precursors are physically introduced to the
I
reagents. This limits the conditions under which reactions can
be initiated. The ability to induce reactions with the imposition
of an external stimulus, such as heat,1 mechanical force,2 or
light,3 allows for exquisite temporal control over the initiation
of catalytic reactions. This is particularly important for certain
materials chemistry applictions, where premature initiation of
polymerization, depolymerization, or cross-linking reactions
can be detrimental to pattern formation or device manufacture.
For example, in the photolithography arena, photoacid
generators (PAGs) are utilized to eject protons upon irradiation
at a defined wavelength, which triggers the catalyzed ester
deprotection of resist polymers in the region of exposure,
allowing for patterning to sub-20 nm resolution. Typical
commercially available and industrially relevant PAGs consist of
triphenylsulfonium salts4 that release protons upon irradiation
with 254 nm wavelength light. Despite the advanced practice of
this chemistry, the reactions that protons can amplify are
limited, and so the ability to photochemically liberate
functionally distinct and potentially more versatile Lewis acid
catalysts is highly desirable.
One important class of Lewis acids are the perfluoroarylbor-
anes,5 typified by the simplest member, tris-
(pentafluorophenyl)borane, B(C6F5)3. First prepared almost
50 years ago,6 B(C6F5)3 exhibits strong Lewis acidity, high
thermal stability, and air and water tolerance. In the 1990s this
borane emerged as an important cocatalyst for olefin
polymerizations mediated by single site catalysts.7 As a
consequence, it became widely available, and new reactions
mediated by this versatile Lewis acid were discovered, most
notably its use in the transition-metal free hydrosilylation8−10 of
a variety of substrates and hydrogenation of imines.11,12 It is
also effective for the clean and rapid construction of siloxane
Received: May 3, 2012
Published: May 31, 2012
© 2012 American Chemical Society
9601
dx.doi.org/10.1021/ja3042977 | J. Am. Chem. Soc. 2012, 134, 9601−9604

Journal of the American Chemical Society
Communication
Scheme 1
Scheme 2
Figure 1. X-ray structure of 2-Ph3S (thermal ellipsoids drawn to 50%
probability level). The tert-butyl group on C4 has been removed for
clarity. Selected bond distances (Å): B1−O1, 1.521(3); C21−O1,
1.310(3); C21−O2, 1.205(3); C21−N1, 1.393(3). Selected non-
bonded distances (Å): S1−C1, 3.54; S1−C6, 3.49; S1−C7, 3.47; S1−
C12, 3.47; S1−N1, 3.52. Selected bond angle (°): B1−O1−C21,
121.9(2).
incorporation of the isotopic label into the carbamato carbon.
Furthermore, thermogravimetric analysis (TGA) shows that it
is thermally stable up to its melting point of 185 °C at which
time it loses mass at a steady rate presumably due to
decomposition (Figure S1, Supporting Information (SI)). It
is stable also in refluxing acetonitrile or tetrahydrofuran
solution under ambient light. However, upon irradiation with
254 nm light in CD2Cl2 solution (2.1 × 10−3 M) in the
presence of 1 equiv of acetonitrile, it efficiently releases
B(C6F5)3, which is trapped as its acetonitrile adduct,
23,24
CH3CN·B(C6F5)3
(Figure 2; Figures S5−7, SI); at this
In a series of sequential steps, the TMEDA stabilized lithium
salt of the carbamic acid derived from 3,6-di-tert-butylcarbazole
(3-LiTMEDA) was synthesized and isolated. Reaction of 3-
LiTMEDA with B(C6F5)3 formed a lithium carbamato borate,
which was converted to the triphenylsulfonium salt 2-Ph3S via
ion exchange with [Ph3S]+[Cl]− in excellent yield. This salt is a
thermally robust, air and moisture stable microcrystalline white
solid that was fully characterized by multinuclear NMR, IR and
absorption spectroscopic techniques as well as X-ray crystallog-
raphy. Key spectroscopic data include characteristic resonances
in the aromatic region of the proton NMR spectrum at 7.42,
7.65, and 7.77 ppm for the phenyl group protons of the cation,
an IR band at 1700 cm−1 for the stretching frequency of the
carbonyl group in the anion, and a resonance at −4.1 ppm in
the 11B NMR spectrum for the boron nucleus of the carbamato
borate moiety. Figure 1 depicts the molecular structure of the
ion pair along with selected metrical data. The B1−O1 distance
of 1.521(3) Å in 2-Ph3S is shorter than those found in carbonyl
function adducts of B(C6F5)321 and similar to that observed for
the formatoborate complex [TMPH][HCO2B(C6F5)3],22 con-
sistent with its formulation as a carbamato borate anion. The
triphenyl sulfonium countercation sits above the five-membered
C4N core of the carbamato function, with nonbonding
distances between S1 and these atoms of ≈3.5 Å.
Figure 2. Plot of the appearance of CH3CN·B(C6F5)3 vs time as a
function of intermittent irradiation (254 nm) of 2-Ph3S (2.1 × 10−3
M
in CD2Cl2) in the presence of 1 equiv of CH3CN.
concentration the photoreaction proceeds to >99% within 12
min with an apparent rate constant of ≈3.3 × 10−3 s−1. The
generation of CH3CN·B(C6F5)3 only proceeds when the light
source is on; turning the lamp off stops the photorelease of
borane immediately. We propose that the release of borane
proceeds, as shown in Scheme 2, via protonation of the
carbonyl group; the resulting salt 2-H spontaneously ejects
borane and the carbamic acid 3-H, which undergoes
decomposition to 3,6-di-tert-butyl carbazole and CO2. This
notion is supported by the observation of a transient signal at
Ion pair 2-Ph3S is a true PhLAG on the basis of the following
observations. Exposure of the compound to an atmosphere of
13CO2 under ambient conditions for several hours results in no
9602
dx.doi.org/10.1021/ja3042977 | J. Am. Chem. Soc. 2012, 134, 9601−9604

Journal of the American Chemical Society
Communication
153.4 ppm (assigned to 3-H) and 13CO2 (126.2 ppm) in the
13C NMR spectrum of the photoproducts of irradiated
solutions of 2-Ph3S selectively 13C labeled in the carbamate
carbon (Figure S6, SI). Furthermore, strong Bronsted acids,
like triflic acid, react rapidly with 2-Ph3S to release B(C6F5)3,
CO2, and carbazole, along with [Ph3S]+[OTf]−. The carbazole
is a weak Lewis base that, shown in separate experiments, does
not bind B(C6F5)3 strongly (Figures S8−10, SI) nor interfere
with catalytic reactions mediated by B(C6F5)3.
Scheme 3
The efficacy of 2-Ph3S as a PhLAG was further demonstrated
by the photoinitiated hydrosilylation of carbonyl functions25
and silation of alcohols and ethers,8 well established reactions
catalyzed by B(C6F5)3. Solutions of substrate, 1 equiv of Et3SiH
and 1 mol % of 2-Ph3S were prepared; no reaction was
observed. Upon irradiation of these solutions at 254 nm for 15
min, silation reactions ensued and were complete in reaction
times consistent with the reactions catalyzed by B(C6F5)3
itself.8,25 The results of these experiments are summarized in
Table S2, SI.
While these results point to photogeneration of B(C6F5)3,
the ability to trigger these sorts of reactions photochemically
provides minimal advantage. Conversely, the ability to
photoinitiate the B(C6F5)3 catalyzed formation of siloxane
structures, i.e., the Piers−Rubinsztajn reaction shown in eq 1,
offers an opportunity to control the formation of thin films of
these materials. Such films have potential application as low k
dielectric materials, and the ability to create photodefined
dielectric films on wafer is of considerable interest.15
Furthermore, siloxane film formation catalyzed by B(C6F5)3
can be achieved at significantly lower temperatures than other
avenues. Model reactions using silanes and silanols, R3SiOH, or
silyl ethers, RSi(OR)3, demonstrate the efficacy of photo-
induced Si−O bond formation (and O−R bond cleavage) using
2-Ph3S as a PhLAG, producing essentially identical results as
those obtained by physical introduction of B(C6F5)3 to
solutions of these reactants. Table S3, SI, shows the results of
these model studies and suggests that 2-Ph3S can be used for
photoinduced cross-linking of suitable precursors in the
formation of siloxane-based thin films.
the absence of TEOS and characterized a soluble polymer
consisting of macrocyclic rings of various sizes due to the partial
ring opening of the cyclic monomer.26 In the present system,
while B(C6F5)3 catalyzes ring-opening reactions of TMCTS,
these reactions are minimized through use of very low B(C6F5)3
loadings, which also ensures eventual degradation of the active
borane catalyst as the solution ages. Furthermore, the use of
eight equivalents of TMCTS controls the MW of these
products and helps to maintain solubility in toluene reaction
solvent and suitability for thin film formation via spin-coating
techniques. The use of TEOS aids the curing of these films by
providing polymer networks with more extensive cross-linking,
although the 29Si NMR spectra of the TMCTS/TEOS
polymers are very similar to those reported by Chojnowski
and co-workers for the polymer prepared exclusively from
TMCTS.26
Toluene solutions of oligomerized 4 prepared as described
above and then diluted 2× with additional toluene were
charged with PhLAG 2-SPh3 (≈1 wt %) and used to form high-
quality thin films which can be subsequently rendered insoluble
in 2-heptanone via a combination of UV exposure (254 nm,
150 mJ/cm2) and baking (105 °C, 1 min, air). Control studies
have shown that both UV activation and subsequent heating are
required to render the films insoluble (Figure S17, SI); either
UV activation or thermal activation alone does not generate
insoluble films. IR analysis of the films show that those formed
via UV exposure and bake exhibit a distinct decrease in
intensity of the bands due to the Si−H stretching (2170 cm−1)
and bending (905 cm−1) modes, while the collection of bands
centered around 1070 cm−1 assignable to various Si−O−Si
modes increases in intensity relative to the film as deposited
prior to irradiation/bake (Figures S18−23, SI). Again, samples
that were only irradiated, or only baked, show minimal change
from the soluble “as deposited” films.
Although more detailed studies are required to fully elucidate
the cross-linking mechanism occurring on the wafer, these
results show that photoreleased B(C6F5)3 utilize Si−H groups
in soluble 4 to form a cross-linked film with more Si−O−Si
bonds. We believe that cross-linking results from either of two
processes: First, B(C6F5)3 can catalyze the dehydrogenative
silation of water absorbed into the film from atmosphere with
Si−H functions from two distinct oligomers.8,27 Alternatively,
in the absence of water, B(C6F5)3 could also catalyze the ring-
opening condensation reaction between two oligomers with
expulsion of MeSiH3 as observed in the solution chemistry
control experiments.26 In both cases, higher MW species will be
formed on the wafer, ultimately leading to films that are no
longer soluble in developing solvent, such as 2-heptanone.
A suitable soluble precursor to siloxane thin films containing
Si−H functions to use in photoinitiated cross-linking reactions
was synthesized from tetraethoxysilane (TEOS) and 2,4,6,8-
tetramethylcyclotetrasiloxane (TMCTS). Simply spin-coating a
silicon wafer with a mixture of TMCTS and TEOS in the
presence of 1% 2-Ph3S did not produce a stable, high-quality
thin film, so solutions of TMCTS:TEOS (8:1, 0.06 M TEOS)
in toluene were aged with 0.006 mol % loading of B(C6F5)3.
1
When such solutions were monitored by H NMR spectros-
copy, evolved ethane was observed, and all Si-OEt groups were
consumed, suggesting that the major product is the TMCTS
“tetramer” 4 (Scheme 3). However, significant quantities of
methylsilane, MeSiH3, were also produced, and the ratio of
remaining Si−Me groups to Si−H groups in the final solution
after three days (when no further spectral changes were
observed and 19F NMR spectroscopy showed complete
decompostion of B(C6F5)3) was approximately 2.7:1 rather
than the 4:1 ratio expected for 4. Control experiments (Figures
S11−16, SI) in which TMCTS was treated with low loadings of
B(C6F5)3 also show the presence of MeSiH3 and suggest that
ring opening of the TMCTS ring precedes elimination of
MeSiH3. Indeed, while this manuscript was in preparation,
Chojnowski et al. published a thorough study concerning the
“hydride transfer ring-opening polymerization” of TMCTS in
9603
dx.doi.org/10.1021/ja3042977 | J. Am. Chem. Soc. 2012, 134, 9601−9604

Journal of the American Chemical Society
Communication
(22) Ashley, A. E.; Thompson, A. L.; O’Hare, D. Angew. Chem., Int.
Ed. 2009, 48, 9839.
Likely, the baking step is required to ensure mobility of the
B(C6F5)3 within the film once it is photogenerated.
(23) Jacobsen, H.; Berke, H.; Doring, S.; Kehr, G.; Erker, G.;
̈
In summary, we have developed a molecule capable of
efficiently releasing the strong organometallic Lewis acid
B(C6F5)3 upon irradiation with UV light. Thus several of the
many reactions catalyzed by this borane5 can now be
photoinitiated. From a materials perspective, perhaps the
most versatile is the dehydrocarbon polycondensation reaction
between silanes and alkoxy silanes,28 the Piers−Rubinsztajn
reaction.13 We have employed this chemistry to develop a
method for photocuring thin films of a siloxane polymer on
silicon surfaces; optimization of the conditions for this process
and the potential of this method for fashioning patterned
structures of this material are currently under investigation.
Frohlich, R.; Meyer, O. Organometallics 1999, 18, 1724.
̈
(24) Bergquist, C.; Bridgewater, B. M.; Harlan, C. J.; Norton, J. R.;
Friesner, R. A.; Parkin, G. J. Am. Chem. Soc. 2000, 122, 10581.
(25) Parks, D. J.; Blackwell, J. M.; Piers, W. E. J. Org. Chem. 2000, 65,
3090.
(26) Chojnowski, J.; Kurjata, J.; Fortuniak, W.; Rubinsztajn, S.;
Trzebicka, B. Macromolecules 2012, 45, 2654.
(27) Zhou, D.; Kawakami, Y. Macromolecules 2005, 38, 6902.
(28) Chojnowski, J.; Rubinsztajn, S.; Cella, J. A.; Fortuniak, W.;
Cypryk, M.; Kurjata, J.; Kazmierski, K. Organometallics 2005, 24, 6077.
́
ASSOCIATED CONTENT
* Supporting Information
■
S
Full experimental details for the synthesis of 2-Ph3S, its
verification as a PhLAG, and its application toward formation of
siloxane-based thin films. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
wpiers@ucalgary.ca
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
Funding for this work was provided by Intel Corp. and the
NSERC of Canada via a CRD Grant to WEP.
■
REFERENCES
■
(1) Thomas, R. M.; Fedorov, A.; Keitz, B. K.; Grubbs, R. H.
Organometallics 2011, 30, 6713.
(2) Piermattei, A.; Karthikeyan, S.; Sijbesma, R. P. Nat. Chem. 2009,
1, 133.
(3) Keitz, B. K.; Grubbs, R. H. J. Am. Chem. Soc. 2009, 131, 2038.
(4) Dektar, J. L.; Hacker, N. P. J. Am. Chem. Soc. 1990, 112, 6004.
(5) Piers, W. E. Adv. Organomet. Chem. 2005, 52, 1.
(6) Massey, A. G.; Park, A. J.; Stone, F. G. A. Proc. Chem. Soc. 1963,
212.
(7) Chen, E. Y.-X.; Marks, T. J. Chem. Rev. 2000, 100, 1391.
(8) Blackwell, J. M.; Foster, K. L.; Beck, V. H.; Piers, W. E. J. Org.
Chem. 1999, 64, 4887.
(9) Blackwell, J. M.; Sonmor, E. R.; Scoccitti, T.; Piers, W. E. Org.
Lett. 2000, 2, 3921.
(10) Parks, D. J.; Piers, W. E. J. Am. Chem. Soc. 1996, 118, 9440.
(11) Chase, P. A.; Jurca, T.; Stephan, D. W. Chem. Commun. 2008,
1701.
(12) Sumerin, V.; Schulz, F.; Nieger, M.; Leskela, M.; Repo, T.;
̈
Rieger, B. Angew. Chem., Int. Ed. 2008, 47, 6001.
(13) Brook, M. A.; Grande, J. B.; Ganachaud, F. Adv. Polym. Sci.
2010, 235, 161.
(14) Thompson, D. B.; Brook, M. A. J. Am. Chem. Soc. 2007, 130, 32.
(15) Volksen, W.; Miller, R. D.; Dubois, G. Chem. Rev. 2009, 110, 56.
(16) Schlesinger, S. I. Photo. Sci. Eng. 1974, 18, 387.
(17) Pobiner, H. Anal. Chim. Acta 1978, 96, 153.
(18) Stephan, D. W.; Erker, G. Angew. Chem., Int. Ed. 2010, 49, 46.
(19) Berkefeld, A.; Piers, W. E.; Parvez, M. J. Am. Chem. Soc. 2010,
132, 10660.
(20) Details on the characterization data for all new compounds can
be found in the SI.
(21) Parks, D. J.; Piers, W. E.; Parvez, M.; Atencio, R.; Zaworotko, M.
J. Organometallics 1998, 17, 1369.
9604
dx.doi.org/10.1021/ja3042977 | J. Am. Chem. Soc. 2012, 134, 9601−9604