Aluminum chloride activation of chloro-boronsubphthalocyanine: A rapid and flexible method for axial functionalization with an expanded set of nucleophiles

Article abstract of DOI:10.1021/ic2016935

We have developed a process whereby chloro-boronsubphthalocyanine (Cl-BsubPc) and other BsubPcs are activated to reaction with oxygen, sulfur, and nitrogen based nucleophiles by treatment with aluminum chloride under mild conditions. This allows for the scope of atoms chemically bound to the boron atom to be expanded beyond those derived from the traditional oxygen and carbon based nucleophiles. The successful formation of thiophenoxy and phenylamino derivatives of BsubPc was confirmed spectroscopically and by X-ray crystallography of single crystals. We have proposed a detailed mechanism for this process based on experimental observation and NMR spectroscopy ( ¹H, ¹¹B, and ²⁷Al) which involves formation of a complex between a halo-BsubPc and AlCl₃ (which we denote BsubPc(Cl) ·Al(Cl′)₃). Our observations indicate that the action of phenol on BsubPc(Cl) ·Al(Cl′)₃ does not involve direct reaction at the boron atom; rather phenol first reacts at the aluminum atom along the way to the formation of a new intermediate complex BsubPc(OPh) ·Al(O′Ph)₃. The consequence is that the rate of this process is independent of the nature of the starting BsubPc. Cl-BsubPc and Br-BsubPc as well as BsubPcs with peripheral substitutents all react to form their respective phenoxy derivatives at the same rate. Quenching of BsubPc(OPh) ·Al(O′Ph)₃ with a Lewis base ultimately produces a new bond between the phenol nucleophile and the boron atom of the BsubPc.

Full text of DOI:10.1021/ic2016935

Article
pubs.acs.org/IC
Aluminum Chloride Activation of Chloro-Boronsubphthalocyanine: a
Rapid and Flexible Method for Axial Functionalization with an
Expanded Set of Nucleophiles
Graham E. Morse† and Timothy P. Bender*,†,‡,§
†Department of Chemical Engineering and Applied Chemistry, University of Toronto, 200 College St., Toronto, Ontario, Canada
M5S 3E5
‡Department of Materials Science and Engineering, University of Toronto, 184 College St., Toronto, Ontario, Canada M5S 3E4
§Department of Chemistry, University of Toronto, 80 St. George St., Toronto, Ontario, Canada M5S 3H6
S
* Supporting Information
ABSTRACT: We have developed a process whereby chloro-
boronsubphthalocyanine (Cl-BsubPc) and other BsubPcs are
activated to reaction with oxygen, sulfur, and nitrogen based
nucleophiles by treatment with aluminum chloride under mild
conditions. This allows for the scope of atoms chemically
bound to the boron atom to be expanded beyond those
derived from the traditional oxygen and carbon based
nucleophiles. The successful formation of thiophenoxy and phenylamino derivatives of BsubPc was confirmed spectroscopically
and by X-ray crystallography of single crystals. We have proposed a detailed mechanism for this process based on experimental
observation and NMR spectroscopy (1H, 11B, and 27Al) which involves formation of a complex between a halo-BsubPc and AlCl3
(which we denote BsubPc(Cl)·Al(Cl′)3). Our observations indicate that the action of phenol on BsubPc(Cl)·Al(Cl′)3 does not
involve direct reaction at the boron atom; rather phenol first reacts at the aluminum atom along the way to the formation of a
new intermediate complex BsubPc(OPh)·Al(O′Ph)3. The consequence is that the rate of this process is independent of the
nature of the starting BsubPc. Cl-BsubPc and Br-BsubPc as well as BsubPcs with peripheral substitutents all react to form their
respective phenoxy derivatives at the same rate. Quenching of BsubPc(OPh)·Al(O′Ph)3 with a Lewis base ultimately produces a
new bond between the phenol nucleophile and the boron atom of the BsubPc.
phenoxylate does not and the presence of base (to sequester
the produced HX) shows little effect on conversion and in
many cases is known to degrade the BsubPc.15 It is generally
accepted that “peripheral donor groups increase the rates and
the yields of substitution, probably by stabilization of an
eventual positive charge on the boron atom”;1 as a corollary,
electron withdrawing groups should slow reaction rates by
reducing the molecules' ability to stabilize a reactive charged
state.
Clearly changing the chemical structure of the BsubPc ligand
for the sole purpose of affecting reaction rates is undesirable, so
we sought out methods to enhance the reactivity of the axial
boron-halide bond in a general way. The aim was to allow for
milder reaction conditions in a facile and inexpensive manner
without degradation of the BsubPc ligand. One option would
be to make use of the other halides. While Cl-BsubPcs are
common, both F-BsubPcs16 and Br-BsubPcs3,8a are thought to
have different reactivities relative to Cl-BsubPc (Br-BsubPc
being faster) although no kinetic data exists in the literature.
Our experience with Br-BsubPc are that it readily hydrolyzes to
hydroxy-BsubPc (HO-BsubPc) at ambient conditions over a
INTRODUCTION
■
Boronsubphthalocyanine (BsubPc) pigments and dyes are
robust polycyclic aromatic systems currently being used in a
broad spectrum of applications.1 Recently there has been
increased interest in the study of BsubPc derivatives because of
their unique bowl-shaped structure,2 their intense orange
fluorescence,3 their intense magenta color (ε = 50,000−
90,000 L M−1 cm−1),4 their organic semiconducting proper-
ties,5 and their selective ion sensing capability.6 In the past 30
years, the synthetically accessible structural variants of BsubPc
have largely been a result of variation around the periphery of
the BsubPc ligand achieved through the use of substituted
phthalonitriles as starting materials (Figure 1). In contrast,
structural variation resulting from the displacement of the labile
axial halide in halo-BsubPcs has been limited to reaction with
oxygen7−10 and carbon11,12 based nucleophiles. In many cases
the displacement requires lengthy reaction times and/or
elevated temperatures.13 Still, such axial substitutions have
been shown to allow for the formation of BsubPc-
oligothiophene hybrids11c and for the control of the solid
state arrangement of BsubPcs.14 The mechanism for the
displacement of the axial halide is complex but is generally
thought to be SN1 in nature when oxygen nucleophiles are
used; simple phenol affects the transformation whereas
Received: August 4, 2011
Published: May 29, 2012
© 2012 American Chemical Society
6460
dx.doi.org/10.1021/ic2016935 | Inorg. Chem. 2012, 51, 6460−6467

Inorganic Chemistry
Article
Figure 1. Generalized synthetic scheme to BsubPcs functionalized in the axial position. Peripheral positions are numbered, axial positions are
indicated, and imine (Ni) and pyrrole (Np) nitrogens are indicated.
period of months when stored outside of a glovebox3,17 This
makes it impractical for use as a starting material in industrial
synthesis or exploratory research of BsubPcs by negating the
benefits of process scale up (i.e., having a large supply on hand
which can be used over a period of time or after a period of
time). Boron dipyrrins, first described in 193718 many decades
before BsubPcs in 1972,19 are structurally similar to BsubPc,
and thus their chemistry presents a rich resource to aid in the
exploration of potential new chemistry for BsubPcs. Boron
dipyrrins are formed by the complexation of dipyrrin with
boron trifluoride producing HF and leaving two residual
boron−fluorine bonds. These bonds are hydrolytically stable
and robust; as a result, chemical transformations of the B−F
bonds are relatively limited. In 2007, however, Tahtaoui et al.
found Lewis acids (aluminum chloride inter alia) to complex
with boron dipyrrins (specifically 4,4-difluoro-4-bora-3a,4a-
diaza-s-indacene (BODIPY)) activating the stable boron−
fluorine bond, causing a rapid displacement upon addition of
phenols20 while requiring stochiometric quantities of the Lewis
acid similar to Friedel−Crafts chemistry.21
in appearance (specific solubility of Cl-BsubPc in chloroben-
zene is 0.2 mg/mL4). Next, 1.5 equiv of aluminum chloride was
added which immediately resulted in a color change of the
solution to deep blue (Figure 2). The solution/mixture was
Figure 2. Absorption spectra of Cl-BsubPc (1a, red), and its complex
with AlCl3 (BsubPc(Cl)·Al(Cl′)3, blue).
Herein we outline a process which begins with the formation
of a complex when Cl-BsubPc (and other selected halo-
BsubPcs) is/are treated with aluminum chloride. Once this
complex is formed a rapid reaction with phenol (or other
selected nucleophiles) is observed, which on quenching with a
Lewis base, yields the desired BsubPc derivative. Our
experimental evidence suggests that the role of aluminum
chloride in this process is not to simply activate the halogen-
boron bond to reaction with nucleophiles (as is the case for
BODIPYs mentioned above20) nor does it facilitate the
then stirred at 60 °C to aid dissolution. After complete
evolution of the blue color (approximately 20−40 min) 5 equiv
of phenol was added to the reaction mixture while a
temperature of 60 °C was maintained. Samples were taken as
a function of time and quenched using a Lewis base (DMF or
pyridine). Analysis by HPLC showed rapid conversion to
phenoxy-BsubPc (PhO-BsubPc, compound 2) by comparison
to a genuine sample. Complete conversion was achieved after 1
h (Figure 3). Conventional workup procedures were used at
this point. In this case the samples were filtered through a plug
of basic alumina and the solvent was removed by rotary
evaporation. We did not observe a difference in the rate of
reaction when Br-BsubPc or selected other BsubPcs were used
in place of Cl-BsubPc (Figure 3a). If residual water was present
HO-BsubPc was formed. This can be avoided by conducting
the process in a glovebox or by prudently drying glassware and
solvents prior to use and operating under a positive pressure of
argon outside of a glovebox. No indication of the formation of
ring expanded products such as metal free phthalocyanine or
chloro-aluminumphthalocyanine was seen. Slower and stalled
conversions were observed when reducing the amount of the
phenol to 1.1 or 0.5 equiv (stalled at 86% and 38% conversion
respectively in under 1 h, Figure 3b). We found that
dichloromethane and chloroform are also useable as solvents
in this process although the former precludes heating the
reaction to 60 °C (in this case complete conversion at room
temperature was achievable in 6−8 h). The use of higher
amounts of Cl-BsubPc (100 mg/mL for example) retarded the
rate of reaction presumably because of a mass transfer and/or
22
formation of BsubPc+ as a prelude to reaction with a
nucleophile. Our investigation suggests that the mechanism is
more complexed. Nonetheless, this process shows high utility,
rapid reactivity which is independent of the choice of
nucleophiles or halo-BsubPc (among those selected for this
study), and uses a readily available, inexpensive, and industrially
practical reagent: aluminum chloride.
Concurrent with our study, Guilleme et al. have shown that
the chloride of Cl-BsubPc can also be extracted with a
halophilic reagent (silver triflate, AgOTf) to form boronsubp-
thalocyanine-triflate (TfO-BsubPc) in situ which can be further
functionalized with a variety of nucleophiles.23 Their approach
requires a 1−2 h reaction to form the reactive TfO-BsubPc,
displays sensitivity to peripheral substitution, and uses silver
triflate in reagent quantities, a salt derived from a precious
metal.
RESULTS AND DISCUSSION
■
The process began with the preparation of a dispersion of 20
mg/mL of Cl-BsubPc (or Br-BsubPc or other BsubPc
derivative) in chlorobenzene which appeared deep pink/red
6461
dx.doi.org/10.1021/ic2016935 | Inorg. Chem. 2012, 51, 6460−6467

Inorganic Chemistry
Article
Figure 4. Stacked plot of H, 11B, and 27Al NMR showing the steps
though a typical reaction sequence (Scheme 1): (i) after stirring Cl-
BsubPc and 1.5 equiv of aluminum chloride in CDCl3; (ii) 30 min,
(iii) 90 min, (iv) 210 min, and (v) 16 h after the addition of 5 equiv of
phenol; (vi) after quenching with excess pyridine (the section between
8.4 and 8.3 is reduced by a factor or 10 because of a large peak from
pyridine); (vii) after filtering through a plug of basic alumina and
rotary evaporation; and (viii) a reference spectra of PhO-BsubPc.
1
Figure 3. (a) Conversion of halo-BsubPc to PhO-BsubPc as a function
of time in the presence of 1.5 equiv of AlCl3 and 5 equiv of phenol (as
indicated). Conversion of Cl-BsubPc to PhO-BsubPc without AlCl3 is
given for comparison; (b) Conversion of Cl-BsubPc to PhO-BsubPc
(or Br-BsubPc as indicated) as a function of time in the presence of 1.5
equiv of AlCl3 and varying equiv of phenol.
aluminum rather than the AlCl4− anion itself. This is consistent
with the observation that the sharpness (or line-width) of an
27Al resonance is proportional to its symmetry.26 We can then
ultimately conclude that the action of Lewis acidic AlCl3 on Cl-
BsubPc results in the formation of an intermediate species
whereby the aluminum complexes to a location close to the
Lewis basic imine nitrogen (see Ni in Figure 1) of Cl-BsubPc
while simultaneously the chloride of Cl-BsubPc interacts with
the aluminum of AlCl3 to form a complex as illustrated
(Scheme 1) which we will denote as BsubPc(Cl)·Al(Cl′)3
(where Cl and Cl′ are used to differentiate chlorides originating
from Cl-BsubPc and AlCl3 respectively).
solubility limitations of both Cl-BsubPc and PhO-BsubPc in
chlorobenzene (0.2 mg/mL and 2.7 mg/mL respectively4).
Owing to the rapid formation of a deep blue colored solution
on addition of AlCl3 to Cl-BsubPc a complex must be formed
between the two constituents. To elucidate the nature of this
complex as well as the remaining mechanism of the reaction
occurring during this process we conducted a spectroscopic
1
investigation of the process. H, 11B, and 27Al NMR were
The formation of the BsubPc(Cl)·Al(Cl′)3 complex is
performed during the course of a typical reaction of Cl-BsubPc
(1a) with AlCl3 and phenol in CDCl3 (20 mg/mL) at 40 °C
(Figure 4). On addition of AlCl3 to Cl-BsubPc, 27Al NMR
showed a broad resonance at δ = 105 ppm (Figure 4i, the other
broad resonance present in the 27Al spectrum is due to
aluminum present in the NMR sample tube glass). The
position of this resonance is commonly a characteristic of an
1
corroborated by the H NMR spectrum which shows that the
1
BsubPc H resonances associated with the outer hydrogen
atoms (H2, H3, H6, H7, H10, H11; see Figure 1 for
Scheme 1. Proposed Mechanism of Cl-BsubPc Activation to
Reaction with Nucleophiles by Treatment with Aluminum
Chloride
24
−
anionic AlCl4 . Indeed, we also acquired the 27Al NMR
spectrum of the ionic liquid 1-ethyl-3-methylimidazolium
tetrachloroaluminate (CAS# 80432-05-9) which contains the
AlCl4− anion and indeed it also shows a sharp resonance at δ =
−
104 ppm (doublet, J = 24.80 Hz). However, if AlCl4 anion is
present then it must have been formed by abstraction of the
chloride from Cl-BsubPc resulting in the presence of the
corresponding positively charged BsubPc+ species. 11B NMR
does not suggest this to be the case. We observed a 11B
resonance at approximately δ = −14.3 ppm (Figure 4i) which is
shifted by approximately +1 ppm from what might be expected
from Cl-BsubPc itself25 (see Supporting Information) and
significantly different from the calculated 11B resonance of
BsubPc+ of δ = +7.4 ppm.22a Given the lack of spectroscopic
evidence for the presence of cationic BsubPc+ and the broad
27Al NMR resonance we must conclude that the resonance in
the 27Al spectrum is in fact indicative of tetra-coordinated
6462
dx.doi.org/10.1021/ic2016935 | Inorg. Chem. 2012, 51, 6460−6467

Inorganic Chemistry
Article
numbering) are slightly shifted downfield by +0.9 ppm to δ =
8.04−8.06 ppm. The resonances of the bay hydrogen atoms
(H1, H4, H5, H8, H9, H12) are significantly broadened on
addition of AlCl3 and appear at δ = 9.05 ppm which is shifted
significantly downfield from the corresponding resonances for
Cl-BsubPc itself (δ = 7.95−7.97 ppm and 8.90−8.92 ppm;
Supporting Information, Figure S1). We conclude that the
significant broadening of the resonances associated with the bay
hydrogens is further indication that the BsubPc(Cl)·Al(Cl′)3
complex rests within the bay of the BsubPc through an
interaction with the Lewis basic imine nitrogens (Scheme 1).
Our attempts to grow crystals of the BsubPc(Cl)·Al(Cl′)3
complex were unsuccessful, although we were able to obtain
a blue/green glassy precipitate during a few attempts.
On addition of phenol several observations can be made. The
first is that on addition of 5 equiv of phenol 11B NMR indicates
that there is formation of two distinct complexes during the
first 30 min which completely convert to one distinct species
between 210 min to 16 h (Figure 4ii−Figure 4v). In each
complex the 11B nucleus is progressively shielded from the
initial BsubPc(Cl)·Al(Cl′)3 complex indicating reaction with
phenol. Second, 27Al NMR does not show a difference in the
position of the 27Al resonance on additional of phenol over the
same time scale indicating the persistence of tetra-coordinated
aluminum (δ = 105 ppm). The resonance is however
significantly sharpened.
Information, Figure S1). Upon addition of pyridine, 27Al NMR
shows the disappearance of the resonance at δ = 105 ppm
characteristic of the tetra-coordinated aluminum species
(Figure 4vi). After a workup by filtration through a plug of
1
basic alumina and then rotary evaporation, a final H and 11B
NMR proves an identical match to a genuine sample of PhO-
BsubPc (Figures 4vii and 4viii, respectively).
We should also note that 5.0 equiv of phenol is not required
to facilitate the complete conversion of PhO-BsubPc. We have
found that only 2.5 equiv of phenol is required (Figure 3b).
Below 2.5 equiv of phenol a plateau in conversion occurs. For
example, upon addition of 1.1 equiv of phenol conversion from
Cl-BsubPc (or Br-BsubPc) to PhO-BsubPc stalls at 86%
(Figure 3b). Following our supposition above, addition of 1.1
equiv of phenol would lead to the formation of a complex
BsubPc(Cl)·Al(Cl′)1.9(OPh′)1.1. Given that quenching of this
complex does produce PhO-BsubPc, the act of quenching must
give rise to the migration of the phenoxy fragment from the
aluminum to the boron atom. If this migration is merely
statistical, one might expect the formation of ∼25% of PhO-
BsubPc on quenching. Given that it yields 86% PhO-BsubPc
indicates a preference to form the stronger B−O bond of PhO-
BsubPc over reforming the B−Cl bond of Cl-BsubPc, although
this preference is clearly not absolute in this case.
To further support our observations and our proposed
mechanism two additional experiments were performed. Each
experiment is designed to show the formation of an
intermediate complex which on quenching results in migration
of a molecular fragment from the aluminum atom to the boron
atom concurrent with the formation of a stronger bond to
boron than existed in the starting material. In the first, we
intercepted our process after the reaction of Br-BsubPc with 1.5
equiv of AlCl3 (60 °C, chlorobenzene, 3 h). The result was the
formation of a deep blue solution. We quenched the resulting
BsubPc(Br)·Al(Cl)3 complex with a Lewis base (pyridine)
without addition of phenol. The workup was as described
above. The result was complete conversion of Br-BsubPc to Cl-
BsubPc which was confirmed by 1H NMR (Supporting
Information, Figure S2) and HR MS (EI, calcd for 430.0905,
found 430.0904, Supporting Information, Figure S3); MS
confirming the absence of Br-BsubPc again indicating the
formation of a stronger B−Cl bond over reforming the weaker
B−Br bond in this case absolutely and without forming a
statistical mixture.
In the second experiment, while not exactly the equivalent of
initiating the reaction with the phenoxylated aluminum
complex BsubPc(OPh)·Al(O′Ph)3 (Scheme 1), we treated Br-
BsubPc with 1.5 equiv of Al(OPh)3 at 60 °C in chlorobenzene
(3 h). The result was a color change of the solution which was
not as dramatic a change as is seen with AlCl3 (likely because of
the lower Lewis acidity of Al(OPh)3). Subsequent treatment of
the intermediate complex BsubPc(Br)·Al(OPh)3 with a Lewis
base (pyridine) resulted in the color reverting back to the
standard magenta/purple color characteristic of BsubPcs and
the complete formation of PhO-BsubPc (workup was as
described above, absence of Br-BsubPc and conversion to PhO-
BsubPc confirmed by 1H NMR (Supporting Information,
Figure S2) and HR MS (EI, calcd for 488.1557, found
488.1559, Supporting Information, Figure S4)). Again a
statistical mixture is not formed and formation of the stronger
B−O bond is favored over reforming the weaker B−Br bond.
Each of these two experiments further corroborate our
proposed mechanism shown in Scheme 1 and the idea that
We can then suppose that on addition of phenol to the
process the BsubPc(Cl)·Al(Cl′)3 complex reacts quickly with
the initial equivalents of phenol discriminately and rapidly
(under 30 min) at the Cl′ position first (the chloride bound to
Al) forming BsubPc(Cl)·Al(O′Ph)3 followed by reaction at the
Cl position relatively slowly (started at 30 min and completed
at approximately 120 min) to form BsubPc(OPh)·Al(O′Ph)3
(Scheme 1).
The supposition that phenol reacts discriminately at the
aluminum center is further supported by the observation that
upon addition of 5 equiv of phenol, we observed a lack of
difference in the observed rates of formation of the phenoxy
derivative of each of Cl-BsubPc, Br-BsubPc, Cl−Cl6BsubPc,
and Cl−Cl12BsubPc (Figure 3a). This is in contrast to the
observation that under standard conditions (without AlCl3)
reaction of this list of derivatives with phenol shows a
differential reactivity relative to each other presumably because
of a difference in the polarization of the B−Cl bond.17 This is
also in contrast to the observations of Guilleme et al. who have
observed when TfO-BsubPc was generated in situ and reacted
there was also a difference in reactivity with different peripheral
substituents.23 Each of these observations which are contrary to
previous observations are further indications that the action of
AlCl3 on Cl-BsubPc produces an intermediate complex
BsubPc(Cl)·Al(Cl′)3 rather than simply enhancing the
formation of BsubPc+ in the SN1-type mechanism which is
typical for phenoxylation of halo-BsubPcs1 and supposed for
TfO-BsubPc.23
Finally, after complete formation of the BsubPc(OPh)·Al-
(OPh′)3 complex (Scheme 1), the addition of a common Lewis
base (we have used both pyridine and DMF in this case) effects
its rearrangement to quantitatively give PhO-BsubPc (Figure
4vi). The 1H and 11B NMR resonances of the quenched sample
are significantly shifted upfield; however, we have confirmed
this is due only to the large portion of pyridine present in the
1
sample (a H NMR spectrum of PhO-BsubPc in the presence
of large amounts of pyridine is illustrated in Supporting
6463
dx.doi.org/10.1021/ic2016935 | Inorg. Chem. 2012, 51, 6460−6467

Inorganic Chemistry
Article
the phenoxy molecular fragment of PhO-BsubPc does indeed
come about by migration from the aluminum atom to the
boron atom on quenching with a Lewis base when less than 5
equiv of phenol are used.
Because we are interested in accessing BsubPc derivatives
with new molecular fragments in the axial positions through
reaction with new nucleophiles, we attempted to use our
process conditions with p-thiocresol, aniline, and N-methylani-
line in place of phenol. We had previously attempted to directly
react these nucleophiles with Cl-BsubPc under conventional
reaction conditions5c,17 without any success; in the case of p-
thiocresol, complete discoloration was observed thereby
indicating the decomposition of the BsubPc unit and in the
case of aniline and N-methylaniline, no reaction was observed.
Each observation is in line with that recently reported by
Guilleme et al.23 Yet, using the process conditions described
herein, conversion of Cl-BsubPc was achieved forming 4-
methylthiophenoxy-boronsubphthalocyanine (3, Scheme 2), N-
Figure 5. Thermal ellipsoid plot of compound 3 at a 35% probability
level. Hydrogen atoms have been omitted for clarity.
Scheme 2. Reaction of halo-BsubPc (1a−d) Yielding Their
a
Respective Axially Substituted BsubPc (Compounds 2−5)
exact analogue to compound 3 containing oxygen (6, Scheme
2)14a thus allowing the comparison of the molecular geometry
changes on introduction of the heavier sulfur atom to the
structure (Table 1). As would be expected, the B−S bond is
longer at 1.923(3) Å as compare to a B−O bond length of
1.436(2) Å. Similarly the S−C bond is longer at 1.776(2) Å
compared to 1.386(2) Å. The B−S−C bond angle was
significantly smaller at 99.2(1)° compared to 115.6(1)° for
B−O−C. What was unexpected was a difference in the bowl
depth of the BsubPc ligand. Bowl depth is defined by the
distance between a line intersecting the boron atom which is
perpendicular to the plane defined by C2|C3|C6|C7|C10|C11
(see numbering in Figure 1). For compound 3 it was
significantly shorter (2.515 Å compared to 2.707 Å) than for
compound 6 indicating that the presence of sulfur has the effect
of flattening the bowl of the BsubPc ligand in the solid state.
We also noted a 5 nm red-shift (to 566 nm) in the absorption
spectrum of 3 relative to 6 in toluene possibly indicating the
flattening of the bowl of BsubPc also happens in solution.
The X-ray determined structure of compound 5 has two
independent molecules in the unit cell showing an axial B−N
bonding distance of 1.508(3) Å and 1.532(3) Å, similar to
those recently described by Guilleme et al.23 The independent
molecules are joined by a single molecule of water despite the
dry solvents being used. The water is hydrogen bound to one of
the nitrogen atoms of the BsubPc ligand of each molecule. In
this case the bowl of the BsubPc ligand in each molecule was
also flattened compared to that of 6, and similarly its absorption
maximum is shifted by 3 nm to 564 nm in toluene. However we
cannot clearly attribute the flattening of the bowl of BsubPc to
a contribution of the nitrogen (in place of oxygen), the N-
methyl group, or the unique crystal structure. Exhaustive
attempts to crystallize 4 were all unsuccessful.
a
Conditions: (i) 1.5 equiv of AlCl3, 5 equiv of phenol, thiophenol,
a
aniline, or N-methylaniline, chlorobenzene, 60 °C. − Compound 6:
4-methyphenoxy-BsubPc is included for reference and was prepared as
detailed in ref 14a.
phenyl-amino-boronsubphthalocyanine (4, Scheme 2) and N-
phenyl-N-methyl-amino-boronsubphthalocyanine (5, Scheme
2) respectively. Each reaction took between 2 h and 8 h to
complete after addition of the nucleophile depending on the
nucleophile. Compound 3 was then purified by sequential
Kauffman column chromatography on silica and then standard
basic alumina with dichloromethane as eluent with an isolated
yield of 44%. Compounds 4 and 5 were isolated by flash
chromatography using basic alumina and dichloromethane,
followed by precipitation from a tetrahydrofuran (THF)
solution into pentane. Final purification was achieved by
sublimation to 210 °C with isolated yields of 30% and 77%
respectively. Crystals suitable for X-ray diffraction of 3 and 5
were obtained by slow vapor diffusion of heptane into
chloroform, unambiguously identifying their respective struc-
tures (Figure 5 and Figure 6). Selected crystallographic
information is given in Table 1 and complete crystallographic
data is given in the Supporting Information accompanying this
Article.
CONCLUSION
■
In summary, we have outlined a new and rapid process to
obtain a variety of BsubPc derivatives with unique molecular
fragments in the axial position. Using a readily available and
inexpensive reagent, aluminum chloride (AlCl3), we have been
able to show a rapid and low temperature (less than 60 °C)
method to produce PhO-BsubPcs in at least three solvents
(chlorobenzene, chloroform, and dichloromethane). This
method can also be used to access a wider range of molecular
The crystal structure of compound 3 presents the first
example of a crystal structure where a B−S bond is attached to
a BsubPc. We have previously synthesized and crystallized the
6464
dx.doi.org/10.1021/ic2016935 | Inorg. Chem. 2012, 51, 6460−6467

Inorganic Chemistry
Article
Figure 6. Thermal ellipsoid plot of compound 5 at a 35% probability level. Hydrogen atoms have been omitted for clarity.
used as received. All 1H NMR spectra where referenced to an internal
standard of 0.05% TMS. All crystal structures were collected using
computer-controlled KappaCCD system and an Oxford Cryostream
variable temperature apparatus. All ultraviolet−visible (UV−vis)
spectroscopy was performed using PerkinElmer Lambda 1050 with a
10 mm path length for solution phase samples. High pressure liquid
chromatography (HPLC) analysis was conducted using a Waters 2695
separation module with a Waters 2998 photodiode array and a Waters
4.6 mm ×100 mm SunFire C18 3.5 μm column. HPLC grade
acetonitrile and DMF were eluted at 0.6 mL/min during operation at a
composition of 80:20, respectively. Mass spectrometry was performed
on a Waters GC Time-of-Flight mass spectrometer with an electron
ionization probe and accurate mass determination.
Phenoxylation of Boronsubphthalocyanines (Compounds
1a−d). The conversion of compounds 1a−d to their phenoxylated
derivative was observed using the following conditions. BsubPc (1a:
0.100 g, 1b: 0.221 g, 1c: 0.148 g, 1d: 0.196 g) was combined with
aluminum chloride (1.5 equiv) and chlorobenzene (21.5 L
chlorobenzene/mol BsubPc (∼20 g BsubPc/L chlorobenzene);
molar concentration kept constant) in a 20 mL scintillation vial
inside a glovebox. The contents were stirred at 60 °C temperature for
between 30 min until complete evolution of the blue color was
achieved. Then 5 (2.5, 1.1 or 0.5) equiv of phenol (1a, 1c, 1d: 0.109 g;
1b: 0.048 g) were added. The contents were stirred for another 5 h at
which point 1 mL of pyridine was added to the reaction mixture and
allowed to stir for 5 min. Each reaction was monitored by HPLC. The
products of each reaction were not isolated.
fragments based on nitrogen and sulfur for BsubPcs (from their
respective nucleophiles).
We have proposed a mechanism by which this trans-
formation occurs which is supported by observation and NMR
spectroscopy (1H, 11B, and 27Al). The mechanism involves
formation of an intermediate complex BsubPc(Cl)·Al(Cl′)3 by
which aluminum chloride complexes with Cl-BsubPc. The
subsequent addition of a phenol results in the formation of
another immediate complex BsubPc(OPh)·Al(OPh′)3 which
upon quenching with a Lewis base results in the formation of a
new bond between the selected nucleophile and the boron
atom of the BsubPc. We also present experimental evidence
which suggests that the reaction of phenol occurs at the
aluminum atom first. This observation is further supported by
the reaction of AlCl3 and Al(OPh)3 with Br-BsubPc producing
Cl-BsubPc and PhO-BsubPc, respectively. The consequence is
that this process does not show the sensitivity to peripheral
BsubPc substitution that other processes do as the polarization
or chemical environment of the halo-B bond is not of
significance. Other nucleophiles such as p-thiocresol, aniline,
and N-methylaniline yield their respective BsubPc derivatives,
despite being entirely unreactive under conventional/standard
axial substitution conditions.
The aim of our group is to use this new process to rapidly
access an ever expanding set of molecular fragments for the
axial position of BsubPc. We have shown that changes in the
axial fragment can affect the physical (including electronic)
properties3,4,27 and solid state arrangements of BsubPcs.14 Such
effort is both warranted and timely as BsubPcs are materials of
current interest for application in the fields listed above.1−6
Chloro-boronsubphthalocyanine (Compound 1a). Synthe-
sized as previously reported.4,10c
Bromo-boronsubphthalocyanine (Compound 1b). Synthe-
sized as previously reported.9a
Chloro-hexachloroboronsubphthalocyanine (Compound
1c). Synthesized as previously reported.27
Chloro-dodecachloroboronsubphthalocyanine (Compound
1d). Cl-Cl12BsubPc was synthesized by adapting the method of
Zyskowski and colleagues.10c 220 mL of 1,2-dichlorobenzene was
added to 11.047 g of tetrachlorophthalonitrile in a 500 mL three-neck
round-bottom flask. 100 mL of 1.0 M solution of BCl3 in heptane was
added. The mixture was then heated, and the heptane was removed
through distillation using a short-path condenser. After distillation the
flask was heated at reflux (180.5 °C) for 1.5 h. A positive pressure of
argon was maintained throughout the experiment. The products were
then cooled, and the solvent was removed by rotary evaporation. Next,
the dry solid was placed in a cellulose thimble and extracted with
methanol and then petroleum ether, each for 20 h in a Soxhlet
EXPERIMENTAL SECTION
■
Methods and Materials. Pyridine, N,N-dimethylformamide
(DMF), acetonitrile, and phenol were purchased from Caledon
Laboratories Ltd. and used as received. Chlorobenzene was obtained
from Caledon Laboratories Ltd. and purified and stored in a Pure Solv
solvent purification system prior to use. Aluminum chloride, p-
thiocresol, aniline, and N-methylaniline were obtained from Sigma
Aldrich Canada. All nuclear magnetic resonance (NMR) spectra were
acquired on a Bruker 400 MHz system in deuterated chloroform
(CDCl3) purchased from Cambridge Isotope Laboratories which was
6465
dx.doi.org/10.1021/ic2016935 | Inorg. Chem. 2012, 51, 6460−6467

Inorganic Chemistry
Article
Crystals of suitable quantity for X-ray diffraction were grown by slow
vapor diffusion of heptane into chloroform.
Table 1. Relevant Crystallographic and Selected Geometrical
Parameters for Compounds 3, 5, and 6
N-Phenylamino-boronsubphthalocyanine (Compound 4).
Cl-BsubPc (0.300 g, 6.97 × 10−4 mol) was combined with aniline
(0.32 mL, 3.48 × 10−3 mol, 5 equiv), aluminum chloride (0.139 g, 1.05
× 10−3 mol, 1.5 equiv), and chlorobenzene (15 mL) in a 20 mL
scintillation vial inside a glovebox. The contents were stirred at 60 °C
temperature for 8 h after which 3 mL of pyridine was added to the
reaction mixture and allowed to stir for 5 min. The reaction mixture
was subsequently purified by flash chromatography through a plug of
standard basic alumina with dichloromethane and the solvent.
Removal of the dichloromethane by rotary evaporation yielded an
oily red/pink fluid. Next, the oil was redissolved in 1 mL of
tetrahydrofuran, precipitated with pentane, and filtered yielding a red/
purple powder. Residual volatile impurities were next removed by
sublimation at 210 °C (100 mg, yield 30%, 99+% pure by HPLC
maxplot). 1H NMR (400 MHz, CDCl3, Me4Si): δ 5.02 −5.06 (2H, d),
6.27−6.32 (1H, t), 6.69−6.73 (2H, t), 7.88−7.91 (6H, m), 8.83−8.86
(6H, m), 13C NMR (400 MHz, CDCl3, Me4Si): δ 113.77, 116.40,
122.02, 128.85, 129.72, 130.88, 144.24, 151.69. HRMS (positive) exact
mass calcd for C30H18BN7, 487.1717; found 487.1728. UV−vistoluene
λmax: 564 nm.
N-Phenyl-N-methyl-amino-boronsubphthalocyanine (Com-
pound 5). Cl-BsubPc (0.300 g, 6.97 × 10−4 mol) was combined
with aniline (0.32 mL, 3.48 × 10−3 mol, 5 equiv), aluminum chloride
(0.139 g, 1.05 × 10−3 mol, 1.5 equiv), and chlorobenzene (15 mL) in a
20 mL scintillation vial inside a glovebox. The contents were stirred at
60 °C temperature for 8 h after which 3 mL of pyridine was added to
the reaction mixture and allowed to stir for 5 min. The reaction
mixture was subsequently filtered through a plug of silica with
dichloromethane, and the solvent was removed by rotary evaporation
yielding an oily red/pink fluid. Next, the oil was redissolved in 1 mL of
tetrahydrofuran, precipitated with pentane and filtered yielding a red/
purple powder. Residual volatile impurities were next removed by
sublimation at 210 °C yielding gold crystals (268 mg, yield 77%, 99+%
pure by HPLC Maxplot). 1H NMR (400 MHz, CDCl3, Me4Si): δ 1.06
(3H, s), 5.44−5.47 (2H, d), 6.50−6.55 (1H, t), 6.71−6.75 (2H, t),
7.87−7.90 (6H, m), 8.81−8.84 (6H, m), 13C NMR (400 MHz, CDCl3,
Me4Si): δ 32.07, 119.31, 119.71, 122.07, 128.02, 129.66, 130.81,
147.78, 151.55, HRMS (positive) exact mass calcd for C31H20BN7,
518.1873; found 518.1991. UV−vistoluene λmax: 564 nm. Crystals of
suitable quantity for X-ray diffraction were grown by slow vapor
diffusion of heptane into chloroform.
a
3
5
6
chemical formula C31H19BN6S
C31H20BN7·0.5(H2O)
502.25
C31H19BN6O
502.33
formula mass
crystal system
a/Å
518.39
monoclinic
10.3506(4)
22.6651(9)
10.7047(3)
90.00
triclinic
triclinic
11.5120(3)
11.7673(3)
19.3022(5)
72.9030(14)
81.8810(12)
75.8510(13)
2416.61(11)
10.1555(3)
10.9258(4)
11.7215(4)
86.3530(17)
78.496(2)
66.9870(18)
1172.88(7)
b/Å
c/Å
α/deg
β/deg
102.582(2)
90.00
γ/deg
unit cell volume/ 2450.99(15)
Å3
temperature/K
space group
150(1)
P21/c
4
150(1)
150(1)
P1
P1
̅
̅
no. of formula
units per unit
cell, Z
4
2
no. of reflections 14105
42414
11050
14448
5309
measured
no. of
independent
reflections
5356
Rint
0.0496
0.0548
0.0641
0.0578
0.0441
0.0465
final R1 values
(I > 2σ(I))
final wR(F2)
values
0.1312
0.1392
0.1138
(I > 2σ(I))
final R1 values
0.1004
0.1583
0.1131
0.1638
0.0651
0.1264
(all data)
final wR(F2)
values
(all data)
goodness of fit
on F2
1.019
1.038
1.058
B-R4 bond
length (Å)
1.923(3)
1.776(2)
99.2(1)
2.515
1.508(3)
1.426(3)
119.8(2)
2.556
1.532(3)
1.407(3)
123.5(2)
2.631
1.436(2)
1.386(2)
115.6(1)
2.707
R4-C bond
length (Å)
B-R4-C bond
angle (deg)
ASSOCIATED CONTENT
■
bowl depth (Å)
a
S
* Supporting Information
Compound 6: 4-methyphenoxy-BsubPc from ref 14a.
Supplementary figures; 1H and 13C NMR spectra for
compounds 3, 4 and 5; and detailed crystallographic data for
compounds 3 and 5. This material is available free of charge via
the Internet at http://pubs.acs.org.
extraction apparatus. The solid was then placed in a vacuum oven at 50
°C overnight. Cl-Cl12BsubPc was used without further purification in
the subsequent steps. Yield 6.344 g (purity 92%, HPLC maxplot).
HRMS exact mass calcd for C24BCl13N6, 843.6141; found 843.6127.
4-Methylthiophenoxy-boronsubphthalocyanine (Com-
pound 3). Cl-BsubPc (0.300 g, 6.97 × 10−4 mol) was combined
with p-thiocresol (0.288 g, 3.48 × 10−3 mol, 5 equiv), aluminum
chloride (0.139 g, 1.05 × 10−3 mol, 1.5 equiv), and chlorobenzene (15
mL) in a 20 mL scintillation vial inside a glovebox. The contents were
stirred at 60 °C temperature for 2 h after which 3 mL of pyridine was
added to the reaction mixture and allowed to stir for 5 min. The
reaction mixture was subsequently poured onto a bed of standard basic
alumina after which it was isolated by sequential Kauffman column
chromatography, first on standard basic alumina and then on standard
silica with dichloromethane as eluent in each case. The solvent was
removed by rotary evaporation yielding 0.160 g of a dark purple solid
(44%, 99+% pure HPLC maxplot and 1H NMR). 1H NMR (400
MHz, CDCl3, Me4Si): δ 2.24 (3H, s), 6.05−6.08 (2H, d), 6.67−6.70
(2H, d), 7.92−7.89 (6H, m), 8.80−8.83 (6H, m). 13C NMR (400
MHz, CDCl3, Me4Si): δ 21.28, 122.38, 128.76, 129.35, 139.99, 130.34,
131.12, 134.37, 150.45 HRMS (positive) exact mass calcd for
C31H19BN6S, 518.1475; found 518.1485. UV−vistoluene λmax: 566 nm.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: tim.bender@utoronto.ca.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We acknowledge the financial support from the National
Sciences and Engineering Research Council (NSERC) for
providing support in the form of a Discovery Grant (T.P.B.)
and a Canada Graduate Scholarship (G.E.M.).
REFERENCES
(1) Claessens, C. G.; Gonzal
2002, 102 (3), 835−854.
■
́
ez-Rodríguez, D.; Torres, T. Chem. Rev.
6466
dx.doi.org/10.1021/ic2016935 | Inorg. Chem. 2012, 51, 6460−6467

Inorganic Chemistry
Article
(2) (a) Claessens, C. G.; Gonzal
́
ez-Rodríguez, D.; Iglesias, R. S.;
ez-Rodríguez,
BODIPY see: (c) Bonnier, C.; Piers, W. E.; Parvez, M.; Sorensen, T. S.
Chem. Commun. 2008, 4593−4595. and (d) Hudnall, T. W.; Gabbai,
F. P. Chem. Commun. 2008, 4596−4597.
Torres, T. C. R. Chim. 2006, 9, 1094−1099. (b) Gonzal
D.; Carbonell, E.; Guldi, D. M.; Torres, T. Angew. Chem., Int. Ed. 2009,
48, 1−6.
(3) Morse, G. E.; Helander, M. G.; Maka, J. F.; Lu, Z. H.; Bender, T.
P. ACS Appl. Mater. Interfaces 2010, 2 (7), 1934−1944.
(4) Morse, G. E.; Paton, A. S.; Lough, A. J.; Bender, T. P. Dalton
Trans. 2010, 39, 3915−3922.
(5) (a) Mutolo, K. L.; Mayo, E. I.; Rand, B. P.; Forrest, S. R.;
Thompson, M. E. J. Am. Chem. Soc. 2006, 128, 8108−8109.
(b) Yasuda, T.; Tsutsui, T. Mol. Cryst. Liq. Cryst. 2007, 462, 3−9.
(c) Morse, G. E.; Castrucci, J. S.; Helander, M. G.; Lu, Z. H.; Bender,
T. P. ACS Appl. Mater. Interfaces 2011, 3 (9), 3538−3544.
(6) (a) Xu, S.; Chen, K.; Tian, H. J. Mat. Chem. 2005, 15 (27−28),
2676−2680. (b) Yang, Li, Y.; Xu, S.; Li, X.; Chen, K.; Tian, H. Chem.
Lett. 2007, 36 (5), 664−665. (c) Palomares, E.; Martínez-Díaz, M. V.;
Torres, T.; Coronado, E. Adv. Funct. Mater. 2006, 16, 1166−1170.
(d) Ros-Lis, J. V.; Martinez-Manez, R.; Soto, J. Chem. Commun. 2005,
52605260.
́
(23) Guilleme, J.; Gonzalez-Rodriguez, D.; Torres, T. Angew. Chem.,
Int. Ed. 2011, 50 (15), 3506−3509.
(24) Del Grosso, A.; Singleton, P. J.; Muryn, C. A.; Ingleson, M. J.
Angew. Chem., Int. Ed. 2011, 50, 1−6.
(25) Paton, A. S.; Morse, G. E.; Castelino, D.; Bender, T. P. J. Org.
Chem. 2012, 77, 2531−2536.
(26) Karlik, S. J.; Tarien, E.; Elgavish, G. A.; Eichhorn, G. L. Inorg.
Chem. 1983, 22, 525−529.
(27) Morse, G. E.; Helander, M. H.; Stanwick, J.; Sauks, J. M.; Paton,
A. S.; Lu, Z. H.; Bender, T. P. J. Phys. Chem., C 2011, 115, 11709−
11718.
́
(7) Phenols: Claessens, C. G.; Gonzalez-Rodríguez, D.; del Rey, B.;
Torres, T.; Mark, G.; Schuchmann, H. P.; von Sonntag, C.;
MacDonald, J. G.; Nohr, R. S. G. Eur. J. Org. Chem. 2003, 14,
2547−2551.
(8) Alcohols: (a) Kasuga, K.; Idehara, T.; Handa, M.; Ueda, Y.;
Fujiwara, T.; Isa, K. Bull. Chem. Soc. Jpn. 1996, 69, 2559−2563.
(b) Engel, M. K.; Yao, J.; Maki, H.; Takeuchi, H.; Yonehara, H.; Pac,
C. Kawamura Rikagaku Kenkyusho Hokoku 1997, 53−65 ,
CA130:39766..
(9) Water: (a) Potz, R.; Goldner, M.; Huckstadt, H.; Cornelissen,
U.; Tutass, A.; Homborg, H. Z. Anorg. Allg. Chem. 2000, 626, 588−
596. (b) Eckert, A. K.; Rodriguez-Morgade, M. S.; Torres, T. Chem.
Commun. 2007, 4104−4106.
(10) Silanols: (a) Del Rey, B.; Martinez-Diaz, M. V.; Barbera, J.;
Torres, T. J. Porphyrins Phthalocyanines 2000, 4, 569−573. (b) Geyer,
M.; Plenzig, F.; Rauschnabel, J.; Hanack, M.; del Rey, B.; Sastre, A.;
Torres, T. Synthesis 1996, 1139−1151. (c) Zyskowski, C. D.; Kennedy,
V. O. J. Porphyrins Phthalocyanines 2000, 4, 707−712.
(11) Grignards: (a) Camerel, F.; Ulrich, G.; Retailleau, P.; Ziessel, R.
Angew. Chem., Int. Ed. 2008, 47, 8876−8880. (b) Ziessel, R.; Ulrich,
G.; Elliott, K. J.; Harriman, A. Chem.Eur. J. 2009, 15, 4980−4984.
(c) Mauldin, C. E.; Piliego, C.; Poulsen, D.; Unruh, D. A.; Woo, C.;
Ma, B.; Mynar, J. L.; Frec
(10), 2833−2838.
́
het, M. J. ACS Appl. Mater. Interfaces 2010, 2
(12) Organolithium: Geyer, M.; Plenzig, F.; Rauschnabel, J.; Hanack,
M.; del Rey, B.; Sastre, A.; Torres, T. Synthesis 1996, 1139−1151.
(13) Fukuda, T.; Olmstead, M. M.; Durfee, W. S.; Kobayashi, N.
Chem. Commun. 2003, 1256−1257.
(14) (a) Paton, A. S.; Morse, G. E.; Lough, A. J.; Bender, T. P.
CrystEngComm 2011, 13, 914−919. (b) Paton, A. S.; Lough, A. J.;
Bender, T. P. CrystEngComm 2011, 13, 3653−3656.
(15) (a) Gonzal
́
ez-Rodríguez, D.; Torres, T. Eur. J. Org. Chem. 2009,
1871−1879. (b) Weitemeyer, A.; Kliesch, H.; Wohrle, D. J. Org. Chem.
1995, 60, 4900−4904.
̈
(16) Rodriguez-Morgade, M. S.; Claessens, C. G.; Medina, A.;
Gonzalez-Rodriguez, D.; Gutierrez-Puebla, E.; Monge, A.; Alkorta, I.;
Elguero, J.; Torres, T. Chem.Eur. J. 2008, 14, 1342−1350.
(17) Morse, G. E., Bender, T. P., unpublished observations.
(18) Fischer, H.; Orth, H. Die Chemie des Pyrrols; Akademische
Verlagsgesellschaft: Leipzig, Germany, 1937.
(19) Meller, A.; Ossko, A. Monatsh. Chem. 1972, 103, 150−155.
(20) Tahtaoui, C.; Thomas, C.; Rohmer, F.; Klotz, P.; Duportail, G.;
Mely, Y.; Bonnet, D.; Hibert, M. J. Org. Chem. 2007, 72, 269−272.
(21) (a) Friedel, C.; Crafts, J. M. Compt. Rend. 1877, 84, 1392−1450.
(b) Calloway, N. O. Chem. Rev. 1935, 17 (3), 327−392.
(22) Boron centered cations are known for BsubPc and related
compounds. For BsubPc+ see: (a) Kato, T.; Tham, F. S.; Boyd, P. D.
W.; Reed, C. A. Heteroat. Chem. 2006, 17 (3), 209−216. For boron
subphophyrin see: (b) Tsurumaki, E.; Hayashi, S.; Tham, F. S.; Reed,
C. A.; Osuka, A. J. Am. Chem. Soc. 2011, 133, 11956−11959. For
6467
dx.doi.org/10.1021/ic2016935 | Inorg. Chem. 2012, 51, 6460−6467