Full text of DOI:10.1021/op200088b

ARTICLE
pubs.acs.org/OPRD
Scalable Syntheses of the Vaulted Biaryl Ligands VAPOL and VANOL
via the Cycloaddition/Electrocyclization Cascade
Zhensheng Ding, Wynter E. G. Osminski, Hong Ren, and William D. Wulff*
Department of Chemistry, Michigan State University, East Lansing, Michigan 48824, United States
ABSTRACT: Synthetic approaches to the VANOL and VAPOL ligands are developed which are straightforward, inexpensive,
efficient, and amenable to large-scale preparation of the ligands since minimum chromatographic purification is required. The key
step in each synthesis is a cycloaddition/electrocyclic ring-opening/electrocyclic ring closure/tautomerization cascade that provides
a direct one-step route to the monomers from which each ligand is prepared. Improved phenol coupling protocols are developed
which provide the racemic ligands. Finally, dramatic improvements in the resolution procedures feature the reduction of the number
of chemical steps and the defining of new crystallization protocols that greatly enhance the ease and reliability of the separation of
diastereomeric salts.
catalyst 5 and the VANOL catalyst 4 are essentially equally effec-
tive for cis-aziridinations.14g,j Recent reports have described
VAPOL catalysts for the asymmetric catalytic chlorination and
Michael reactions of oxindoles16 and for the benzoyloxylation of
oxindoles.17 Finally, VANOL catalysts have recently been shown
to be superior to both VAPOL and BINOL catalysts for the
catalytic asymmetric amino-allylation of aldehydes.18
1. INTRODUCTION
BINOL and its derivatives are among the most important class
of ligands for asymmetric synthesis.1,2 We have introduced the
vaulted biaryl ligands VANOL 2 and VAPOL 3 with an eye on
redesigning the location of the major groove of the ligand relative
to the active site containing the phenol functions.3,4a As illus-
trated for VAPOL 3 in Scheme 1, as a result of the relocation of
the major groove, there is a substantially larger chiral pocket
around the active site in VAPOL 3 than there is in BINOL 1. In
validation of this design, the VANOL and VAPOL ligands have
been shown to be effective in chiral catalysts for a number of
reactions. In some, VANOL led to the superior catalyst, and for
others it was VAPOL. Aluminum derivatives of VAPOL were
more effective than VANOL for DielsÀAlder reactions,4 and
zirconium derivatives of VAPOL were more effective than those
of VANOL for Mannich reactions.5 It was interesting to find that
VANOL was far more effective than VAPOL in aluminum-medi-
ated BaeyerÀVilliger reactions.6 VANOL and VAPOL have also
been incorporated into phosphoric acid esters to produce chiral
Brønsted acids, and the VAPOL derivative was shown to be more
effective in the amidation7 and imidation8 of imines and in the
asymmetric reduction of imines,9 while both showed effective-
ness in the desymmetrization of aziridines depending on the
nucleophile.10 The VANOL and VAPOL ligands both showed
essentially equal ability to serve in catalysts for the Petasis
reaction11 and the hydroarylation of alkenes.12 Heteroatom DielsÀ
Alder reactions of imines with Danishefsky’s diene were both
faster and more enantioselective with a boron-VAPOL catalyst
than with the corresponding boron-VANOL catalyst.13 How-
ever, perhaps the most important application of these ligands is
the catalytic asymmetric aziridination of imines with diazo compounds
which utilizes the same boron-based catalyst as the heteroatom
DielsÀAlder reaction.14,15 Recent studies have revealed that this
boron-based catalyst is a rather unique chiral polyborate
Brønsted acid that contains a boroxine ring in which one of the
borons is spiro-fused to the VANOL or VAPOL ligands.14i,l,m In
the aziridination of imines with diazo compounds, the VANOL
catalyst 4 is superior for trans-aziridinations,14k and the VAPOL
The success of asymmetric catalysts derived from the VANOL
and VAPOL ligands has resulted in the need for syntheses of
these ligands that are more efficient than the original procedures
that we first published.3,4a In this report we describe cost-effective
and scalable racemic syntheses of both the VANOL and VAPOL
ligands and, in addition, resolution methods that reliably allow
for the separation of the two enantiomers of each ligand by
crystallization. The key step in both syntheses is the cycloaddi-
tion/electrocyclization cascade (CAEC) that is initiated by the
[2 + 2] cycloaddition of phenylacetylene with either phenyl or
2-naphthyl ketene (Scheme 2). Improved procedures are also
developed for the oxidative phenol coupling reactions that are
used to generate racemic VANOL and VAPOL. Finally, the pro-
cedures for the resolution of each ligand involving alkaloid salts
of their corresponding hydrogen phosphate derivatives have
been improved such that highly reproducible protocols are
now available for the crystallization of each of the diastereomeric
salts of each ligand.
2. BACKGROUND
We have recently developed a new synthesis of the VANOL
ligand that utilizes inexpensive materials and requires neither
special equipment nor low temperatures, which is easy to scale-
up since no chromatographic separations are required.19 This
synthesis begins with the chlorination of 1-naphthol, and treat-
ment of the resulting 4-chloro-1-naphthol with AlCl3 in benzene
Special Issue: Asymmetric Synthesis on Large Scale 2011
Received: June 7, 2011
Published: July 12, 2011
r
2011 American Chemical Society
1089
dx.doi.org/10.1021/op200088b Org. Process Res. Dev. 2011, 15, 1089–1107
|

Organic Process Research & Development
ARTICLE
Scheme 1
Scheme 2
prepared this complex on a 250 g scale. The initial product of the
benzannulation is a 1-phenanthrol that is acetylated to give phen-
anthrene 14. A thiol will cleave the methyl ether in the presence
of aluminum chloride and will simultaneously effect the reductive
deacetylation to give 2-phenyl-4-phenanthrol 11 in high yield.
This four-step process is quite efficient giving 11 in 52% overall
yield from 1-bromonaphthalene and relatively easy to scale up to
100 g or more since chromatography can be avoided in all
purifications. For example, the benzannulation reaction of car-
bene complex 13 and phenyl acetylene has been carried out on a
250 g scale.23 However, a significant deficit of this approach is the
cost of chromium hexacarbonyl, which at $6 per gram is not an
issue on small scale, but on a 100 g scale or larger it becomes
prohibitive. Thus, a more cost-effective synthesis of VAPOL was
sought and is the subject of the present study. We have also
examined the synthesis of VAPOL utilizing the Snieckus phenol
synthesis, an approach that requires the use of low temperatures
(À78 °C) and lachrymators.24 The yields for the key step fell off
when the scale was increased and some of the steps required
purification by chromatography, and thus this approach was not
considered in seeking a more cost-effective scalable synthesis
of VAPOL.
leads to a dienoneÀphenol rearrangement of an in situ gener-
ated cyclohexa-2,5-dienone and the formation of 3-phenyl-1-
naphthol 8 (Scheme 3). The last step is the oxidative phenol
coupling of 3-phenyl-1-naphthol 8 with air which gives racemic
VANOL in high yield. Considering the fact that the synthesis of
4-chloro-1-naphthol 7 has been reported on a 500 kg (3472 mol)
scale,20 that the remaining two steps in the synthesis of VANOL
2 shown in Scheme 3 employ very inexpensive reagents, and that
no chromatographic separations are needed, this synthesis
should be scalable for large-scale production of the VANOL
ligand.
An attempt to synthesize VAPOL by the approach to VANOL
shown in Scheme 3 would begin with the chlorination of
4-phenanthrol 9 (Scheme 4). The key intermediate is 2-phen-
yl-4-phenanthrol 11 which has been employed in the previous
synthesis of VAPOL via an oxidative dimerization.3 The dienoneÀ
phenol type rearrangement set up by the treatment of 1-chloro-4-
phenanthrol 10 with AlCl3 in benzene has never been investi-
gated. The reason is that while the chlorination of 4-phenanthrol
9 is known to give 1-chloro-4-phenanthrol 10 (along with 41% of
3-chloro-4-phenanthrol),21 4-phenanthrol 9 is not readily avail-
able, and the necessity for its synthesis22 would make this
approach far too long to be economical on a large scale.
The current method for the synthesis of VAPOL is shown in
Scheme 5 and involves the benzannulation of the naphthyl
carbene complex 13 with phenylacetylene as the key step.3
The carbene complex is prepared by the addition of 1-naphthyl-
lithium to chromium hexacarbonyl and then methylation by di-
methyl sulfate. The carbene complex is a crystalline red solid and
can be readily purified by crystallization, and we have routinely
3. CYCLOADDITION/ELECTROCYCLIZATION CASCADE
(CAEC) APPROACH TO VANOL AND VAPOL
We have previously published3 the three-step synthesis of
2-phenyl-4-phenanthrol 11 shown in Scheme 6 that begins with
the commercially available 2-naphthaleneacetic acid 15, the
parent member of a class of nonsteroidal anti-inflammatory
agents such as naproxen.25 Conversion of the carboxylic acid
15 to its corresponding acid chloride 16 is then directly followed
without purification by thermolysis with neat phenyl acetylene
1090
dx.doi.org/10.1021/op200088b |Org. Process Res. Dev. 2011, 15, 1089–1107

Organic Process Research & Development
ARTICLE
Scheme 3
Scheme 5
Scheme 6
Scheme 4
which gives the carboxylic ester 17. This intermediate also is not
purified, and the crude product is directly saponified to give
2-phenyl-4-phenanthrol 11 in 27% yield from the acid 15. The
yield for this process may not be viewed in such a negative light
when it is realized that two equivalents of the acid chloride 16 is
required for this reaction, one to generate the 2-phenyl-4-
phenanthrol 11 and one to trap it to give the ester 17. While
2-naphthaleneacetic acid 15 can be recovered from the saponi-
fication and recycled, the need to do this greatly reduces the effi-
ciency of this synthesis of VAPOL. Despite the flagging interest
in this approach that the low yield of 11 engendered, the opti-
mization of this synthesis of VAPOL was undertaken given the
simplicity of the chemical steps and the inexpensive nature of the
reagents.
In early attempts to scale up the VAPOL synthesis shown in
Scheme 6, it was found that the 2-phenyl-4-phenanthrol 11 that
was obtained after saponification could only be purified by care-
ful chromatography on silica gel.3 The crude product was
obtained as a black sticky tarry material from which no attempts
at crystallization were successful in providing 11 in any form with
increased purity. The mechanism by which the transformation of
16 to 17 occurs begins with the loss of HCl from the acid chloride
16 to give the ketene 18 that then is followed by a [2 + 2]
cycloaddition with phenyl acetylene to give the cyclobutenone
19 (Scheme 7). The electrocyclization cascade begins with ele-
ctrocyclic ring opening of the cyclobutenone to give the vinyl
ketene 20 which undergoes a 6eÀ electrocyclic ring closure to
give cyclohexa-2,4-dienone 21 and then, upon tautomerization,
gives rise to 2-phenyl-4-phenanthrol 11 that is then either
trapped by the acid chloride 16 or the ketene 18 to provide
the ester 17.
This cycloaddition/electrocyclization cascade (CAEC) in
the past has always been performed without a solvent,26 but a
recent report by Redic and Schuster27 finds that the yields can
be improved somewhat if decalin is employed as a solvent. In
our original report3 of the reaction without solvent, the acid 15
is converted to the acid chloride 16, and then the excess SOCl2
is removed under high vacuum to leave the acid chloride as a
yellow solid. This material is then heated with phenyl acet-
ylene to give the phenanthrol 11 in 27% yield after reductive
cleavage of the ester 17 with LAH (Table 1, entry 1). Inspired
by the report of Redic and Schuster,27 we repeated this
reaction in decalin to find that the yield slightly increases to
37% (entry 2). It was also found that mineral oil could give
1091
dx.doi.org/10.1021/op200088b |Org. Process Res. Dev. 2011, 15, 1089–1107

Organic Process Research & Development
ARTICLE
essentially the same yield (entry 3). The lack of tertiary
hydrogens in mineral oil gives it a safety advantage at these
high temperatures in that it should be less reactive with air at
these higher temperatures. Further optimization found that
190 °C is the optimal temperature, and 1.3 equiv of phenyl
acetylene gives material that was much easier to purify and was
cleaner after purification.
Given the low yield observed for 2-phenyl-4-phenanthrol 11
and the fact that half of the starting 2-naphthaleneacetic acid 15
is wasted when phenanthrol 11 is trapped, the idea was
conceived that a trapping agent could be found that could
compete with the in situ generated ketene in the reaction with
the phenol function in 11 but not otherwise affect the reaction.
In developing a synthesis of any phenol-containing compound,
protection of the phenol function is often required to prevent
side reactions of the hydroxyl group. The phenol function is
most often protected as either an ether or an ester. Formation of
ethers usually requires the presence of base since in most cases
the hydroxyl group itself is not a strong nucleophile. Another
problem associated with ether-forming agents is that they are
usually too volatile to be useful for reactions requiring condi-
tions with high temperatures. On the other hand, esterification
reagents (acid halides and anhydrides) are generally reactive to
the phenol function, thermally stable, and typically inexpensive.
A number of such reagents were examined, and the results are
summarized in Table 2. All reactions were carried out with 1.3
equiv of phenylacetylene to ensure that there is sufficient alkyne
present for the trapping agent to do the job as intended in which
case a minimum of 1.0 equiv would be needed. The reaction
with SOCl2 as the trap was carried out by not removing the
excess SOCl2 from the previous step in which the acid 15 is
converted to its acid chloride. This did not lead to the isolation
of phenanthrol 11 but rather to the formation of an unchar-
acterizable black tar (Table 2, entry 1). The reaction with
POCl3 added as the trap faired a little better, giving a small
amount of 11. A low yield of 11 was obtained with acetyl
chloride, but the quality of the isolated product was significantly
improved, as indicated by both the melting point and color of
the final product. This low yield could be related to the low
boiling point of acetyl chloride which is consistent with the fact
that benzoyl chloride gives nearly double the yield as acetyl
chloride. Nonetheless, it was still disappointing to find that the
50% yield ceiling could not be overcome. This was to be realized
with anhydrides as the trapping agent and, surprisingly, with
only one particular anhydride, iso-butanoic anhydride which
gave the phenanthrol 11 in 75% yield. It is really quite
remarkable that it gives a 20% higher yield than the isomeric
n-butanoic anhydride, and the reason for this is not understood
at this time. It was also unexpected that the n-butanoic ester
would be much more difficult to hydrolyze with aqueous KOH
than the iso-butanoic ester (entries 9 vs 10). The phenanthrol
11 could be liberated with either KOH or LAH, but the former
was deemed to be more desirable on large scale due to cost and
to the fact that the latter results in a quite exothermic reaction.
The fact that a 75% yield of phenanthrol 11 is obtained when
iso-butyric anhydride is used as the trap indicates that at least
three-fourths of the carboxylic acid 15 takes part in the
cycloaddition/electrocyclization cascade and that a maximum
of one-fourth of the phenanthrol 11 comes from the hydrolysis
of the 2-naphthaleneacetic acid ester 17 via the trapping of the
ketene 18 (Scheme 7) or the acid chloride 16 (Scheme 6). This
Scheme 7
Table 1. Solvents for the Cycloaddition/Electrocyclization Cascade (CAEC)a
entry
HCtCPh (equiv)
temp (°C)b
solvent (9 mL)
% yield 11c
mp 11 (°C)d
color of 11
1e
2
0.67
0.67
0.67
0.67
0.67
1.31
185
190
190
170
210
190
none
27
37
38
22
37
41
nd
brown
brown
brown
brown
brown
grey
decalin
144À5
140À1
150À1
147À8
140À1
3
mineral oil
mineral oil
mineral oil
mineral oil
4
5
6f
a Unless otherwise specified, each reaction was carried out on 76 mmol of 15 with 3.6 equiv of SOCl2. The phenanthrol 11 was liberated from 17 with 1.1
equiv of LAH in 100 mL of ether. b Oil bath temperature. c Isolated yield after chromatography on silica gel. d mp of 11 is 154À5 °C (ref 3). e Data from
ref 3: ester was converted to 11 with KOH in MeOH/H2O for 20 h at 25 °C. f 1.72 equiv of LAH used.
1092
dx.doi.org/10.1021/op200088b |Org. Process Res. Dev. 2011, 15, 1089–1107

Organic Process Research & Development
ARTICLE
Table 2. Survey of Trapping Agents for Phenanthrol 11a
entry
trap
SOCl2
trap bp (°C)
ester cleavage
% yield 11b
mp 11 (°C)c
color of 11d
1e
2
3f
76
105
51
LAH
LAH
LAH
KOH
KOH
LAH
LAH
LAH
KOH
KOH
KOH
ND
16
nd
nd
POCl3
128À30
148À9
152À4
125À7
152À3
153À5
153À4
151À2
149À50
nd
black
MeCOCl
23
light yellow
dk brown
chocolate
light yellow
light yellow
yellow
4
PhCOCl
198
196
140
140
168
182
198
203
44
51g
5
n-heptylCOCl
(MeCO)2O
(MeCO)2O
(EtCO)2O
(i-PrCO)2O
(n-PrCO)2O
(ClCH2CO)2O
6
7h
17
48
8
31
9
75
55i
orange
10
11
orange
trace
nd
a Unless otherwise specified, each reaction was carried out on 76.4 mmol (14.2 g) of 15 with 3.6 equiv of SOCl2. The phenanthrol 11 was liberated from
17 either with 1.1 equiv of LAH in 100 mL of ether at 25 °C for 12 h or with 4.7 equiv of KOH in 100 mL of H2O at 100 °C for 12 h. b Isolated yield after
chromatography on silica gel. ND = not detected. c mp of 11 after silica gel chromatography (lit3 mp 154À5 °C). nd = not determined. d Color of 11 after
silica gel chromatography. e Excess SOCl2 not removed prior to thermolysis with HCtCPh. f Reaction time is 36 h. g Yield of octanoate ester of 11.
h Reaction time is 64 h. i Hydrolysis performed twice at 120 °C with 11.7 equiv of KOH.
Table 3. Optimization of the Synthesis of Phenanthrol 11 with (i-PrCO)2Oa
entry
SOCl2 (equiv)
(i-PrCO)2O (equiv)
time (h)
temp (°C) b
% yield 11c
mp 11 (°C)d
color of 11e
% yield 26f
recovery 15f
1g
2
3.6
3.6
3.6
3.6
3.6
1.8
1.8
3.6
3.6
3.6
3.6
2
2
1
4
2
2
2
2
2
2
2
24
24
24
24
24
24
48
48
48
48
48
190
190
190
190
210
190
190
190
170
190
190
68
71
43
55
51
65
81
79
67
77
47
151À2
147À8
149À51
152À3
152À3
152À3
154À5
152À3
153À4
156À7
nd
brown
orange
orange
orange
orange
orange
yellow
yellow
orange
yellow
brown
nd
nd
nd
nd
nd
nd
nd
8
nd
nd
nd
nd
nd
nd
nd
13
25
nd
ND
3
4
5
6
7
8
9
11
8
10h
11i
ND
a Unless otherwise specified, each reaction was carried out on 76.4 mmol (14.2 g) of 15 with no solvent. The phenanthrol 11 was liberated from its
iso-butyl ester by refluxing with 5.8 equiv of KOH in 100 mL of H2O for 12 h. b Oil bath temperature. c Isolated yield after chromatography on silica gel.
nd = not determined. d mp of 11 after silica gel chromatography (lit.3 mp 154À5 °C). e Color of 11 after silica gel chromatography f Isolated yield after
chromatography on silica gel. ND = not detected. g Reaction performed in mineral oil (9 mL) as solvent, and 11 was liberated by heating with 5.8 equiv of
KOH in 100 mL of H2O at 120 °C (oil bath temperature) for 12 h. h The alkyne was added in four equal portions at 0, 6, 12 and 24 h. i Alkyne was added
by syringe-pump over a period of 30 h, and the scale was 306 mmol of 11.
was verified by carrying out the reaction without the final
hydrolysis which led to the isolation of the iso-butyric ester
22 in 74% yield and the 2-naphthaleneacetic acid ester 17 in 7%
yield (eq 1).
1093
dx.doi.org/10.1021/op200088b |Org. Process Res. Dev. 2011, 15, 1089–1107

Organic Process Research & Development
ARTICLE
Scheme 8
26 and which is generally obtained in ∼10% yield (Scheme 8).
The origin of this product is most likely from a [2 + 2]
cycloaddition of phenylacetylene with the ketene 20, a process
which is in competition with electrocyclic ring closure to give the
phenanthrol 11. The resulting cyclobutenone 23 could then be
expected to begin its own electrocyclization cascade that begins
with electrocyclic ring opening to give the dienyl ketene 24 and
then an electrocyclic ring closure to give the cyclohexadienone
25 and finally tautomerization to the byproduct 26. The by-
product phenol 26 can be separated from the desired phenol 11
by chromatography on silica gel, but on large scale, it is not clear if
the two phenols will be easily separable by crystallization. Thus,
some effort was put forth to minimize the formation of this
byproduct. Mechanistically, the partition point is the ketene 20,
and the amount of the two phenol products is dependent on the
competition between the electrocyclic ring closure to 21 and the
[2 + 2] cycloaddition to give 23. This competition will thus be
dependent on the concentration of alkyne, and thus the distribu-
tion between the two products was examined as a function of the
rate of addition of the alkyne that, in all but the last two entries of
Table 3, is added all at once. In entry 10, the alkyne is added in
four equal portions over a 24 h period, but the ratio of 11 to 26 is
virtually unchanged (entries 8 vs 10). It is not clear why the
portion-wise addition did not illicit a change in the ratio of the
two products. The formation of the byproduct could be sup-
pressed by the slow addition of the alkyne over a period of 30 h by
syringe in a reaction that was carried out on 4-fold increased scale
(entry 11). None of the phenol 26 could be detected, but this was
gained at an immense price since the yield of the phenanthrol 11
fell to 14%. No further attempts to minimize the formation of 26
were pursued.
The purpose of the 11 reactions summarized in Table 3 is to
identify factors that will lead to optimization of the reaction with
iso-butyric anhydride as the trapping agent. The first finding is
that the use of mineral oil as solvent is unnecessary (entries 1 vs 2),
and this is useful in simplifying the purification of the
phenanthrol 11. The results in entries 2À4 served to identify
the optimal amount of trapping agent as two equivalents. As was
found by the survey of conditions without a trapping agent
(Table 1), the optimal temperature for the reaction with the
iso-butyric anhydride trap is 190 °C (entries 2 and 5 and 8 and 9).
A notable decrease in yield (20%, entries 2 vs 5) was observed when
the temperature was raised to 210 °C. The reason for this is not
clear, but it was observed that iso-butyric anhydride (bp 182 °C)
began to boil at this temperature, and it is possible that this
retarded the condensation of the alkyne (bp 144 °C) and its
return to the reaction flask. Finally, it was observed from this
study that the amount of excess SOCl2 is not crucial (entries 2
and 6), and a reaction time of 48 h is superior to that of 24 h
(entries 6 and 7). The chlorination of acid 15 can also be
performed with (COCl)2 instead of SOCl2. When the reaction
indicated in entry 2 of Table 3 was repeated with the acid chloride
16 generated from the acid 15 with oxalyl chloride according to
the procedure of Bandarage,28 the phenanthrol 11 was obtained
in 67% yield if the oxalyl chloride was added all at once and in a
lower yield (47%) if it was added slowly over a period of 1 h. The
highly vigorous reaction that ensued when oxalyl chloride was
added all at once is a decided detriment, and along with the price
difference, thionyl chloride is thus deemed the reagent of choice
for the large-scale synthesis of phenanthrol 11.
The two-alkyne phenol 26 could be separated from the de-
sired phenanthrol 11 by chromatography on silica gel, and on
small scale (14.2 g) this proved to be practical (Table 4). The
reaction run under the optimal conditions established in Table 3
(entry 8) gave a 75% isolated yield of 11 by chromatography on
silica gel (Table 4, entry 1). In an effort to develop a protocol that
avoids chromatography, the reaction was repeated (entry 2), and
the product was crystallized from the crude reaction mixture
from dichloromethane. The first crop gave pure 11 in 37% yield,
but it did not prove possible to obtain any additional 11 in pure
form by further crystallization under any condition. The second
crop was typically impure 26. However, loading the concentrated
mother liquor onto a silica gel column gave additional pure 11 for
a total of 57% yield. An additional complication arose from the
presence of some nonpolar black impurities that interfered with
both chromatographic and crystallization purification techni-
ques. The amount of these impurities and the ultimate quality
of the product 11 were found to be dependent on the source of
2-naphthylacetic acid 15. One particular batch of 15 gave
significantly more of these black impurities than others (entry 3).
The source of this problem was not obvious since the acid 15
used in entry 3 was nearly white in appearance and had a clean 1H
NMR spectrum, and its melting point matched published values.
Crystallization of the material used in entry 3 did not markedly
reduce the amount of black impurities formed in the reaction. It is
suspected without evidence that this compound may have been
prepared by a WillgerodtÀKindler reaction29 from 2-naphthyla-
cetophenone and that residual sulfur is not completely removed
by crystallization. The amount of these black impurities was
minimal from the reaction performed with the acid 15 obtained
from Aldrich Chemical Co. (Table 4, entry 4).
The cycloaddition/electrocyclization cascade produces a by-
product which was identified as 2-(2-naphthyl)-3,5-diphenylphenol
1094
dx.doi.org/10.1021/op200088b |Org. Process Res. Dev. 2011, 15, 1089–1107

Organic Process Research & Development
ARTICLE
Table 4. Optimization of the Synthesis of Phenanthrol 11 with (i-PrCO)2O
entry
15 (g)
source of 15
purification
% yield 11a
% yield 26
1
2
3
4
5
6
14.2
14.2
14.2
14.2
57
Milestone
Milestone
O-Chem
Aldrich
column only
crystallizationb/column
75
57
61
74
57
64
8
7
column only
crystallizationb/column
crystallizationb/filter/crystallizationc/crystallizationd
crystallizationb/filter/crystallizationc/crystallizationd
nd
8
Milestone
5
57
Milestone
5
a Combined isolated yield of 11. b The first crystallization gives 11 from CH2Cl2. The second crystallization gives 26 from i-PrOH. d The third
c
crystallization gives additional 11 from hexanes/dichloroethane (2:1).
In an effort to find a method for the separation of 11 and 26 by
crystallization, a number of solvents were screened, and it was
found that isopropanol was superior to ethanol, methanol, ethyl
acetate, and toluene. It was noted that while 157.4 g of 11 would
dissolve in 100 mL of boiling iso-propanol this stood in stark
contrast to only 1.75 g of phenanthrol 26. This suggested that the
byproduct be removed first by initial crystallization from iso-
propanol. However, all attempts to do this invariably led to the
collection of a solid that was a mixture of phenol 26 and
phenanthrol 11. Thus, it was realized that any final protocol
would involve collection of the phenanthrol first, and the opti-
mized procedure is shown in entries 5 and 6 in Table 4. The crude
product is first crystallized from CH2Cl2 to give the phenanthrol
11 (42%, entry 5). It is difficult to crystallize either product from
the residue of the mother liquor because of the presence of the
nonpolar black impurities. Thus, it was found best to filter the
residue through a bed of silica gel in a sintered glass funnel with
suction from a water aspirator with a 1:1 mixture of CH2Cl2 and
hexanes. The first liter of eluent was discarded and consisted of
∼10 g (the amount depends on the source of 15) of a black tarlike
substance which contained a small amount of 26. The remaining
material (a mixture of 11 and 26) was flushed from the silica gel
and crystallized from isopropanol to give a 5% yield of the pure
phenol 26. A final crystallization from hexanes and 1,2-dichlor-
oethane gives an additional 15% yield of pure 11. This protocol
proved tobe reproducible overa numberof runs, and anadditional
example is given in Table 4 (entry 5). It should be noted that for
the 2-naphthylacetic acid 15 obtained from Aldrich Chemical Co.
very little of these black impurities were observed (Table 4, entry 4),
and thus the crystallization of 11 was not impeded. In this case,
crystallization gave a 58% yield of phenanthrol 11 from the first crop
from dichloromethane. An additional 16% of 11 was obtained by
column chromatography, but it is suspected that additional pure
11 could have been obtained by taking a second crop.
Table 5. Optimization of the Synthesis of 3-Phenyl-1-
naphthol 8 with (i-PrCO)2Oa
% yield of 8
entry 27 (mmol) 1st cropb 2nd cropc columnd total 8 (%)e total 8 (g)e
1
2
3
74.7
374
374
À
43
44
À
13
17
70f
11
7
70
67
68
11.5
54.7
56.0
a After neutralization, the crude reaction mixture is washed with sat. aq
Na2CO3 to remove iso-butyric acid 28. b The first crystallization gives 8
from hexanes/CH2Cl2 (3À4:1). c The second crystallization gives 8
from hexanes/CH2Cl2 (3:1). d The mother liquor is stripped of solvent
and loaded onto a silica gel column and eluted with hexanes/CH2Cl2
(2:1). Collection of fractions containing 8 gives a yellow solid that is not
completely pure. Crystallization from hexanes/CH2Cl2 (3:1) gives
additional pure 8 in the indicated yield. e Total isolated yield of 8. f The
iso-butyric acid 28 is removed by distillation, and then the entire crude
reaction mixture was loaded onto a silica gel column and eluted with
hexanes/CH2Cl2 (2:1). Collection of the fractions containing 8 gives
pure 8 as a biege solid.
steps from the commercially available 4-chloro-1-naphthol 7 or
three steps from the less expensive 1-naphthol.19
With the successful development of a scalable protocol for the
preparation of VAPOL via the CAEC cascade, the immediate
question is that would this work as well for the synthesis of
VANOL? At first blush, this might be considered as unnecessary
since as summarized in Scheme 3 we have recently developed an
inexpensive and technically simple synthesis of VANOL in two
1095
dx.doi.org/10.1021/op200088b |Org. Process Res. Dev. 2011, 15, 1089–1107

Organic Process Research & Development
ARTICLE
reaction has gone to about 50À60% completion. It was found
best to separate out the product by crystallization of the reaction
mixture and then remove the solvent from the mother liquor and
subject the oily black residue to the reaction conditions. This
procedure was performed a total of six times to give an overall
81% yield of racemic VANOL after a total reaction time of 89 h.
The extremely tedious nature of this procedure prompted a
search for a suitable solvent for this reaction such that constant
contact of the naphthol 8 and air could be maintained.
The provision of a more homogeneous milieu should assist in
the more efficient diffusion of air into the reaction mixture and
thus hopefully greatly reduced times to effect complete conver-
sion. After consideration of several solvents with suitable high
boiling points, the final choice proved to be mineral oil. Given the
exposure of the solvent to high temperatures in the presence of
air for extended times, it was desired to employ a solvent without
tertiary hydrogens to avoid peroxide formation. In addition,
mineral oil should be easy to remove by simply washing the
crude reaction mixture with hexanes since racemic VANOL and
VAPOL are essentially insoluble in hexanes. In this regard, it is
interesting to note that optically pure VANOL and VAPOL are
highly soluble in hexanes. As indicated by the data in Table 6, the
use of mineral oil as solvent for the phenol coupling was very
successful and has several advantages over the neat reaction. First
the reactions go to completion without stopping half way, and in
a much shorter period of time, and unexpectedly, it was found
that once the reaction reaches completion the crude product is an
orange solid instead of a black tarry material. VANOL can be
easily isolated and purified by simple filtration and subsequent
crystallization.
Not all of the optimized parameters translated from the neat
reaction to the reaction in mineral oil. This was especially true of
the reaction temperature which was optimal at 175 °C for the
neat reaction; however, in mineral oil the yield sharply fell off at
this temperature (46%) (Table 6, entry 7). A survey of tempera-
tures found that reactions at 135 and 145 °C did not go to
completion even with extended reaction times (entries 2 and 3).
The reaction at 155 °C does go to completion in 38 h, but the
reaction at 165 °C was found to be ideal since it goes to
completion in only 17 h and gives a higher yield (89%) than
the reaction at 155 °C. A reaction at half the concentration gives a
slightly lower yield and requires double the reaction time (entries
5 vs 6), and thus the optimal conditions for the reaction in
mineral oil are those in entry 5. It was demonstrated that with the
optimal conditions the reaction could be readily scaled-up to 50
g, and this proved to be reliably reproducible giving 85À86%
yield in three different runs (entries 8, 10, and 11). Thus, in the
search for an improved and more general access to VANOL, the
possibility that the presence of solvent might ameliorate the
destructive effects of the harsh conditions motivated us to carry
out the reaction with mineral oil as the reaction media. As a result,
the incomplete conversions and associated extended processing
time that plagued the earlier studies gave way to this more
synthetically useful methodology.
However, the CAEC cascade has a potentially significant advan-
tage in a VANOL synthesis in that it can provide for the rapid
synthesis of a family of VANOL ligands (eq 2). The commercial
availability of a significant number of substituted phenyl acetic
acid derivatives empowers the CAEC approach to VANOL since
it would enable direct access to a diverse family of chiral ligands.
This diversity could, in all likelihood, not be matched by the
dienoneÀphenol approach to VANOL (Scheme 3) given the
paucity of substituted 1-naphthols that are commercially avail-
able and the fact that the chlorination and dienoneÀphenol
rearrangement steps may be influenced by substituents present in
the naphthol.
Thus, it was pleasing to find that the CAEC cascade to
VANOL proceeded smoothly with the conditions optimized
for VAPOL to provide access to 3-phenyl-1-naphthol 8 in good
yields on large scale (Table 5). In this case, the acid chloride 27
is commercially available and inexpensive, and the CAEC
reaction is a lot cleaner since the black nonpolar impurities
seen in the CAEC cascade in the synthesis of 11 are not
observed here. The major side product in this reaction is iso-
butyric acid 28, and this can be easily removed by washing the
crude reaction mixture with aqueous sodium carbonate. The
naphthol 8 is not lost in this extraction, even though it can be
extracted into aqueous sodium hydroxide.3 Two crops of the
product can be taken by crystallization from hexanes and
methylene chloride which provides the majority of the pure
product 8 as a white fluffy solid in 56À61% yield. An additional
7À11% yield of 8 can be obtained from the residue from the
mother liquor upon purification by column chromatography on
silica gel. The latter may or may not be cost-effective in terms of
money and time given the relatively low cost of the starting
materials.
4. OXIDATIVE PHENOL COUPLING IN AIR TO GIVE
RACEMIC VANOL AND VAPOL
The final step in the synthesis of VANOL and VAPOL is the
oxidative phenol coupling of 3-phenyl-1-naphthol 8 and 2-phen-
yl-4-phenanthrol 11, respectively. While ferric chloride is the
oxidant of choice for the phenol coupling of 2-naphthol in
the preparation of BINOL, we found in our initial studies in the
synthesis of VANOL and VAPOL that metal oxidants were
unsuitable for the related coupling step for these ligands.3
We took a cue from the total synthesis of gossypol by Edwards
and Cashaw in which they found that oxidative phenol coupling
of a substituted 1-naphthol was best performed with air as the
oxidant.31 Following their protocol, the first synthesis of VANOL
was achieved simply by melting 3-phenyl-1-naphthol 8 in a test
tube in the presence of air to give racemic VANOL 2 in 87% yield
(1 g scale).3 The major drawback of this method is that racemic
VANOL has a melting point (231À3 °C) that is above the opti-
mal temperature that is needed for the phenol coupling step. This
leads to a solidification of the reaction mixture as the reaction
progresses and unreacted naphthol 8 becomes trapped and
removed from exposure to air. On small scale (1 g), this can be
countered by occasionally breaking up the solid mass in small
pieces with a stirring rod which allows the reaction to go to
completion. This, however, becomes a serious issue when the
reaction is scaled up as is illustrated by the 100 g scale reaction
shown in entry 1 of Table 6.32 In this reaction, the melted
naphthol 8 is heated under a slow flow of air until the stir bar
stops which on a 100 g scale is about 20 h at which time the
The original procedure3 for the preparation of VAPOL via the
oxidative coupling under neat conditions on large scale was
problematic for the same reasons that the original synthesis of
VANOL was problematic. With a simple and high-yielding
procedure using mineral oil for the synthesis of VANOL
(Table 6) in hand, effort was turned to extend the same protocol
for the dimerization of 2-phenyl-4-phenanthrol 11. Indeed, the
same protocol applied to the syntheses of racemic VAPOL led to
1096
dx.doi.org/10.1021/op200088b |Org. Process Res. Dev. 2011, 15, 1089–1107

Organic Process Research & Development
ARTICLE
Table 6. Phenol Coupling in the Preparation of Racemic VANOL 2a
entry
8 (g)
time (h)b
[8] (M)
temp (°C)
% yield crude 2
mp crude 2 (°C)c
% yield 2d
mp 2 (°C)c
1e
2f
3f
4
100
5
89
72
68
38
17
40
20
32
24
32
24
À
175
135
145
155
165
165
175
165
165
165
165
nd
nd
nd
nd
91
89
nd
93
88
93
92
nd
81
80
74
80
89
83
46
86
72
86
85
nd
0.92
0.92
0.92
0.92
0.46
0.92
0.92
0.92
0.92
0.92
nd
nd
5
nd
nd
5
nd
nd
5
5
224À9
nd
231À2
nd
6
5
7
5
nd
nd
8
20
50
50
50
224À9
215À7
224À7
224À9
231À2
230À1
231À2
231À2
9g
10
11
a Unless otherwise specified, each reaction was carried out at the indicated concentration in mineral oil with an air flow of 0.13 L/min directed over the
surface of the reaction mixture. b Unless otherwise specified, the time is indicated for complete conversion as monitored by TLC. c The reported mp for
racemic 2 is 231À3 °C (ref 3). d The yields are isolated material obtained by crystallization from CH2Cl2. e This reaction was run neat. The reaction was
run until the stir bar stopped (20 h) and then worked up and the product 2 collected by crystallization from CH2Cl2. The crude mixture containing
unreacted 8 was resubjected to the reaction conditions until the stir bar stopped again at which point more of the product 2 was collected and the
unreacted 8 resubmitted to the reaction conditions. This cycle was performed a total of 6 times to give a total of an 81% yield of 2 in 89 h of reaction time.
f These reactions were stopped prior to full conversion due to the late appearance of byproduct. g Reaction was initially stopped after 16 h but was not
complete. The crude mixture was then exposed to the same conditions for an additional 8 h at which point the reaction was complete.
Table 7. Phenol Coupling in the Preparation of Racemic VAPOL 3a
entry
11 (g)
time (h)b
temp (°C)c
% yield crude 3
mp crude 3 (°C)d
% yield 3e
mp 3 (°C)d
1
2
3
4
5
6
7
4.3
5
33
36
36
36
36
36
36
190
180
170
170
170
170
170
86
93
95
91
98
91
99
nd
84
70
84
81g
89
93
97
nd
nd
nd
5
nd
nd
50f
50f
50f
48h
293À5
297À9
296À9
300À1
310À1
308À9
306À9
310À1
a Unless otherwise specified, each reaction was carried out at 0.75 M in mineral oil with an air flow of 0.15 L/min directed over the surface of the reaction
mixture. b The time is indicated for complete conversion as monitored by TLC. c Oil bath temperature. d The reported mp for racemic 3 is 312À313.2 °C
(ref 30). e The yields are isolated material obtained by crystallization from CH2Cl2 (2 crops) in entry 1, from EtOAc (2 crops) in entries 2À3, and from
toluene (2 crops) in entries 4À7. f 11 prepared as indicated in Scheme 5. g 1.6 g of 11 was recovered. h 11 prepared as indicated in Table 4.
the same advantages seen for this procedure with VANOL,
including a much cleaner reaction and much greater ease in
purification of the product (Table 7). Upon completion of the
reaction, removal of the mineral oil by washing with hexanes
1097
dx.doi.org/10.1021/op200088b |Org. Process Res. Dev. 2011, 15, 1089–1107

Organic Process Research & Development
ARTICLE
Scheme 9
gives the crude product as a brown powder rather than as a black
tarry material that is observed under the neat reaction.3 The tem-
perature is not as crucial to the success of the coupling reaction
here as it is for the synthesis of VANOL (entries 1À3), and
170 °C was chosen as optimal for this reaction. The reaction can
be scaled up to 50 g with no detrimental effects, and the racemic
VAPOL can be purified by crystallization from toluene (2 crops)
to give 89À97% yield of 3 as light brown crystals. The cleanest
material and highest yield was obtained from the reaction of a
batch of 2-phenyl-4-phenanthrol 11 that was produced by the
CAEC cascade (Table 4) and gave 3 with a mp of 310À1 °C
(lit.30 312À313.2 °C). It is interesting to note that racemic
VAPOL has a melting point that is 86 °C higher than the optically
pure VAPOL, a differential that is among the largest ever
recorded for an organic compound.30
temperature is sufficient to generate 29 in 96% yield in 6 h.
The original method for the separation of the diastereomeric salts
was to dissolve both in ethanol and selectively crystallize Salt I
(31) which contains the (S)-enantiomer of the VANOL hydro-
gen phosphate.33 With time we found that ethanol did not always
give as high of purity of salt I as our original specifications would
indicate.3 We found that the source of variability was the very low
solubility of Salt I in ethanol which required the use of undesired
large quantities of solvent on large scale. In pursuit of alternative
solvents, it was found that both salts are soluble in acetone and
toluene, while neither is particularly soluble in ethyl acetate,
hexane, and isopropanol. Reasonable solubility differentials were
found for dichloroethane and dichloromethane with Salt II being
the least soluble. Thus, it was decided to attempt to reverse the
order of salt collection.
Extensive studies on dichloromethane as the solvent for reso-
lution never led to the isolation of Salt II in high optical purity in a
single crystallization. This repeatedly resulted in a salt that was
greatly enriched in Salt II, but one was always left with the need
for enhancement strategies. 1,2-Dichloroethane, on the other
hand, gave much higher yields of Salt II in the first crop, but the
de was only about 95%. This was found to be due to the fact that
Salt II was coming out too fast. To slow this down, it was found
that the slow addition of a solution of brucine in dichloroethane
to a solution of VANOL hydrogen phosphate 29 in dichlor-
oethane gave much higher purity for Salt II. It was also found that
performing these manipulations under nitrogen leads to signifi-
cantly improved yields and purities for Salt II, and this may be
related to the formation of a white precipitate when solutions of
brucine are overexposed to air. Finally, differences were found
between technical grade (À)-brucine and 98% (À)-brucine. The
higher purity grade gave lower yields of Salt II but with a higher
purity. It was found that a higher purity of Salt II could be
5. RESOLUTION OF VANOL WITH (À)-BRUCINE
The protocol for the resolution of VANOL outlined in
Scheme 9 is the culmination of significant effort over an extended
period of time devoted to streamline the process and to make it
more reliable for providing access to optically pure VANOL in
high yields with crystallization methods alone. Only the final
purification of the optical pure ligand employs column chroma-
tography. The resolution is based on the separation of diaster-
eomeric salts formed from the reaction of racemic VANOL
hydrogen phosphate 29 with (À)-brucine 30. The former is
generated in high yield from the reaction of racemic VANOL
with POCl3 followed by hydrolysis with water. Our original
resolution procedure3 involved heating racemic VANOL with
POCl3, but we have since found that this is unnecessary and can
on occasion result in the formation of significant unidentified
byproduct. The reaction of VANOL with POCl3 at room
1098
dx.doi.org/10.1021/op200088b |Org. Process Res. Dev. 2011, 15, 1089–1107

Organic Process Research & Development
ARTICLE
Scheme 10
achieved with the less expensive technical grade brucine if care
was taken not to transfer small amounts of insoluble material
present in the dichloroethane solutions of brucine. All of these
observations taken together lead to the formulation of a protocol
that produces an 87% yield of Salt II with greater than 99% de in a
single crop. Furthermore, an 83% yield of Salt I of greater than
99% de can be obtained in a single crop from the mother liquor
by crystallization from ethanol. Finally, the pure enantiomers of
VANOL can be liberated from these salts in a single step by
reduction with Red-Al in high yields. This is a dramatic improve-
ment of the original method3 which involved treatment of the
salts with HCl to generate the optically pure enantiomers of
VANOL hydrogen phosphate which are then first methylated to
give the VANOL methyl phosphates, and these are then finally
reduced with Red-Al to give the pure ligands.
failing was the finding that Salt II can be much more reliably and
efficiently crystallized from a mixture of acetone and isopropanol.34
After crops of Salt I and Salt II are each collected, this sequence is
then repeated until all of the Salts have been collected (usually a
total of three cycles). This process has proven to be virtually
infallible giving high yields and diastereomeric purity of both salts
each time it is performed. Finally, the optically pure (R)- and (S)-
enantiomers of VAPOL can be liberated from Salt I 35 and Salt II
36 with the same procedure described above for VANOL in 90
and 91% yields, respectively.
7. TIME-LAPSED PERFORMANCE OF THE VANOL AND
VAPOL LIGANDS IN AN AZIRIDINATION REACTION
The methods described herein for the synthesis and resolution
of VANOL and VAPOL involve relatively inexpensive starting
materials, are reliable and reproducible, do not involve the chro-
matographic separation of any intermediate, and thus are immi-
nently suitable for the large-scale preparation of these chiral
ligands. It was deemed useful that the development of these
efficient methods for the production of VANOL and VAPOL
be coupled with an investigation of long-term stability of these
ligands which in turn would be reflective of the long-term ability
of these ligands to function as effective ligands in asymmetric
catalysts. We choose to monitor this effectiveness with time in
terms of their ability to provide catalysts for the asymmetric
aziridination of imines with ethyl diazoacetate.14g This was
monitored for the reaction of imine 37 to generate aziridine 39
with catalysts prepared from both the VANOL and VAPOL
6. RESOLUTION OF VAPOL WITH (À)-CINCHONIDINE
The original method for the resolution of VAPOL is related to
that of VANOL but uses (À)-cinchonidine instead of (À)-brucine.
The overall process presented in Scheme 10 is a modification of
the original version and features a much more reliable crystal-
lization protocol and the same streamlined method for liberation
of the optically pure VAPOL ligands that was described above for
VANOL. The original crystallization protocol that we published
in 1996 involved the crystallization of both salts from ethanol.3 In
practice, we have found that this works much better for Salt I than
it does for Salt II. It is often disappointing to find that Salt II is
obtained in low yields and with variable de’s. A solution to this
1099
dx.doi.org/10.1021/op200088b |Org. Process Res. Dev. 2011, 15, 1089–1107

Organic Process Research & Development
ARTICLE
Table 8. Evaluation of Storage Protocols for the VANOL and VAPOL Ligandsa
0 years
% yield 39c
0.5 years
1.5 years
2.5 years
storage conditionsb
ligand
% ee 39
% yield 39c
% ee 39
% yield 39c
% ee 39
% yield 39c
% ee 39
A
B
refrigerator (3 °C)
under nitrogen
room temp under argon
Al foil
VANOL
VAPOL
VANOL
VAPOL
VANOL
VAPOL
VANOL
VAPOLd
90
84
91
85
83
83
83
93
93
90
90
90
90
90
94
90
86
81
91
85
86
85
90
85
93
92
92
92
92
91
93
91
90
89
87
83
86
85
88
89
93
92
94
94
92
94
91
92
87
86
89
85
84
87
91
91
95
95
91
92
93
92
94
95
C
D
room temp
under argon
room temp
under air
a Ligand quality was judged by their performance in aziridination reactions which were carried out over a period of 2.5 years. The aziridination catalyst
was prepared, and the azirdination reactions were performed by procedure F in ref 14g. Unless otherwise specified, the ligands were all off-white in color,
and no change in color or appearance was noted in the course of the study. The VAPOL used in this study was a crystalline form consisting of two
molecules of VAPOL and one molecule of CH2Cl2. The VANOL did not contain CH2Cl2. b Condition A: ligands stored in a refrigerator (3 °C) in a
brown bottle under nitrogen and sealed with parafilm. Condition B: ligands stored in a cabinet in a brown bottle under argon sealed with parafilm and
wrapped in aluminum foil. Condition C: ligands stored on a bench in a brown bottle under argon and sealed with parafilm. Condition D: ligands stored
on a bench in a brown bottle under air and sealed with parafilm. c Isolated yield after chromatography on silica gel. d After 6 months of storage, a thin light
orange layer appeared on the surface of the VAPOL sample. This was stirred into the sample and was not observed to reappear after 1.5 or 2.5 years.
ligands over a period of two and a half years (Table 8). This
includes samples of each ligand that were stored in four different
ways: A, in a refrigerator under nitrogen; B, at room temperature
under argon and protected from light by aluminum foil; C, at
room temperature under argon; and D, at room temperature stored
in the presence of air. It was surprising to find that there was not
significant variation in either the yield or asymmetric induction in
any of these reactions with either ligand under any of the four
conditions of storage, even when stored at room temperature in
the presence of air. The color of all samples of the ligands is an
off-white, and with a single exception, no change in either the
color or appearance of the ligands was observed. For the sample
of the VAPOL ligand stored under air at room temperature, a
thin orange layer was observed on the top of the ligand sample
after 6 months, but this did not appear to have an effect on the
aziridination reaction. This small amount of colored material was
mixed in with the rest of the sample, and curiously, this dis-
coloration did not make a reappearance after 18 months or after
30 months.
In conclusion, the present work describes a synthesis of the
VANOL and VAPOL ligands that represent a dramatic improve-
ment in terms of cost-effectiveness and scalability over our
previously published syntheses in 1996.3 The key step in each
synthesis is a cycloaddition/electrocyclization cascade (CAEC)
that begins with a [2 + 2] cycloaddition of phenylacetylene with
an aryl-substituted ketene. The essential finding that enabled this
process as synthetically efficient was that the yields could be
improved from 27 to 81% if isobutryic anhydride is added to trap
the phenanthrol product 11 before it reacts with the ketene or its
acid chloride precursor. The scalability of both syntheses was
greatly enhanced by the finding that the air-mediated oxidative
phenol coupling step in the formation of the racemic VANOL
and VAPOL could be carried out in mineral oil. Finally, more
reliable methods for the resolution of each ligand were developed
involving VANOL and VAPOL hydrogen phosphate. Classic
resolution procedures employing alkaloids (brucine for VANOL
and cinchonidine for VAPOL) with the proper solvent systems
and proper procedural controls have produced highly reprodu-
cible protocols for the separation of the diastereomeric salts for
each ligand by crystallization.
8. EXPERIMENTAL SECTION
General Procedure to Prepare Acid Chloride 16. A single-
neck 250 mL round-bottom flask equipped with a large 48 Â 18 mm
oval magnetic stir bar and a condenser was charged with 14.23 g
(76.42 mmol, from Milestone Pharm Tech USA Inc., Lot #:
T1161) of acid 15 and SOCl2 (20 mL, 274 mmol). The top of the
condenser is vented to a bubbler and then into a beaker filled with
aq NaOH to trap acidic gases (HCl and SO2). The mixture was
heated to reflux for 1 h in a 90 °C oil bath, and then all of the
volatiles were carefully removed by swirling it under high vacuum
(1 mmHg) for 1 h with a second liquid N2 trap to protect the
pump. Crude 16 (ca. 16.7 g) could be obtained as a yellow solid
(color was different with 15 from other venders). Spectral data
for 16: white wax-like solid; 1H NMR (CDCl3, 300 MHz) δ 4.28
(s, 2H), 7.34À7.37 (dd, 1H, J = 4.5, 1.5 Hz), 7.47À7.52 (m, 2H),
7.35 (s, 1H), 7.98À7.86 (m, 3H).
Preparation of 2-Phenyl-4-phenanthrol 11 on a 14 g Scale
with Purification by Chromatography (Table 4, entry 1). Acid
15 (14.23 g, 76.42 mmol) from Milestone Pharm Tech USA Inc.
(Lot #: T1161) was used to prepare acid chloride 16 with SOCl2
(20 mL, 274 mmol) in the same manner as described above. The
flask containing the resulting 16 (ca. 16.7 g) was filled with argon
1100
dx.doi.org/10.1021/op200088b |Org. Process Res. Dev. 2011, 15, 1089–1107

Organic Process Research & Development
ARTICLE
and charged with phenylacetylene (11 mL, 100.2 mmol) and
(i-PrCO)2O (25 mL, 146.2 mmol). The flask was fitted with a
condenser flushed with nitrogen with a Teflon sleeve in the joint
and Teflon tape wrapped around the joint to secure a tight seal.35
The mixture was heated and stirred in a 190 °C oil bath for 24 h
with a gentle nitrogen flow across the top of the condenser. The
brown reaction mixture was cooled to below 100 °C (ca. 60 °C,
oil bath temperature), and then aqueous KOH (25 g, 445.6 mmol
in 100 mL of H2O) was slowly added. This two-phase mixture
was stirred in a 100 °C oil bath overnight (15 h), and the color
changed to orange. The mixture was cooled to rt, and ether
(100 mL) was added and stirred for 10 min before the organic
layer was isolated in a separatory funnel. The water layer was
extracted twice with ether (100 mL Â 2), and the combined
organic layer was washed with brine (100 mL) and dried over
MgSO4 and filtered.
761 w, 700 m cmÀ1; mass spectrum, m/z (% rel intensity) 373.8
(30), 372.5 M+ (100), 371.3 (15), 352.2 (13), 326.3 (8), 265.3
(10), 252.2 (8), 175.1 (10), 91.1 (5). HRMS calcd for C28H19O
m/z 371.1436, meas 371.1430.
Preparation of 2-Phenyl-4-phenanthrol 11 on a 14 g Scale
with Purification by Crystallization and Chromatography
(Table 4, entry 4). Acid 15 (14.23 g, 76.42 mmol from Sigma-
Aldrich Inc., Lot #: 03824CE) was used to prepare 16 with SOCl2
(20 mL, 274 mmol) in the same manner as described above. The
flask containing the resulted white solid was filled with argon and
charged with phenylacetylene (11 mL, 100.2 mmol) and (i-PrCO)2O
(25 mL, 146.2 mmol). The flask was fitted with a condenser
flushed with nitrogen with a Teflon sleeve in the joint and Teflon
tape wrapped around the joint to secure a tight seal.35 The mix-
ture was heated and stirred in a 190 °C oil bath for 48 h with a
gentle nitrogen flow across the top of the condenser. The orange
reaction mixture was cooled to below 100 °C (ca. 60 °C, oil bath
temperature), and then aq KOH (25 g, 445.6 mmol in 100 mL
H2O) was added. After stirring in a 100 °C oil bath overnight (15 h),
the orange solution was cooled to 25 °C, and ether (150 mL) was
added and stirred for 10 min before the organic layer was isolated
in a separatory funnel. The water layer was extracted twice with
ether (150 mL Â 2), and the combined organic layer was washed
with brine (100 mL) and dried over MgSO4 and filtered. The
solvent was completely removed in vacuo, and the brown solid
(22 g) was crystallized under N2 as follows: to a 250 mL round-
bottom flask was added a stir bar and 50 mL of CH2Cl2. The flask
was equipped with a condenser (flushed with N2 in advance) and
heated to reflux. Then CH2Cl2 (60 mL) was added slowly via
syringe until a clear dark brown solution formed, which was
cooled to 25 °C then to À20 °C overnight. The crystals of 11
(11.92 g, 44.15 mmol, 57.8%, mp 156À7 °C) were collected by
filtration and washed with 15 mL of CH2Cl2/hexanes (1:2).
The dark-colored mother liquor was combined with silica gel
(about 100 mL). After inserting a piece of cotton into the neck of
the trap of the rotary evaporator and removing the solvents, both
the flask and trap were put on high vacuum (0.5 mmHg) for
30 min. A chromatography column (6 cm diameter) was pre-
pared by filling the column with hexanes, and then silica gel was
added such that after settling a depth of 15À20 cm (about
350 mL silica gel) had been reached. The dried silica gel with the
preadsorbed product was added to the solvent above the pre-
pared bed and allowed to settle. The solvent level was lowered to
the top of the silica gel, and then a layer of sand was applied
immediately. The column was then eluted with a 1:2 mixture of
CH2Cl2:hexanes under gravity (about 1 h). A void volume of
about 700 mL was collected and discarded. This was followed by
a yellow fast-moving mixture of impurities (1.7À1.8 L). When
TLC indicated that the byproduct (26) began to elute, a second
fraction was collected that contained this byproduct (about
700À800 mL, containing 2.3 g), followed by an overlapping
third fraction containing a mixture of the two products
(300À400 mL, from which 1.1 g of solid could be isolated after
removal of solvents). This material contained 0.71 g of 11 based
on its 1H NMR spectrum and could be further purified by either
crystallization from CH2Cl2 or column (under the same con-
ditions). When 26 had finished eluting, elution was continued
under N2 pressure to collect a fourth fraction containing the pure
desired product (total of 3À3.5 L), which was stripped of solvent
by rotary evaporator and then dried in vacuo overnight to afford
3.45 g (12.7 mmol, 16.6%, mp 151À3 °C) of 11 as an orange
solid. The combined yield was 74.4%.
The dark-colored solution was combined with silica gel (about
100 mL). After inserting a piece of cotton into the neck of the
trap of the rotary evaporator and removing the solvents, both the
flask and trap were put on high vacuum (2 mmHg) for 30 min. A
chromatography column (6 cm diameter) was prepared by filling
the column with hexanes, and then silica gel was added such that
after settling a depth of 15À20 cm (about 350 mL silica gel) had
been reached. The dried silica gel with the preadsorbed product
was added to the solvent above the prepared bed and allowed to
settle. The solvent level was lowered to the top of the silica gel,
and then a layer of sand was applied immediately. The column
was then eluted with a 1:2 mixture of CH2Cl2:hexanes under
gravity (about 1 h). A void volume of about 700 mL was collected
and discarded. This was followed by a dark colored (dark brown
to black) fast-moving mixture of impurities, which smells strongly
like phenylacetylene (about 2 g from 1.5 to 2 L of solution after
the solvents were removed). When TLC indicated that the
byproduct (26) began to elute, a second fraction was collected
that contained 26 (about 700 mL, containing 2.24 g, 6.04 mmol,
8%, Rf = 0.50 using 1:3 EtOAc:hexane). When 26 had finished
eluting, elution was continued under N2 pressure to collect the
desired product (total of 4À5 L). The product fraction was
stripped of solvent by rotary evaporator and then dried in vacuo
overnight to afford 15.5 g (57.4 mmol, 75.1%, mp 152À153 °C,
lit. value:3 154À5 °C) of 11 as a yellow solid. Spectral data for 11:
1
Rf = 0.33 (1:5 EtOAc:hexane); H NMR (CDCl3, 500 MHz)
δ 5.71 (s, 1H), 7.18 (s, 1H), 7.35 (t, 1H, J = 7.5 Hz), 7.45 (t, 2H, J =
7.6 Hz), 7.55 (t, 1H, J = 7.2 Hz), 7.72 (t, 1H, J = 7.4 Hz),
7.64À7.70 (m, 5H), 7.84 (d, 1H, J = 7.8 Hz), 9.58 (d, 1H, J = 8.5
Hz); 13C NMR (CDCl3 75 MHz) δ 112.20, 118.52, 119.86,
125.99, 126.65, 127.21, 127.62, 128.25, 128.42, 128.89, 130.13,
132.57, 135.27, 139.17, 140.09, 154.61 (2 sp2 C’s not located).
Spectral data for 26: white crystals, mp 207À208 °C; Rf = 0.22
1
(1:1 CH2Cl2:hexane); H NMR (CDCl3, 500 MHz) δ 5.37
(s, 1H), 7.13À7.16 (m, 3H), 7.17À7.20 (m, 3H), 7.35 (dd, 2H, J =
7.5, 2.0 Hz), 7.41 (t, 1H, J = 8.0 Hz), 7.48 (s, 1H), 7.50 (s, 1H),
7.51À7.55 (m, 2H), 7.72À7.74 (m, 2H), 7.71 (d, 1H, J = 8.5
Hz), 7.71 (q, 2H), 7.86 (s, 1H); 13C NMR (CDCl3, 75 MHz)
δ 113.32, 121.72, 125.75, 126.74, 126.88, 127.37, 127.91, 128.07,
128.20, 129.08, 129.16, 129.21, 129.79, 129.95, 132.76, 132.82,
133.67, 140.70, 141.25, 142.29, 142.96, 153.70, two carbons not
located; 13C NMR (d6-acetone, 75 MHz) δ 113.50, 120.14,
126.03, 126.11, 126.76, 126.90, 127.20, 127.87, 127.99, 128.13,
128.22, 129.42, 130.12, 130.29, 130.42, 132.26, 133.36, 135.29,
140.84, 141.11, 142.17, 143.53, 156.30, one carbon not located;
IR (thin film) 3522 br vs, 1701 w, 1684 w, 1653 m, 1558 m, 1506 w,
1101
dx.doi.org/10.1021/op200088b |Org. Process Res. Dev. 2011, 15, 1089–1107

Organic Process Research & Development
ARTICLE
Isolation of the Ester Intermediates of 17 and 22 (No KOH
Workup). Acid 15 (14.23 g, 76.42 mmol) from Milestone Pharm
Tech USA Inc. (Lot #: T1161) was used to prepare acid chloride
16 with SOCl2 (20 mL, 274 mmol) in the same manner as
described above. The flask containing the resulting 16 (ca. 16.7 g)
was filled with argon and charged with phenylacetylene (11 mL,
100.2 mmol) and (i-PrCO)2O (25 mL, 146.2 mmol). The flask
was fitted with a condenser flushed with nitrogen with a Teflon
sleeve in the joint and Teflon tape wrapped around the joint to
secure a tight seal.35 The mixture was heated, and the contents
were stirred in a 190 °C oil bath for 48 h and then cooled to
25 °C. The volatiles wereremoved underhigh vacuum (1 mmHg),
and the residue (27 g) was dissolved in a minimum amount of
CH2Cl2 and loaded onto a silica gel column wet loaded with
hexanes. The column was eluted with a mixture of CH2Cl2 and
hexanes (1:2) to afford a fraction of the first component (22 was
the major component) and then a fraction of the second com-
ponent (17 was the major component). Both compounds were
crystallized from boiling CH2Cl2 saturated with hexanes by
cooling to 25 °CandthenÀ20 °Ctogetpure17 (2.34g, 5.34mmol,
7.0%) as white cotton-like crystals and 22 (19.3 g, 56.8 mmol,
74.3%) as yellow leaf-like crystals. The byproduct 26 was not
located. Spectral data for 17: mp 177À8 °C; Rf = 0.22 (1:1
CH2Cl2: hexane); 1H NMR (CDCl3, 500 MHz) δ 4.32 (s, 2H),
7.07 (t, 1H, J = 7.0 Hz), 7.40 (t, 1H, J = 7.0 Hz), 7.45À7.51 (m,
3H), 7.52À7.56 (m, 2H), 7.57 (d, 1H, J = 1.5 Hz), 7.66 (dd, 1H,
J = 8.5 Hz, J = 2.0 Hz), 7.72À7.78 (m, 4H), 7.83À7.84 (d, 1H, J =
7.5 Hz), 7.88À7.90 (m, 2H), 7.93 (d, 1H, J = 7.5 Hz), 8.00 (s,
1H), 8.01 (d, 1H, J = 2.0 Hz), 8.78 (d, 1H, J = 8.5 Hz); 13C NMR
(CDCl3, 75 MHz) δ 42.73, 121.07, 122.45, 125.17, 126.41,
126.68, 126.75, 126.82, 126.89, 127.43, 127.60, 127.84, 128.02,
128.07, 128.58, 128.93, 129.18, 130.55, 133.02, 133.16, 133.88,
135.28, 139.35, 139.77, 149.17, 170.05, four carbons not located;
13C NMR (d6-acetone/d6-DMSO 75 MHz) δ 41.86, 121.23,
122.41, 124.82, 126.37, 126.68, 126.74, 127.12, 127.37, 127.65,
128.03, 128.04, 128.38, 128.46, 128.58, 128.61, 128.74, 129.04,
129.16, 129.44, 131.66, 133.00, 133.22, 133.93, 135.37, 139.13,
139.27, 149.64, 170.40, one carbon not located; IR (salt plate)
1745 vs, 1622 w, 1454 m, 1385 m, 1235 m, 1116 m, 812 m,
753 m cmÀ1; mass spectrum, m/z (% rel intensity) 438.1 M+ (8),
271.2 (18), 270.0 (100), 238.9 (15), 168.0 (10), 140.9 (44),
114.9 (12); HRMS calcd for C32H23O2 m/z 439.1698, meas
439.1702. Spectral data for 22: yellow crystal; mp 121À2 °C; Rf =
0.50 (1:3 EtOAc: hexane); 1H NMR (CDCl3, 500 MHz) δ 1.54
(d, 6H, J = 7.0 Hz), 3.14À3.2 (m, 1H), 7.44 (t, 1H, J = 7.5 Hz),
7.54 (t, 2H, J = 7.5 Hz), 7.58 (d, 1H, J = 2.0 Hz), 7.62À7.68 (m,
2H), 7.77À7.82 (m, 4H), 7.92À7.94 (m, 1H), 8.04 (s, 1H), 9.18
(d, 1H, J = 8.0 Hz); 1H NMR (d6-acetone, 500 MHz) δ 1.48 (d,
6H, J = 7.0 Hz), 3.21 (octet, 1H, J = 7.0 Hz), 7.43 (t, 1H, J = 7.5
Hz), 7.53 (t, 2H, J = 7.5 Hz), 7.65 (td, 1H, J = 7.0 Hz, J = 1.0 Hz),
7.69À7.73 (m, 1H), 7.75 (d, 1H, J = 1.5 Hz), 7.81À7.90 (m,
4H), 7.97 (d, 1H, J = 8.0 Hz), 8.18 (d, 1H, J = 2.0 Hz), 9.22 (d,
1H, J = 8.5 Hz); 13C NMR (CDCl3, 75 MHz) δ 19.33, 35.13,
121.11, 122.67, 125.00, 126.84, 126.88, 127.58, 127.65, 128.06,
128.52, 129.09, 129.14, 129.19, 130.09, 133.27, 135.29, 139.46,
139.94, 149.42, 175.83; 13C NMR (d6-acetone, 75 MHz)
δ 19.33, 34.77, 121.09, 122.70, 124.66, 126.74, 127.10, 127.16,
127.26, 127.37, 127.63, 128.07, 128.19, 128.47, 129.06, 129.16,
129.30, 133.32, 135.38, 139.30, 139.50, 149.82, 175.45; IR (salt
plate) 2973 w, 1754 s, 1453 m, 1386 m, 1176 m, 1107 vs, 882 m,
749 m, 695 m cmÀ1; mass spectrum, m/z (% rel intensity) 341.3
(6), 340.1 M+ (30), 270.3 (35), 270.1 (100), 239.0 (60), 165
(10), 138.8 (10), 119.6 (17), 70.9 (22), 43.0 (58). Anal. Calcd for
C24H20O2: C, 84.68; H, 5.92. Found: C, 84.34; H, 5.86.
Preparation of 2-Phenyl-4-phenanthrol 11 on a 57 g Scale
with Purification by Crystallization (Table 4, entry 5). A
single-neck 1 L round-bottom flask equipped with a large 48 Â
18 mm oval magnetic stir bar and a condenser was charged with
57 g (306 mmol, Milestone Pharm Tech USA Inc., Lot #: T1161)
of acid 15 and SOCl2 (45 mL, 618 mmol). The top of the
condenser is vented to a bubbler and then into a beaker filled with
aqueous NaOH to trap acidic gases (HCl and SO2). The mixture
was heated to reflux for 1 h in a 90 °C oil bath, and then all of the
volatiles were distilled off. It was then put on high vacuum (2 mmHg)
and swirled until the residue solidified with a second liquid N2
trap to protect the pump. The extra liquid N2 trap was then
removed, and the residue was kept under vacuum for 1 h. The
flask containing the yellow crude acid chloride 16 was filled with
argon, and then phenylacetylene (45 mL, 410 mmol) and
(i-PrCO)2O (100 mL, 603 mmol) were added. The flask was fitted
with a condenser flushed with nitrogen with a Teflon sleeve in
the joint and Teflon tape wrapped around the joint to secure a
tight seal.35 The reaction mixture was heated and stirred in a
190 °C oil bath for 48 h with a gentle nitrogen flow across the top
of the condenser. The brown reaction mixture was cooled to
about 60 °C (oil bath temperature), and aq KOH (100 g, 1.8 mol
in 400 mL of H2O) was slowly added. After stirring in a 100 °C
oil bath overnight (15 h), the orange solution was cooled to rt,
ether (400 mL) added, and the mixture stirred for 30 min before
the organic layer was isolated in a 2 L separatory funnel. The
water layer was extracted twice with ether (400 mL Â 2), and the
combined organic layer was washed with brine (400 mL), dried
over MgSO4, and filtered. The dark-colored organic solutions
1
were combined together (5 drops were collected for H NMR
analysis), and the solvents were removed in vacuo. The residue
was collected in a 500 mL flask and dried under high vacuum
(1 mmHg) overnight to give ca. 80 g of the dark brown crude
product. The 1H NMR spectrum of the crude product indicated
it was a mixture of 11 and 26 with a 1:0.15 ratio.
To the crude mixture in the 500 mL flask was added a stir bar
and 100 mL of CH2Cl2, and the resulting solution was brought to
a boil under an atmosphere of N2. More CH2Cl2 was then added
in 50 mL aliquots until all was dissolved (total of 200 mL). The
solution was allowed to cool to rt and then to À20 °C in a
refrigerator overnight. The mixture was filtered and the solid
washed with a mixture of CH2Cl2:hexanes (1:2, 15 mL Â 2) to
give 34.4 g (127.2 mmol, 41.5%) of 11 as beige crystals (mp
156À7 °C, lit.3 154À5 °C). The dark-colored mother liquor was
collected in a 500 mL flask and combined with silica gel
(ca. 150 mL). After inserting a piece of cotton into the neck of the
trap of the rotary evaporator and removing the solvents, both
the flask and trap were put on high vacuum (0.5 mmHg) for 1 h. The
mixture was filtered through a short column of silica gel as
follows. A short pad of Celite (30 g) was prepared in a sintered
glass funnel (OD 10 cm, 18 cm long), and then silica gel
(400 mL) was added followed by the crude mixture absorbed
on silica gel and finally a thin layer of sand (2À3 cm). The
mixture was then eluted with hexanes:CH2Cl2 = 1:1 with a
vacuum produced by a water aspirator. The first fraction of ca. 1 L
was discarded. This fraction by TLC contains fast running
impurities (ca. 10 g, black material) and a small amount of
double-inserted byproduct 26. An additional 2.5 L of hexanes:
CH2Cl2 = 1:1 was passed through and collected, and this
product-containing fraction was concentrated on a rotary
1102
dx.doi.org/10.1021/op200088b |Org. Process Res. Dev. 2011, 15, 1089–1107

Organic Process Research & Development
ARTICLE
evaporator to afford a mixture of 11 and 26 (∼25 g) as an
orange solid.
Collection of a third crop gave material that was not suffi-
ciently pure by 1H NMR. Therefore, the third crop and mother
liquor residue were combined (∼35 g) and purified via column
chromatography on silica gel. This mixture was dissolved in CH2Cl2
and added to 40 mL of silica gel. After removal of volatiles, the
silica gel mixture was loaded onto a silica gel column (5 Â 25 cm)
that was wet loaded with hexanes. Elution with hexanes/CH2Cl2
(2:1) and combining the fractions containing the product gave
18 g of an off-white solid that was shown to contain small
amounts of impurities by 1H NMR. This material was crystallized
from 80 mL of a 3:1 mixture of hexanes and CH2Cl2 to give the
pure product 8 in 11% yield (9.41 g, 43 mmol) with a mp of
97.5À98 °C. Spectral data for 8: white solid; mp 98.5À99.5 °C
This solid was refluxed with 10 mL of iPrOH under N2 to
effect dissolution, cooled to 25 °C, and then to À20 °C over-
night. The solid was filtered and washed with iPrOH (3 Â
10 mL) to give pure 26 (5.93 g, 15.9 mmol) as a white solid. The
mother liquor was stripped of solvent in vacuo, and to the orange
residue (22 g, mp 146À7 °C) was added 100 mL of a 2:1 mixture
of hexanes and 1,2-dichloroethane. The mixture was brought to a
boil under an atmosphere of nitrogen. More of this solvent
mixture was added (in 50 mL aliquots, total of about 350 mL)
until all was dissolved. The mixture was cooled to 25 °C and then
to À20 °C to give a second crop of 16b (12.4 g, 45.9 mmol, 15%,
mp 157À8 °C). The total yield was 56.5%.
1
(lit.3 96À97.5 °C); Rf = 0.48 (1:3 EtOAc/hexane). H NMR
The SOCl2 (50 mL) could be recovered from this process.
Acid 15 could be recovered from aqueous KOH as follows: after
acidification with HCl (6 N) to pH∼1 and extraction with ether
(3 Â 100 mL), the combined organic layer was washed with brine
(100 mL), dried over MgSO4, and filtered. Ether was removed via
a rotary evaporator, and the remaining orange oil was a mixture
of 15 in ca. 90 mL of isobutyric acid. The isobutyric acid was
distilled off under a vacuum produced by a water aspirator
(80 mmHg/81À2 °C) to give 7.20 g of crude 15 as an orange
solid. This material was crystallized from EtOH/H2O to give
5.15 g (27.5 mmol, 9%) of the acid starting material 15 as yellow
leaf-like crystals.
(CDCl3, 300 MHz) δ 5.32 (s, 1H), 7.06 (s, 1H), 7.34 (t, 1H, J = 9
Hz), 7.41À7.50 (m, 4H), 7.62À7.64 (m, 3H), 7.82 (d, 1H, J = 10
Hz), 8.13 (d, 1H, J = 9 Hz); 13C NMR (CDCl3, 75 MHz) δ
108.41, 118.73, 121.39, 123.47, 125.34, 126.86, 127.20, 127.37,
127.99, 128.75, 134.85, 138.73, 140.67, 151.47; mass spectrum,
m/z (% rel intensity) 220 M+ (100), 191.0 (45), 189.0 (30),
165.0 (23), 95 (23), 55 (21), 43 (25).
Preparation of Racemic VANOL 2 on a 50 g Scale with
Purification by Crystallization (Table 6, entry 11). An oil bath
(19 cm id containing 1.5 L oil) was heated to 165 °C and stirred
with a magnetic stirrer. The naphthol 8 (50.05 g, 227.5 mmol)
was introduced by a funnel into a 2 L 3-necked round-bottom
flask equipped with a 60 Â 18 mm oval magnetic stir bar and a
400 mm Allihn water-cooled condenser. This was followed by
the addition of 250 mL of light mineral oil through the same
funnel. A glass tube (6 mm id) was introduced into the flask via
the second neck to about 5 cm above the surface of the aphthol
solution and was used to provide a stream of house air which is
maintained at a flow rate of 0.15À0.20 L/min.36 The third neck
was sealed with a rubber septum. The stir bar in the oil bath was
removed before the flask was introduced into the oil bath to
warm it up for about 15 min until the solid was melted. Airflow
was allowed to flow into the flask while the molten 18 was stirred
as fast as possible. The airflow was switched to N2 after the
reaction was kept at 165 °C for 24 h. The flask was removed from
the oil bath and cooled to ambient temperature before hexanes
(500 mL) were added to the flask. The mixture was stirred for
30 min, and then it was cooled to À20 °C overnight (12 h) before
the solid was collected by suction filtration. The crude product
(about 46 g, brown powder, softened at 216À219 °C and melted
at 224À227 °C) was dried on high vacuum and crystallized from
600 mL of hot CH2Cl2. The dark-colored solution was cooled to
room temperature and then to À20 °C overnight (12 h). The
brown crystals were collected via suction filtration, washed with
hexanes (3 Â 50 mL), and dried under vacuum to give the first
crop product of (()-VANOL (37.4 g, 85.3 mmol, 74.7%, mp
231À232 °C, lit.3 231À233 °C). The mother liquor was dried, and
the residue was crystallized from 60 mL of hot CH2Cl2 and cooled
to À20 °C overnight to give 5.04 g of additional (()-VANOL
as brown crystals (11.5 mmol, 10.1%, mp 230À231 °C). The
combined yield was 42.4 g (96.8 mmol, 85.2%).
Preparation of 3-Phenyl-1-naphthol 8 on a 59 g Scale with
Purification by Crystallization (Table 5, entry 2). To a flame-
dried single neck 2 L flask equipped with a magnetic stir bar was
added 2-phenylacetyl chloride (50.5 mL, 59.0 g, 374 mmol),
phenyl acetylene (55 mL, 501 mmol), and isobutyric anhydride
(125 mL, 731 mmol). The flask was fitted with a condenser
flushed with nitrogen with a Teflon sleeve in the joint and
Teflon tape wrapped around the joint to secure a tight seal.35
The mixture was stirred at 190 °C for 48 h with a gentle nitrogen
flow over the top of the condenser. The reaction was cooled to
room temperature, and aq KOH (125 g, 2.23 mol in 500 mL of
H2O) was added. The reaction mixture was stirred at 100 °C
overnight (13À15 h). The solution was cooled to 0 °C and
acidified with 6 N HCl to pH ∼ 6 (100À110 mL). The mixture
was then transferred to a separatory funnel using 400 mL of
ether. The organic layer was separated, and the aqueous layer
was washed with ether (3 Â 100 mL). The combined organic
layers were washed with sat Na2CO3 (3 Â 100 mL) and brine
(100 mL) and dried over MgSO4. After filtration through Celite,
the solvent was removed by a rotary evaporator to give a dark
brown oil. Hexanes (3 Â 50 mL) were added and then removed
by a rotary evaporator to give a dark brown solid (92 g) with a
mp of 84À89 °C (begins to soften at 71À73 °C). The crude
product was taken up in 900 mL of refluxing hexanes/CH2Cl2
(4:1), and the hot solution was poured into a 1 L Erlenmeyer
flask leaving some white solid behind (1À2 g) which was taken
up in a small amount of hot dichloromethane. Both solutions
were covered and allowed to cool to room temperature over-
night, and then the solids from each were collected together in a
5 in. Buchner funnel and rinsed with cold hexanes (0 °C, 2 Â
200 mL) to give the first crop of 8 as a white fluffy solid in 43%
yield (34.9 g, 159 mmol) with a mp of 98.5À99.5 °C. The
mother liquor was concentrated to dryness, and the product was
crystallized again using hexanes/CH2Cl2 (3:1, 400 mL) to give a
second crop of 8 as a white fluffy solid in 13% yield (10.37 g,
47 mmol) with a mp of 98.5À99.5 °C.
Preparation of Racemic VAPOL 3 on a 50 g Scale with
Purification by Crystallization (Table 7, entry 4). An oil bath
(19 cm id containing about 1.5 L of oil) was heated to 170 °C
while the oil was stirred with a magnetic stirrer. Phenanthrol 11
(50.0 g, 185.2 mmol) was introduced by a funnel into a 2 L
3-necked round-bottom flask equipped with a 60 Â 18 mm oval
magnetic stir bar and a 400 mm Allihn water-cooled condenser.
1103
dx.doi.org/10.1021/op200088b |Org. Process Res. Dev. 2011, 15, 1089–1107

Organic Process Research & Development
ARTICLE
This was followed by the addition of 250 mL of light mineral oil
through the same funnel. A glass tube (6 mm id) was introduced
into the flask via the second neck to about 5 cm above the surface
of the solution of 11 and was used to provide a stream of house
air, which was maintained at a flow rate of 0.15À0.20 L/min.36
The third neck was sealed with a rubber septum. The stir bar in
the oil bath was removed before the flask was introduced to warm
the mixture for about 30 min until the solid melted. Airflow was
allowed to flow into the flask while the solution of 11 was stirred
as fast as possible. The airflow was switched to N2 gas after the
reaction was kept at 170 °C for 36 h. The flask was removed from
the oil bath and cooled to ambient temperature before hexane
(500 mL) was added to the flask. The mixture was stirred for
30 min, and then it was cooled to À20 °C overnight (12 h) before
the solid was collected by suction filtration. The crude product
(45.1 g, brown powder, softened at 286À287 °C and melted at
293À295 °C) was dried under high vacuum and crystallized from
a minimum amount of boiling toluene (about 1 L). The dark-
colored solution was cooled to room temperature and then to
À20 °C overnight. The liquid portion of the solution was filtered
without disturbing the precipitate to collect the fine suspension
of particles in solution, and then a new piece of filter paper was
used to collect the brown crystals via suction filtration. This solid
was washed with hexanes (2 Â 50 mL) and dried over vacuum to
give the first crop of (()-VAPOL (33.3 g, 61.9 mmol, 66.8%, mp
310À311 °C, lit.30 312À313.2 °C). The mother liquor was dried,
and the residue was crystallized from 150 mL of hot toluene and
then cooling to room temperature for 12 h followed by cooling to
À20 °C overnight to give an additional 7.0 g of (()-VAPOL
(13.0 mmol, 14.0%, mp 298À299 °C). The combined yield was
40.3 g (74.9 mmol, 80.9%).
Preparation of Racemic VANOL Hydrogen Phosphate 29.
To a single-necked 500 mL round-bottom flask flushed with N2
was added 40.0 g (91.3 mmol) of racemic VANOL via a powder
funnel. As the contents of the flask were stirred with a 48 Â
18 mm oval magnetic stir bar, 150 mL of pyridine was added and
used to rinse the funnel. The flask was fitted with a rubber septum
and a nitrogen balloon. To the transparent brown solution was
added phosphorus oxychloride (17.0 mL, 182.4 mmol) dropwise
over a period of 10 min via a plastic syringe. Upon addition of
POCl3, the flask becomes hot, and a beige precipitate forms but
does not stop the stirrer. The resulting suspension was stirred for
6 h at room temperature. Water (120 mL) was added slowly in 3
to 4 portions, and the addition of each subsequent portion was
delayed until boiling had subsided. The resulting biphasic sus-
pension was stirred at room temperature for 2 h. The pyridine
was removed by rotary evaporator, and the residue was dissolved
in 250 mL of CH2Cl2 to give a brown solution which was washed
twice with 500 mL of 1 N HCl. The solution was dried over
MgSO4 and filtered through Celite. The solvent was removed to
give the crude product, which was dried overnight under high
vacuum. This left the crude racemic VANOL hydrogen phos-
phate as a yellow amorphous solid which was used directly in the
resolution without further purification. The yield: 44.0 g (87.9
mmol, 96.4%). Spectral data for 29: white solid, mp 245À250 °C
dec; 1H NMR (CDCl3 300 MHz) δ 5.80 (br s, 1H), 6.45 (d, 4H,
J = 7.5 Hz), 6.88 (t, 4H, J = 7.4 Hz), 7.06 (t, 2H, J = 7.2 Hz), 7.47
(s, 2H), 7.49À7.51 (m, 4H), 7.77 (d, 2H, J = 7.4 Hz), 8.46 (d,
2H, J = 7.3 Hz).
equipped with a condenser and a stir bar (48 Â 18 mm) and
which had been flushed with a N2 stream that was introduced via
a needle in a septum on one of the necks and monitored by a
bubbler attached to the condenser. To the flask was added
dichloroethane (200 mL), and the contents of the flask were
stirred and brought to a boil by heating the flask with a heating
mantel. To a separate 250 mL pear-shaped flask was added
brucine dihydrate (38.0 g, 88.6 mmol, technical grade, 92.6%)
and 90 mL of dichloroethane. The contents of the flask were
purged with N2 for about 2 min, and then the flask was sealed
with a rubber septum which was fitted with a nitrogen balloon.
This flask was heated by swirling in hot flowing water (50 °C)
until a clear colorless solution (volume about 110 mL) resulted
which had a small amount of grey insoluble impurities floating on
top. The resulting brucine solution was placed in a hot water bath
(50 °C), and 60 mL of the solution was removed by a syringe
equipped with a 12 gauge needle and added to the solution of 29
over a period of about 1 min to give a brownish clear solution.
The remainder of the brucine solution (about 50 mL) was loaded
into the syringe leaving the insoluble material in the flask. The
brucine solution was slowly added to the solution of 29 via
syringe pump (addition rate: 100 mL/h). Solid began to form
when about 22 mL of the solution was left in the syringe. The
addition was paused while the suspension was refluxed for 15 min,
and then the addition was resumed to complete the addition to
the suspension. An additional portion of dichloroethane (10 mL)
was used to rinse the pear flask leaving the insoluble material in
the flask. The final washing was added over a period of one
minute. The resulting suspension was cooled slowly to room
temperature and allowed to settle for 48 h without disturbance.
The white solid was collected by filtration through a Buchner
funnel and washed three times with dichloroethane (25 mL) and
then dried over high vacuum for 12 h to afford 36.3 g (40.5 mmol,
87%, >99% de) of the Salt II 32.
The dark yellow mother liquor was stripped of volatiles by
rotary evaporator and combined with 100 mL of dichloroethane
in a 2 L single-necked round-bottom flask fitted with a condenser.
The resulting mixture was stirred and brought to a boil to dissolve
the solid residue. A total of 1000 mL of ethanol was added in
portions as follows: first, 300 mL of ethanol was added, and the
resulting solution was returned to a boil and then additional
ethanol (200 mL) was added and again returned to a boil. At this
point, another 200 mL portion of ethanol was added, and the
mixture was refluxed for 15 min until it appeared that no further
accumulation of precipitate was occurring. Finally, an additional
300 mL portion of ethanol was then added, and the mixture was
refluxed for 10 min before being allowed to cool to room
temperature overnight undisturbed. The beige solid was isolated
by filtration in a Buchner funnel and dried over high vacuum for
12 h to afford 34.8 g (38.9 mmol, 83%, >99% de) of Salt I 31. The
filtrate was stripped of volatiles to give 10.50 g of a light brown
solid (Salt I, 32% de).
Liberation of (R)-VANOL by Reductive Cleavage of the
Brucine Salt II. To an oven-dried 1 L round-bottom flask was
added Salt II 32 (52.7 g, 58.8 mmol), 250 mL of reagent-grade
toluene, and a 48 Â 18 mm oval magnetic stir bar. The flask was
equipped with a 1 L pressure compensating addition funnel
which was sealed with a rubber septum. The system was flushed
with N2 for at least 30 min, and then the mixture was cooled to
0 °C. A nitrogen balloon was used to balance the pressure. Red-Al
(75 mL, 65 wt % in toluene, 246 mmol) was added to the funnel
and then slowly added to the flask over 3 h with stirring. After
Resolution of Racemic VANOL Hydrogen Phosphate 29
with (À)-Brucine. The hydrogen phosphate 29 (45.2 g, 90.3 mmol)
was placed in a 500 mL 3-necked round-bottom flask that was
1104
dx.doi.org/10.1021/op200088b |Org. Process Res. Dev. 2011, 15, 1089–1107

Organic Process Research & Development
ARTICLE
addition, the flask was warmed to room temperature for 4.5 h.
The flask was then cooled to 0 °C, and chilled (0 °C) HCl (6 N,
500 mL) was added slowly to quench the reaction as follows: the
first 20 mL of HCl was added a pipet full at a time down inside the
wall of the flask, and the remainder was added in 3À5 portions
over 10 min. The mixture was put into a 2 L separatory funnel,
and the organic layer was collected. The water layer was extracted
three times with 500 mL of ethyl acetate. The combined organic
layer was washed with 400 mL of brine, dried over MgSO4, and
filtered through Celite. Upon removal of the solvent the residue
(ca. 25 g) was dissolved in a minimum amount of CH2Cl2 (ca.
50 mL) and loaded onto a silica gel column (4.5 cm OD, silica gel
was filled to a depth of 40 cm) and eluted by hexanes:CH2Cl2
(1:2). When TLC indicated that VANOL had begun to elute, a
2 L round bottomed flask was used to collect all of the VANOL.
When the elution was complete (ca. 2 L), 5 drops of this solution
was saved for HPLC analysis. The rest of solution was stripped of
solvents to give a foamlike solid. This solid was dissolved in a
minimum (ca. 50 mL) of CH2Cl2, and then 300 mL of hexanes
was added. The solution was slowly evaporated by a rotary evap-
orator to dryness. The residue was again taken up in a minimum
of (ca. 50 mL) of CH2Cl2, and then 300 mL of hexanes was
added. The resulting mixture was shaken vigorously under a
strong N2 flow until a solid crashed out. This mixture was then
slowly evaporated to dryness. This process gave (R)-VANOL as a
white powder-like solid that was more stable to long-term storage
if this procedure were not employed. Yield: 22.7 g (51.9 mmol,
88.1%); mp 199À201 °C, [R]D 316 (CHCl3, c 1.0) on 99% ee
material. Spectral data for (R)-VANOL: 1H NMR (CDCl3, 500
MHz) δ 5.82 (s, 2H), 6.63 (d, 4H, J = 7.5 Hz), 6.94 (t, 4H, J =
8.0 Hz), 7.07 (t, 2H, J = 7.5 Hz), 7.32 (s, 2H), 7.54À7.56 (m,
4H), 7.78 (d, 2H, J = 7.0 Hz), 8.35 (d, 2H, J = 8.0 Hz); 13C NMR
(CDCl3, 125 MHz) δ 112.72, 122.03, 122.79, 122.92, 125.66,
126.61, 127.44, 127.51, 127.68, 128.89, 134.62, 140.19, 140.70,
150.37. The optical purity of (R)-VANOL was determined as
follow: 5 drops of the saved (R)-VANOL solution was diluted
with 2 mL of hexanes, from which ca. 4 μL was injected into an
HPLC with a Pirkle D-phenylglycine column and eluted with a
98:2 mixture of hexane/isopropanol (254 nm, flow rate: 2 mL/min
over 30 min). Under these conditions, the retention time of
(R)-VANOL was 18.3 min, and that of (S)-VANOL was 20.6 min.
The ee of (R)-VANOL was >99%. The same procedure can be
applied to obtain (S)-VANOL from the reductive cleavage of Salt
I 31 in 81% yield and >99% ee, mp 200À201 °C, [R]D À319
(CHCl3, c 1.0).
(d, 2H, J = 9 Hz), 7.87 (d, 2H J = 7.5 Hz), 9.87 (d, 2H, J = 8.5 Hz).
White solid, mp > 350 °C.
Resolution of Racemic VAPOL Hydrogen Phosphate 33
with (À)-Cinchonidine. To a boiling suspension of 33 (40.0 g,
66.7 mmol) in 2 L of absolute ethanol in a 4 L Erlenmeyer flask
was added (À)-cinchonidine (20.0 g, 68.0 mmol), while the
solution was stirred on a hot plate (Corning) with a 60 Â 18 mm
oval magnetic stir bar. Ethanol (500 mL) was used to rinse all
(À)-cinchonidine into the suspension, which was gently boiled
for 30 min before it was cooled to room temperature overnight.
The solution was filtered without disturbing the solid until most
of the liquid went through the Buchner funnel. Then the filter
paper was replaced by a new one, and the remaining solid was
broken, filtered, and washed with ethanol (50 mL). This solid
(17.6 g, 19.7 mmol, 58%) was revealed by its 1H NMR spectrum
to be Salt I 35. The mother liquor was dried by a rotary
evaporator and combined with 2 L of 2-propanol, which was
heated to a boil in a 4 L Erlenmeyer flask. Acetone was added
slowly with stirring until the solid dissolved (about 500 mL). This
solution was cooled to room temperature overnight and then to
À20 °C. Crystals (26.1 g, 28.7 mmol, 87%) were collected and
1H NMR revealed that were pure Salt II 36. The filtrate was dried
and dissolved in a minimum amount of boiling CH2Cl2 (about
50 mL), to which ethanol (100 mL) was added in two portions
and heated for 15 min until a solid crashed out. More ethanol was
then introduced (400 mL in two portions) and heated to a boil
and then allowed to cool to room temperature overnight. The
solid was filtered and washed twice with 25 mL of ethanol and
was identified as pure Salt I 35 by 1H NMR (9.24 g, 10.3 mmol,
31%). The mother liquor was dried by rotary evaporator and
combined with 500 mL of 2-propanol, which was heated to a boil,
and acetone was slowly added while stirring until the solid
dissolved (about 50 mL). This solution was cooled to room
temperature overnight and then to À20 °C. Crystals (3.58 g,
4.0 mmol, 12%) were collected, and 1H NMR showed them to be
pure Salt II 36. The filtrate (a 3:1 mixture of Salt I and Salt II by
1H NMR spectrum, total of 8.6 g) was dried and dissolved in a
minimum amount of boiling CH2Cl2 (15 mL), to which ethanol
was added in three portions (25 + 25 + 50 mL), and after each
addition the solution was returned to a boil for 10 min. A solid
crashed out, and more ethanol (150 mL, which made the total
volume of ethanol at 250 mL) was introduced and heated to a
boil and then cooled to room temperature overnight. The solid
was filtered and washed with 25 mL of ethanol, which was
confirmed to be pure Salt I 35 by 1H NMR (2.96 g, 3.31 mmol,
10%). The total amount of Salt I and Salt II: 29.8 g (99%) and
29.7 g (99%), respectively. Spectral data for Salt I 35: white solid;
1H NMR (CDCl3 500 MHz) δ 0.75À0.85 (m, 1H), 1.50À1.80
(m, 6H), 2.28 (s, 1H), 2.38 (s, 1H), 2.62À2.81 (m, 2H), 4.10
(s, br, 1H), 4.70 (d, 1H, J = 17.1 Hz), 4.83 (d, 1H, J = 10 Hz),
5.08À5.12 (m, 1H), 5.92 (s, 1H), 6.49 (d, 4H, J = 7.8 Hz), 6.60
(br s, 2H), 6.77 (t, 1H, J = 7.7 Hz), 6.85 (t, 4H, J = 7.6 Hz), 7.01
(t, 2H, J = 7.3 Hz), 7.15 (m, 3H), 7.19À7.27 (m, 4 H), 7.43
(s, 2H), 7.55À7.61 (m, 6H), 7.75 (d, 1H, J = 8.4 Hz), 8.67 (d, 1H,
J = 4.4 Hz), 10.08 (d, 2H, J = 8.6 Hz), 11.79 (s, br, 1H). Spectral
Preparation of Racemic VAPOL Hydrogen Phosphate 33.
To a 500 mL round-bottom flask charged with N2 was added
30.91 g (57.4 mmol) of racemic VAPOL and pyridine (170 mL).
As this mixture was stirred, phosphorous oxychloride (10.7 mL,
114.8 mmol) was added dropwise over a period of 10 min. After
all the solid had dissolved, the resulting solution was stirred for
6 h at room temperature (a lot of solid formed after about
45 min). Water (75 mL) was added slowly, and the resulting
biphasic suspension was stirred at room temperature for 2 h.
Pyridine was removed by rotary evaporator, and the residue was
combined with 250 mL of 1 N HCl and then was filtered and
washed twice with 250 mL of 1 N HCL. The crude product was
dried over high vacuum overnight and used directly for resolu-
tion without further purification. The yield: 34.35 g (54.4 mmol,
100%). The spectrum of 33: 1H NMR (CDCl3) δ 6.51 (d, 4H, J =
7.2 Hz), 6.93 (t, 4H, J = 7.3 Hz), 7.09 (t, 2H, J = 7.5 Hz), 7.23
(s, 4H), 7.52À7.55 (m, 4H), 7.66 (d, 2H, J = 8.8 Hz), 7.76
1
data for Salt II 36: light yellow crystal; H NMR (CDCl3 500
MHz) δ 0.97À1.05 (m, 1H), 1.20À1.29 (m, 1H), 1.50À1.58 (m,
1H), 1.62À1.73 (m, 2H), 1.98À2.07 (m, 1H), 2.18À2.19
(m, 1H), 2.40À2.50 (m, 3H), 3.10À3.15 (m, 1H), 4.05À4.15 (m,
1H), 4.75 (d, 1H, J = 17.2 Hz), 4.81 (d, 1H, J = 10.4 Hz),
5.22À5.32 (m, 1H), 6.39 (s, 1H), 6.55 (d, 4H, J = 7.4 Hz), 6.61
(br s, 1H), 6.89 (t, 4H, J = 7.6 Hz), 6.98 (t, 2H, J = 7.8 Hz), 7.04
1105
dx.doi.org/10.1021/op200088b |Org. Process Res. Dev. 2011, 15, 1089–1107

Organic Process Research & Development
ARTICLE
(t, 2H, J = 7.3 Hz), 7.36À7.31 (m, 4H), 7.50 (s, 2H), 7.59À7.68
(m, 6H), 7.75 (d, 2H, J = 7.8 Hz), 8.16 (d, 1H, J = 7.7 Hz), 8.40
(d, 1H, J = 7.9 Hz), 8.69 (d, 1H, J = 5.0 Hz), 9.94 (d, 2H, J =
8.6 Hz), 11.85 (s, 1H).
127.54, 128.44, 128.85, 128.86, 129.30, 130.33, 132.86, 135.32,
139.77, 141.59, 153.44.
(R)-VAPOL can be obtained from the reductive cleavage of
salt II 36 with Red-Al utilizing the same procedure described
above in 90% yield and >99% ee. After removal of the CH2Cl2 as
described above, (R)-VAPOL was obtained as a white powder
with mp 223À227 °C (lit.30 226.2À227.9 °C) and [R]D À148.3
(CHCl3 c 1.0).
Liberation of (S)-VAPOL by Reductive Cleavage of the
Cinchonidine Salt I 35. A 500 mL 3-necked round-bottom flask
equipped with a stir bar was sealed with one rubber septum and
two adapters. One adapter was connected to a bubbler, and the
flask was flushed with N2 for 1 h via the other adapter. Salt I 35
(36.3 g, 40.6 mmol) was put into the flask, and the adapters were
switched to two rubber septums. A N2 balloon was connected to
the flask via a needle. Toluene (200 mL) was added followed by a
48 Â 18 mm oval magnetic stir bar. The mixture was cooled in an
iceÀwater bath for 1 h and then stirred while Red-Al (54 mL,
177.0 mmol as a 65 wt % solution in toluene) was added via
syringe pump over 3 h using a 12-gauge needle. The suspension
was allowed to stir at room temperature overnight and then was
chilled in an iceÀwater bath for 1 h. Precooled HCl (6 N, 150 mL,
0 °C) was added, and the mixture was stirred at room tempera-
ture for 5 min before it was poured into a 1 L separatory funnel.
The organic layer was separated, and the aqueous layer was ex-
tracted three times with 100 mL of ethyl acetate. The combined
organic layer was washed with brine (100 mL) and then dried
over MgSO4. After filtration through Celite, the solvent was removed
in vacuo. The residue (ca. 20 g) was dissolved in a minimum
amount of CH2Cl2 (about 110 mL, a heat gun may help dis-
solution) and was loaded onto a chromatography column (6 cm
diameter) prepared by filling the column with a 1:20 mixture of
CH2Cl2 and hexanes and then the addition of silica gel such that
after settling a depth of ca. 15 cm had been reached. When the
solvent level was lowered to the top of the silica gel, the column
was eluted with a 1:1 mixture of hexanes and CH2Cl2 and tracked
with TLC until the product had eluted. All of the fractions
containing the product were combined together (about 1.5 L),
from which 5 drops was collected by pipet which was later used
for optical purity determination. The solvent was slowly removed
by a rotary evaporator to give shiny yellow crystals. To remove
the color, the crystals were combined with 30 mL of CH2Cl2 and
swirled for about 1 min, and then 100 mL of hexane was added;
this mixture was filtered and rinsed with hexanes (2 Â 25 mL).
The solid (19.8 g, 36.8 mmol, 90.6% as light yellow crystals) was
collected and dried overnight on high vacuum. The optical purity
of (S)-VAPOL (S-3) was determined to be >99% by HPLC using
a Pirkle D-phenylglycine column as described below: the 5 drops
of solution saved from above were diluted with 5 mL of hexanes,
from which 4 μL was injected onto the HPLC with the following
conditions: 260 nm, 2 mL/min, 25:75 mixture of 2-propanol and
hexanes, and the tR = 14.0 min for (S)-VAPOL and tR = 23.9 min
for (R)-VAPOL.
’ AUTHOR INFORMATION
Corresponding Author
*E-mail: wulff@chemistry.msu.edu.
’ ACKNOWLEDGMENT
This work was supported by a grant from the National Science
Foundation.
’ REFERENCES
(1) Brunel, J. M. Chem. Rev. 2005, 105, 857–897.
(2) Chen, Y.; Yekta, S.; Yudin, A. K. Chem. Rev. 2003, 103, 3155–3211.
(3) Bao, J.; Wulff, W. D.; Dominy, J. B.; Fumo, M. J.; Grant, E. B.;
Rob, A. C.; Whitcomb, M. C.; Yeung, S.-M.; Ostrander, R. L.; Rheingold,
A. L. J. Am. Chem. Soc. 1996, 118, 3392–3405.
(4) (a) Bao, J.; Wulff, W. D.; Rheingold, A. L. J. Am. Chem. Soc. 1993,
115, 3814–3815. (b) Bao, J.; Wulff, W. D. Tetrahedron Lett. 1995,
36, 3321–3324. (c) Heller, D. P.; Goldberg, D. R.; Wulff, W. D. J. Am.
Chem. Soc. 1997, 119, 10551–10552. (d) Heller, D. P.; Goldberg, D. R.;
Wu, H.; Wulff, W. D. Can. J. Chem. 2006, 84, 1487–1503.
(5) Xue, S.; Yu, S.; Deng, Y.; Wulff, W. D. Angew. Chem., Int. Ed.
2001, 40, 2271–2774.
(6) Bolm, C.; Frison, J.-C.; Zhang, Y.; Wulff, W. D. Synlett 2004,
1619–1621.
(7) Rowland, G. B.; Zhang, H.; Rowland, E. B.; Chennamadhavuni,
S.; Wang, Y.; Antilla, J. C. J. Am. Chem. Soc. 2005, 127, 15696–15607.
(8) Liang, Y.; Rowland, E. B.; Rowland, G. R.; Perman, J. A.; Antilla,
J. C. Chem. Commun. 2007, 4477–4479.
(9) Li, G.; Liang, Y.; Antilla, J. C. J. Am. Chem. Soc. 2007, 129, 5830–5831.
(10) (a) Rowland, E. B.; Rowland, G. B.; Rivera-Otero, E.; Antilla,
J. C. J. Am. Chem. Soc. 2007, 129, 12084–12085. (b) Della Sala, G.;
Lattanzi, A. Org. Lett. 2009, 11, 3330–3333. (c) Larson, S. E.; Baso, J. C.;
Li, G.; Antilla, J. C. Org. Lett. 2009, 11, 5186–5189.
(11) Lou, S.; Schaus, S. E. J. Am. Chem. Soc. 2008, 130, 6922–6923.
(12) Harada, H.; Thalji, R. K.; Bergman, R. G.; Ellman, J. A. J. Org.
Chem. 2008, 73, 6772–6779.
(13) (a) Newman, C. A.; Antilla, J. C.; Chen, P.; Predeus, A. V.;
Fielding, L.; Wulff, W. D. J. Am. Chem. Soc. 2007, 129, 7216–7217.(b)
Newman, C. A. Ph. D. Thesis, Michigan State University, 2007.
(14) (a) Antilla, J. C.; Wulff, W. D. J. Am. Chem. Soc. 1999,
121, 5099–5100. (b) Antilla, J. C.; Wulff, W. D. Angew. Chem., Int. Ed.
2000, 39, 4518–4521. (c) Loncaric, C.; Wulff, W. D. Org. Lett. 2001,
3, 3675–3678. (d) Patwardan, A.; Pulgam, V. R.; Zhang, Y.; Wulff, W. D.
Angew. Chem., Int. Ed. 2005, 44, 6169–6172. (e) Deng, Y.; Lee, Y. R.;
Newman, C. A.; Wulff, W. D. Eur. J. Org. Chem. 2007, 2068–2071. (f) Lu,
Z.; Zhang, Y.; Wulff, W. D. J. Am. Chem. Soc. 2007, 129, 7185–7194. (g)
Zhang, Y.; Desai, A.; Lu, Z.; Hu, G.; Ding, Z.; Wulff, W. D. Chem.—Eur.
J. 2008, 14, 3785–3803. (h) Zhang, Y.; Lu, Z.; Desai, A.; Wulff, W. D.
Org. Lett. 2008, 10, 5429–5432. (i) Hu, G.; Huang, R. H.; Wulff, W. D.
J. Am. Chem. Soc. 2009, 131, 15615–15617. (j) Mukherjee, M.; Gupta,
A. K.; Lu, Z.; Zhang, Y.; Wulff, W. D. J. Org. Chem. 2010, 75, 5643. (k)
Desai, A.; Wulff, W. D. J. Am. Chem. Soc. 2010, 132, 13100–13103. (l)
Vetticatt, M.; Desai, A.; Wulff, W. D. J. Am. Chem. Soc. 2010,
132, 13104–13107. (m) Hu, G.; Gupta, A. K.; Huang, R. H.; Mukherjee,
M.; Wulff, W. D. J. Am. Chem. Soc. 2010, 132, 14669–14675. (n) Ren, H.;
Wulff, W. D. Org. Lett. 2010, 12, 4908–4911.
The (S)-VAPOL obtained as described above exists as a
solvate with CH2Cl2 containing two molecules of VAPOL per
molecule of CH2Cl2.30 To remove the CH2Cl2 all of the material
above was taken up in 250 mL of hexanes, and then the volatiles
were removed by rotatory evaporation and then under high vacuum
(0.1 mmHg) for 48 h. This gave an off white powder with mp 223À
226 °C (lit.30 226.2À227.9 °C) and [R]D 143.5 (CHCl3 c 1.0).
1
Spectral data for (S)-VAPOL: H NMR (CDCl3, 500 MHz)
δ 6.60 (s, 2H), 6.69 (d, 4H, J = 8.0 Hz), 6.96 (t, 4H, J = 8.0 Hz),
7.07 (t, 2H, J = 7.5 Hz), 7.45 (s, 2H), 7.63 (t, 2H, J =
8.5 Hz), 7.67À7.71 (m, 4H), 7.83 (d, 2H, J = 8.5 Hz), 7.94 (d,
2H, J = 8.5 Hz), 9.75 (d, 2H, J = 8.5 Hz); 13C NMR (CDCl3, 125
MHz) δ 115.84, 118.16, 123.25, 126.36, 126.83, 126.99, 127.03,
1106
dx.doi.org/10.1021/op200088b |Org. Process Res. Dev. 2011, 15, 1089–1107

Organic Process Research & Development
ARTICLE
(15) For a review, see: Zhang, Y.; Lu, Z.; Wulff, W. D. Synlett
2009, 2715–2739.
(16) Zheng, W.; Zhang, Z.; Kaplan, M. J.; Antilla, J. C. J. Am. Chem.
Soc. 2011, 133, 3339–3341.
(17) Zhang, Z.; Zheng, W.; Antilla, J. C. Angew. Chem., Int. Ed. 2011,
50, 1135–1138.
(18) Ren, H.; Wulff, W. D. J. Am. Chem. Soc. 2011, 133, 5656–5659.
(19) Ding, Z.; Xue, S.; Wulff, W. D. Chem--Asian J 201110.1002/
asia.201000804.
(20) Donaldson, N. The Chemistry and Technology of Naphthalene
Compounds; Edward Arnold Publishers: London, 1958; pp 238À242
and 427.
(21) Ota, E.; Iso, F. Nippon Kagaku Kaishi 1985, 1715–1720.
(22) (a) Mosettig, E.; Duvall, H. M. J. Am. Chem. Soc. 1937, 59,
367–369. (b) Behrend, R.; Ludewig, W. J. Lieb. Chem. 1911, 379,
351–362.
(23) Waters, M. L.; Wulff, W. D. Org. React. 2008, 70, 211–623.
(24) Yu, S.; Rabalakos, C.; Mitchell, W. D.; Wulff, W. D. Org. Lett.
2005, 7, 367–369.
(25) (a) Harrison, I. T.; Lewis, B.; Nelson, P.; Rooks, W.; Rosz-
kowski, A.; Tomolonis, A.; Fried, J. H. J. Med. Chem. 1970, 13, 203–205.
(b) Munoz-Muniz, O.; Juaristi, E. Tetrahedron Lett. 2003,
44, 2023–2026.
(26) Kipping, C.; Schiefer, H.; Schonfelder, K. J. Prakt. Chem. 1973,
315, 887–894.
(27) Redic, R.; Schuster, G. B. J. Photochem. Photobiol. A: Chem.
2006, 179, 66–74.
(28) Bandarage, U. K.; Come, J. H.; Green, J. Tetrahedron Lett. 2006,
47, 8079–8081.
(29) Alam, M. M.; Adapa, S. R. Synth. Commun. 2003, 33, 59–63.
(30) Price, C. P.; Matzger, A. J. J. Org. Chem. 2005, 70, 1–6.
(31) Edwards, J. D., Jr.; Cashaw, J. L. J. Am. Chem. Soc. 1957, 79,
2283–2285.
(32) An alternate solution is to coat the starting material onto glass
beads, but this becomes quite cumbersome upon scale-up.
(33) The original assignment that Salt I contains the (S)-VANOL
ligand has been reaffirmed:Polavarapu, P. L.; Petrovic, A. G.; Vick, S. E.;
Wulff, W. D.; Ren, H.; Ding, Z.; Staples, R. J. J. Org. Chem. 2009, 74,
5451–5457.
(34) This observation was made by Gang Hu and Chunrui Wu.
(35) Leakage of phenylacetylene from this joint is the most common
source of failure for this reaction.
(36) The same yield is obtained if the flow of air is excluded and
reflux is simply open to air: personal communication from Jon Antilla.
’ NOTE ADDED AFTER ASAP PUBLICATION
This paper was published on the Web on July 12, 2011, with a
production error in Table 8. The value for % ee 39 (0 years) has
been corrected for Al foil. The corrected version was reposted on
July 22, 2011.
1107
dx.doi.org/10.1021/op200088b |Org. Process Res. Dev. 2011, 15, 1089–1107