Isolable and Readily Handled Halophosphonium Pre-reagents for Hydro- and Deuteriohalogenation

Article abstract of DOI:10.1021/jacs.6b12653

Although the addition of acid halides across olefins is well-studied, limitations remain with a number of substrate classes that possess leaving groups, polyunsaturation, and acid-sensitive moieties, particularly polyenes prone to cyclization. The process is also challenging when conducted on a small scale, and moreover, methods for the addition of their deuterated counterparts typically require special techniques, especially when control of stoichiometry is required. Herein is described a readily synthesized and handled reagent class which can accomplish the controlled and selective Markovnikov addition of both HCl and HBr across several alkene classes under mild reaction conditions tolerant of diverse functionality. The process is particularly valuable on a laboratory scale, and direct comparisons to other methods are provided. As a result of in-depth mechanistic studies seeking to understand how these novel tools work and the active species behind their efficacy, the means to easily add DCl and DBr using a controlled amount of D₂O was discovered along with the critical role of hydrolysis in leading to active hydrohalogenation species.

Full text of DOI:10.1021/jacs.6b12653

Article
pubs.acs.org/JACS
Isolable and Readily Handled Halophosphonium Pre-reagents for
Hydro- and Deuteriohalogenation
Florian T. Schevenels,†,‡,§ Minxing Shen,†,‡,§ and Scott A. Snyder*,†,‡
†Department of Chemistry, The Scripps Research Institute, 130 Scripps Way, Jupiter, Florida 33458, United States
‡Department of Chemistry, University of Chicago, 5735 South Ellis Avenue, Chicago, Illinois 60637, United States
S
* Supporting Information
ABSTRACT: Although the addition of acid halides across
olefins is well-studied, limitations remain with a number of
substrate classes that possess leaving groups, polyunsaturation,
and acid-sensitive moieties, particularly polyenes prone to
cyclization. The process is also challenging when conducted on
a small scale, and moreover, methods for the addition of their
deuterated counterparts typically require special techniques, especially when control of stoichiometry is required. Herein is
described a readily synthesized and handled reagent class which can accomplish the controlled and selective Markovnikov
addition of both HCl and HBr across several alkene classes under mild reaction conditions tolerant of diverse functionality. The
process is particularly valuable on a laboratory scale, and direct comparisons to other methods are provided. As a result of in-
depth mechanistic studies seeking to understand how these novel tools work and the active species behind their efficacy, the
means to easily add DCl and DBr using a controlled amount of D2O was discovered along with the critical role of hydrolysis in
leading to active hydrohalogenation species.
surface-supported methods (such as PBr3/SiO2);4d,g,11 anti-
INTRODUCTION
■
Markovnikov side products are possible, though, if the reaction
Among all organic reaction transformations, the hydro-
is not under strict protection from air, light, and/or trace
halogenation of alkenes is one of the most fundamental;
peroxides. Overall, this reaction process is much less developed.
indeed, it has been studied since the dawn of the field and is
Additionally, few effective and practical methods exist for either
often among the first reactions taught to students in their
DCl or DBr addition, particularly on a laboratory scale.12
introductory organic chemistry courses.1 Moreover, the alkyl
Herein, we describe a new group of stoichiometric pre-
halide products of such additions have proven invaluable as
reagents, materials that combine a halophosphonium salt and a
synthons for further chemistry, particularly in the form of
Lewis acid, which, once hydrolyzed, can smoothly add HCl,
nucleophilic substitutions and metal-catalyzed cross couplings
HBr, and their deuterated counterparts to a variety of alkenes.
when the halide is primary or secondary and a range of
These materials include substrates that have historically proven
emerging approaches when the halide is tertiary.2
difficult to hydrohalogenate (or deuteriohalogenate), with the
Conventionally, Markovnikov HCl addition employs dry
developed technology both complimenting and enhancing
HCl gas or condensed liquid HCl,3 tools with sufficient
existing capabilities, particularly for compounds possessing
operational difficulty that a number of alternatives have been
advanced. These include the use of phase-transfer conditions
employing lipophilic phosphonium catalysts and aqueous HCl
additional elements of unsaturation in the form of alkenes and
alkynes. In several cases, direct comparisons of the efficiency of
these pre-reagents versus other available tools are presented.
solution, the in situ generation of HCl from highly reactive
Finally, the structural basis behind their unique properties is
inorganic/organic chlorides, as well as solid surface-promoted
also investigated, pointing to a mechanistic picture more
and metal-mediated hydrochlorinations;4 several indirect
complicated than serving simply as progenitors of hydrogen
halides.
approaches have been developed as well.5−7 Of significance,
Carreira8 recently demonstrated a radical-based process to
directly hydrochlorinate terminal alkenes with acid-sensitive
RESULTS AND DISCUSSION
■
moieties that are hard to functionalize through cationic
processes.9 Nevertheless, limitations still exist with these
1. Reagent Discovery and Initial Optimization. Our
entry to this field of study occurred during efforts to improve
methods, largely due to issues with unwanted leaving group
displacements, cyclization reactions, and other polymerization
events that can occur because of the inherent reaction
the reactivity profile of CDSC (Et2SCl·SbCl6, 1, Scheme 1),13,14
a highly reactive chloronium source15 we identified that could
achieve direct, chloronium-induced polyene cyclizations of
conditions, especially if conducted on a small scale. Unlike
HCl, solutions of HBr can readily add across alkenes10 but
typically do so under conditions that are quite harsh, with few
Received: December 9, 2016
alternatives identified for Markovnikov additions outside of
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.6b12653
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
and 6 as effective reagents in this regard, noting that their
structures were, at this point, only proposals due to challenges
in characterization beyond melting point analysis due to rapid
hydrolysis in standard NMR solvents (see the Supporting
Information). However, that structural uncertainty was viewed
as initially acceptable (though worth further analysis, vide infra)
given their ability to convert 7 predominantly into the
dichlorinated product 11 in a deliberate attempt to react both
alkenes by using an excess of reagent (5.0 equiv) in CH3NO2.
These reagents were themselves prepared in 63%, 97%, and
77% yield, respectively, by combining 1.1 equiv of the
phosphine ligand [either 1,2-bis(diethylphosphino)ethane
(depe), 1,2-bis(diphenylphosphino)ethane (dppe), or 1,3-
bis(diphenylphosphino)propane (dppp)] with 1.0 equiv of
Cl2 in 1,2-dichloroethane for 5 min at −30 °C, followed by the
addition of 1.1 equiv of SbCl5 and stirring for another 1−4 h at
ambient temperature; this yield value was based on the
assumed stoichiometry of 1:1:1 of the ligand, molecular
halogen, and the Lewis acid counterion in the reagent form
drawn. Following an extensive screen of additional Lewis acid
derived counterions,18,19 we found not only that reagent 12
based on TiCl4 was effective for HCl addition but also that the
TiBr4 and HfBr4 salts 13 and 14 (preliminary structural
proposals as noted above) could effect HBr addition under the
same general design.
Scheme 1. Development of New HCl and HBr Addition
Agents Based on Diphosphine Ligands Combined with
Molecular Halogen and MXy
Thus, with these tools in hand, we sought first to explore
their scope with a range of alkene substrates, selecting 5, 6, 13,
and 14 for these studies because of their overall ease of
synthesis, isolation, and stability as well as their general
reactivity profiles based on our initial screens. In section 4, we
return to efforts to determine their true structures as part of
mechanistic explorations.
2. Exploration of HCl Addition Scope. Starting with HCl
addition, a range of alkenes were readily functionalized using
either 5 or 6 (2.2 equiv; amount based on the proposed
structures in Scheme 1) with stirring in CH3NO2 at 23 °C for
3−16 h. The yields in Figure 1 are indicated specifically for
which reagent provided the outcome, noting that because 5 and
6 afforded very similar reactivity profiles, we did not test every
substrate with both of these tools. As shown, monoalkene
substrates with varying degrees of substitution afforded
products (15−26) with tolerance for common protecting
groups as well as heteroaromatic substituents with basic atoms,
including a phenyl sulfide (16), an unsubstituted indole (17), a
benzyl ether (23), and various ester moieties (21, 22, 24−
26).20
Next, in tests of chemoselectivity, the tools readily selected
for 1,1-disubstituted and trisubstituted alkenes over terminal/
internal alkynes (27 and 28) as well as monosubstituted
terminal alkenes (29) and 1,2-disubstituted alkenes (30 and
31). In fact, in initial tests using simple acyclic substrates
possessing only a monosubstituted terminal alkene or a 1,2-
disubstituted alkene, no reaction was observed with either 5 or
6.21 Importantly, these reagents can differentiate between
electronically similar alkenes in distinct steric environments as
highlighted by the formation of 32. In addition, 1,1-
disubstituted and trisubstituted alkenes could be selected over
an alkene in an α,β-unsaturated system (33) as well as a
tetrasubstituted (and partially deactivated) alkene (34). A
highly reactive benzyl bromide (35) was also obtained largely
intact, with less than 10% displacement by chloride; neither this
material, nor the alkyne-containing products, would likely have
survived radical-based hydrochlorination.8,22
substrates such as 7 into 8, albeit in modest yields and
diastereoselectivity.13 One design that we explored was
exchanging the sulfur-derived Lewis base of 1 for a
phosphine.16 Initial experiments revealed that the tri-n-
butylphosphine derivative 2 was a reagent that was too
unstable to be isolated, while its triphenylphosphine analog
(3) was a discrete solid. When this new material (3) was
exposed to polyene 7, the major product was not the
chlorinated and cyclized adduct 8 but instead the proton-
cyclized product 9; intriguingly, we also observed small
amounts of monohydrochlorinated and noncyclized 10, a
material we had not previously obtained to any appreciable
extent. As such, we sought to determine if we could select for
this product exclusively.
Rationalizing that the above result with substrate 7 reflected
rapid hydrolysis of 3 in CH3NO2 (vide infra), we attempted
next to discover a more stable form of the reagent by screening
commercial phosphines including bidentate phosphine ligands.
Our thought was that an additional phosphorus atom could
potentially complex with the proposed P−Cl+ cation through a
Lewis acid/Lewis base interaction (one which might be
stabilizing), noting that some complexes made from chlorine
and bidentate phosphines had already been reported.17 After
exploring several such variants, we identified complexes 4, 5,
B
DOI: 10.1021/jacs.6b12653
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 2. Additional HCl addition products generated with reagents 5
and 6 using polyene substrates. Yields in blue are either literature
outcomes or reactions we performed with the alternative reagent
indicated.
yields were still acceptable, however. One possible explanation
for these differential yields could be that a diminished ability of
the ester carbonyl group (comparing a trifluoroacetate to an
acetate) to trap an intermediate carbocation affords higher
product throughput.
Globally, while simple monoalkene substrates can certainly
be hydrochlorinated with other tools, we believe that the main
advantage of 5 and 6, apart from ease of use and handling, is for
polyunsaturated alkenes, where careful control of stoichiometry
is required on laboratory scale. We do note that their
preparation does require the use of Cl2 gas; however, that
use can be on a single occasion to make a large amount of 5
and/or 6 for multiple, future uses. As will be noted in section 4
when we return to the issue of reagent structure, it might be
possible to prepare these reagents solely with inorganic sources
of chlorine such as PCl3 as well.
3. Exploration of HBr Addition Scope. Figure 3 presents
the analogous results for HBr addition with a smaller number of
substrates using 13 and 14, separating the two reagents based
on some reactivity differences observed with certain substrates.
Of note, benzyl ethers and basic nitrogen atoms were tolerated
with much of the same chemoselectivity observed in the
hydrochlorinations above retained, though in some cases, such
as with product 47, yields were reduced. In most cases,
however, the throughput was effective, such as the formation of
43 in 91% yield.
The main difference between these hydrobrominating tools
and their chlorine counterparts, however, was with polyene
substrates. Hf-based reagent 14 proved capable of generating
monohydrobrominated products such as sulfone 46 in a
reasonable yield (61%), while Ti-based 13 instead afforded full
alkene consumption of both the distal and internal alkenes,
yielding the dihydrobrominated product 48 under the standard
reaction conditions.24 The difference in reaction times between
the two tools is not responsible for the double addition
observed with 13; the longer time was used to drive the process
to completion, noting that early stoppage of the reaction gave
an inseparable mixture of both mono- and dihydrobrominated
material.
Figure 1. HCl addition products generated with reagents 5 and 6.
Yields in blue are either literature outcomes or reactions we performed
with the alternative reagent indicated.
Finally, a series of polyenes were monohydrochlorinated at
their terminal olefins with good selectivity to deliver 36−40
(Figure 2), even when reactive leaving groups were present
such as the trifluoroacetate moiety within 39c. Our reagents
worked especially well compared to the known distal alkene
hydrochlorination methods,23 such as that promoted by TiCl4
on sulfone-type substrates 36 (97% vs 25%)23b and 37 (95% vs
12%) and geranyl benzoate (40, 82% vs 42%). Intriguingly, for
reasons that are not entirely clear, the yields observed with
geranyl acetate (39a, 40% vs 97%) and geranyl pivalate (39b,
57% vs 94%) were inferior both to TiCl4 and the outcome
obtained with our tools in generating 39c, an arguably more
sensitive substrate because of its terminal functional group; the
The key point for this chemistry, however, is that controlled
monohydrobromination, particularly for polyunsaturated sub-
strates, is much more challenging to achieve with other
C
DOI: 10.1021/jacs.6b12653
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 2. Mechanistic Studies Reveal the Source of the
Hydrogen Atom and Allow for Facile Incorporation of
Deuterium along with Cl and Br in Various Substrates Using
D2O
incorporation based on available detection methods.25,26
Outside of indicating the key role of water (with quantification
experiments described at the end of this section, vide infra), this
result formed the basis for a highly efficient means to effect DCl
and DBr addition across alkenes using the cheapest and most
readily available source of deuteron. As illustrated for the
additional substrates in Scheme 2, the method afforded
deuteriohalogenated products 52−5527 with equivalent func-
tional group tolerance, as well as in similar yields, as the original
reaction processes.
Figure 3. HBr addition products generated with reagents 13 and 14.
Yields in blue are either literature outcomes or reactions we performed
with the alternative reagent indicated.
available approaches. As highlighted by the yields in blue, use of
HBr/AcOH in our hands proved challenging with several of the
selected substrates, either affording decomposition (as with
benzyl ether 41) or full conversion into dihydrobrominated
products (as was observed with the substrates leading to alkyne
44 and alkene 47). Of note, we were also able to form 43 in
85% yield with HBr/AcOH, noting that a previous literature
report with this same combination documented 53% yield on a
larger scale.2c
Based on this knowledge and other evidence as documented
below, we believe we possess a reasonable picture for the
overall process behind these reactions, especially for those
systems using antimony(V) as source of the Lewis acid
counterion. First, we observed that when we exposed the
natural product (−)-nootkatone (56, Scheme 3) to 2.2 equiv of
5, unique product formation occurred to deliver 57 in 91%
yield wherein both alkene hydrochlorination and conversion of
the ketone to a vinyl chloride occurred; this outcome suggested
to us that an initial reagent structure with reactivity like POCl3
or PCl5 could be plausible, implicating a P−Cl species.28,29
That supposition was verified when very careful reagent
preparation and rapid X-ray crystallographic analysis under
strictly anhydrous conditions revealed the true nature of
reagents 5 and 6 as drawn in the middle section of Scheme 3,
differing only in terms of the counterion where antimonate
reduction has occurred in both cases;30 as indicated in the
Supporting Information, we also obtained some NMR evidence
for 5 and 6 in this form following their exposure to anhydrous
CD3CN. What this finding suggests, when 5 is used for
purposes of illustration as drawn in the bottom portion of
Scheme 3, is that after the generation of the initially proposed
reagent structure in terms of its cation component a redox
reaction with SbCl5 afforded the dichlorinated ligand along with
both [SbCl4]− and Cl−. That new antimonate ion (itself never
observed in our crystal structures) either merged with Cl− to
generate [SbCl5]2− or with another molecule of itself to make
[Sb2Cl8]2−. We believe that both of these ions, as well as
potentially Cl−, populate the anionic portion of the reagent, and
different crystals might then likely possess different counterions
as we observed with 5 and 6.31
4. Mechanistic Investigation and DCl/DBr Additions.
Given this body of results, we sought to understand both the
mechanism for the reaction process as well as establish the
structures of our reagents and/or active species, since the
structures as drawn within Scheme 1 as noted previously are
only proposals based on stoichiometry with no evidence for
their connectivities. We began these investigations by
attempting to identify the source of the proton in the final
products. Critically, we noticed that when our reaction solvent
(CH3NO2) was scrupulously dry, the reaction proceeded in far
inferior yield, suggesting that adventitious water might be the
source of proton as well as a requisite part of the transformation
in generating an active reagent (vide infra); indeed, NMR
characterization studies (see the SI) showed that all of our
1
reagents produced unique spectra based on 31P, 13C, and H
NMR following dissolution for varied times in different
solvents, suggesting hydrolysis as a critical mechanistic process.
In initial support of that idea, we had utilized commercial,
nondried CH3NO2 as the solvent in all of the reactions of the
preceding sections, a solvent which possessed between 500 and
2000 ppm of H2O as measured by Karl−Fischer titration, and
we did not flame-dry our reaction flasks.
As shown in Scheme 2, that supposition was further verified
with the use of CD3NO2 affording no D incorporation within
product 15, while the use of 5 in D2O-saturated CH3NO2
generated 51 in 65% yield with effectively complete (>95%) D
As already noted, however, 5 and 6 are best described as pre-
reagents, with rapid hydrolysis affording a new species as
observed in those same NMR spectra taken in CD3NO2; it is
D
DOI: 10.1021/jacs.6b12653
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 3. Initial Mechanistic Insights Gained by Unique
Product Formation and X-ray Data for 5 and 6 Revealing
Their True Structures
Scheme 4. X-ray Structure of Complex 59 and Proposed
Structure for the Active Hydrochlorinating Reagent as
Compound 58
a
other species as part of alkene hydrochlorination mechanistic
pictures such as that drawn in the right-hand portion of the
inset of Scheme 4.33 Equally critical, we obtained an X-ray
structure of a hydrolyzed variant of 5 from wet CH3NO2 in the
form of 59, a dimeric protonated bis(phosphine oxide)
containing half of an [Sb2Cl8]2− anion and one molecule of
CH3NO2 in the unit cell. This compound can be referred to as
[dppeO2H]+(1/2)[Sb2Cl8]2− based on the nature of ions or as
denoted within the scheme and as will become clear shortly, as
dppeO2·SbCl3·HCl based on its overall composition and key
atom oxidation states, ignoring the solvent molecule.34 Part of
the molecule is shown in Scheme 4, highlighting the placement
of the proton attached to one of the phosphine oxides; see the
Supporting Information for full drawings of this complex.
In an effort to further verify the structure of the active
1
species, we compared the H and 31P NMR spectra of various
independently prepared phosphine dioxides in CD3CN as well
as performed a number of experiments with 59 and other
controls.35 For example, we attempted an independent
preparation of 59 by combining equimolar amounts of
dppeO2, SbCl3, and HCl; as shown in Table 1, that material
1
Table 1. Key H and 31P NMR Shifts for Several Phosphine
Dioxides and Reagent 5 before and after Hydrolysis
entry
reagent
1H δ (ppm, m) 31P δ (ppm)
1
2
3
4
dppeO2·SbCl3·HCl
dppeO2·SbCl3·HCl + HCl (ex)
5 after hydrolysis
2.86 (d)
3.05 (d)
3.04 (d)
nd
43.3
50.7
48.9−50.4
73.9
a
A mechanism to account for the formation of these materials is also
5 before hydrolysis
provided at the bottom of the Scheme.
1
the resultant material that is expected to be the active
hydrochlorinating species. As shown in Scheme 4, we believe
that this new species is the protonated bis(phosphine oxide)
HCl complex 58, drawn here as only the cationic portion with
the assumption being that several antimonate species and/or
chloride are also present as already discussed above. As such,
HCl coordination to the other components of the reagent is
critical for the selectivity and reactivity observed; i.e., it is more
than simple in situ generation of HCl, though free HCl could
reasonably be expected to convert a portion of the alkene into
the hydrochlorinated product as well.
did not possess H and 31P spectra which matched that of the
species generated by dissolving 5 in CD3CN, assuming that
adventitious water afforded the active reagent in the latter case.
However, the observed chemical shifts of dppeO2·SbCl3·HCl +
HCl was a near match to 59 (i.e., 5 after hydrolysis), with the
exact stoichiometry of HCl possibly accounting for the subtle
chemical shift differences.
As indicated in Scheme 5, exposure of substrate 50 to 1.0 M
HCl in dry CH3NO2 afforded product 15 in only 23% yield
with significant starting material remaining after an extended
reaction time (40 h). If the oxidized dppe ligand (dppeO2) was
added to 1.0 M HCl in dry CH3NO2, the yield increased to
64% within only 16 h of reaction time, suggesting the role of
In support of this picture, complexes of type 58 are known,32
and modern texts illustrate similar types of HCl coordination to
E
DOI: 10.1021/jacs.6b12653
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
activates another molecule of HCl, thus enhancing its reactivity
(as shown in 58, Scheme 4).
Scheme 5. Further Support for the HX Addition Mechanism
Finally, given the seemingly indispensible role of water in the
reaction as noted above, including the observation of some
differential yields based on water content as already mentioned
in the context of different reagent mixtures within the confines
of Scheme 5, we sought to provide a more quantitative
assessment of how much water is required/tolerated. In a
similar vein, we wanted to see if reagent age (and attendant
degree of hydrolysis) could be a relevant factor, aiming to
ensure reproducibility of the processes described earlier. Thus,
we set up reactions of substrate 50 for a total stir time of 4 h
with both freshly prepared 5 and a batch of 5 stored for >1
year, using CH3NO2 with three different levels of water
content: 30, 670, and >30000 ppm (i.e., H2O-saturated
CH3NO2); the water content was determined by Karl−Fischer
titration by measuring a blank flame-dried experiment flask
containing solvent as well as all reagents except for 5.
As shown in Table 2, we projected yields for both recovered
starting material and product based on the idea that 1 equiv of
water can maximally generate 2 equiv of HCl in hydrolyzing
pre-reagent 5 into 59. As indicated, the year-old reagent worked
in all cases, affording product yields which decreased somewhat
as water content increased, while reactions with fresh reagent
cases closely followed initial projections, though consistently
went further to completion than might have been expected.
Our analysis of these outcomes is that a certain degree of
reagent hydrolysis had also occurred before 5 was added to the
solution; for year-old reagent, stored in a bottle opened
multiple times for different experimental uses,37 that hydrolysis
was likely already complete, while for freshly prepared reagent,
it was a level of partial hydrolysis which provided more active
reagent than would be expected based on the reaction vessel
water-content alone. A key finding, however, is that too much
water does not cause problems given the yields observed;
sufficient water content and/or reagent hydrolysis is, on the
other hand, necessary for complete conversion. As already
denoted, commercial CH3NO2 in combination with the use of a
nonflame-dried flask (measured water content >500 ppm) was
sufficient to effect the reactions without the need for addition of
extra water.
a
b
2.2 equiv of all reagents used at 23 °C. Stopped after incomplete
c
conversion following 40 h of stirring. Stopped after 16 h of stirring.
complexation of HCl in effecting the process as drawn within
58 (Scheme 4); this result also suggests that the antimonate
components of our reagents are spectator ions. To then try and
generate 58 “in situ” from 59, we added HCl to 59 in both wet
and dry CH3NO2, finding that only in dry solvent did the
reaction proceed with similar efficacy (68% yield versus 23%
yield). This finding mirrors that of our NMR studies where the
addition of HCl to independently prepared dppeO2·SbCl3·HCl
was necesssary to generate spectra that matched hydrolyzed 5.
In addition, this result indicates that initial reagent hydrolysis is
important but that overall water content might matter, a feature
we will discuss in more detail shortly.
For now, though, it is important to note that despite the
ability to seemingly reproduce the active species using 59, from
our perspective, both in terms of final yield (79% versus 68%
here) as well as operational simplicity, the in situ generation of
the active species from 5 is still preferred as it requires simple
addition to the alkene of interest in commercial, undried
CH3NO2 without any need to measure/handle HCl or
rigorously dry the CH3NO2 solvent. As a further observation,
31P NMR monitoring during the course of the reaction of 50
with 5 showed the reagent 31P peak moving upfield gradually,
broadening but getting closer to the value of the protonated
dioxide antimony(III) complex 59 (43.3 ppm), suggesting a
progressive loss of HCl. Thus, the superior reactivity observed
in the in situ protocol could reflect a Brønsted-assisted
Brønsted acid as proposed by Yamamoto,36 where in this
case the protonated oxygen atom on the phosphine oxide
In addition, we sought to understand the necessity of
employing 2.2 equiv of the pre-reagents in order to obtain
reasonable yields of product as originally determined in our
optimization experiments. As shown in Figure 4, using three
Table 2. Further Efforts To Determine the Role of Water Content Amount and Reagent Age on the Reproducibility and Facility
of the Hydrochlorination of Substrate 50
reagent age
(2.2 equiv)
H2O content (ppm)
projected starting material
recovered (%)
isolated starting material
recovered (%)
projected product
yield (%)
isolated product
yield (%)
a
b
entry
(1.5 mL of solvent)
1
2
3
4
5
6
1 year
1 year
1 year
fresh
30
670
97
41
0
0
0
3
59
57
45
44
11
46
65
>30000
30
0
100
3
97
41
0
62
29
0
fresh
670
59
fresh
>30000
100
a
b
Reactions were all performed on a 0.215 mmol scale using 2.2 equiv of reagent 5 in CH3NO2 at 23 °C, stopping the reaction after 4 h. H2O
content of the CH3NO2 solutions was determined by blank parallel experiments.
F
DOI: 10.1021/jacs.6b12653
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Detailed experimental procedures, copies of all spectral
data, X-ray data, and full characterization (PDF)
X-ray data for compound 5 (CIF)
X-ray data for compound 6 (CIF)
X-ray data for compound 59 (CIF)
AUTHOR INFORMATION
Corresponding Author
*sasnyder@uchicago.edu
■
ORCID
Scott A. Snyder: 0000-0003-3594-8769
Author Contributions
§F.T.S. and M.S. contributed equally to the science developed
in this manuscript.
Figure 4. Rate of starting material disappearance in the reaction
converting 50 into 15 using freshly prepared 5 in varying amounts.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
different amounts of reagent 5 with substrate 50 under the
standard reaction conditions and measuring the starting
material remaining by NMR analysis, each of the reactions
proceeded quickly at the start and then slowed down as time
progressed. However, in no case did consumption of the
substrate stop; even minor consumption continued after 6 h
with 1.1 or 1.65 equiv of 5 used. This outcome might reflect the
fact the hydrochlorination relies on a certain concentration of
active “HCl” and its consumption obviously lowers its
concentration and thus slows down the reaction. Another
possibility is that the initially generated active complexes
between phosphine oxide and HCl dissociate slowly over time;
as such, 2.2 equiv might reflect a necessary stoichiometry to
ensure complete conversion on a reasonable time-scale without
the formation of excessive amounts of side-products.
As a final comment, while we have a strong body of collated
evidence for the properties of reagents 5 and 6 at this juncture,
the overall process for reagents 12, 13, and 14 may be more
complex in terms of the intermediate species produced,38
though all evidence points similarly to the importance of
reagent hydrolysis in leading to the active species; further work
is needed to clarify the specific details.
We thank Dr. Kevin Gagnon (Lawrence Berkeley) for
obtaining an X-ray crystal structure of 59; the Advanced
Light Source is supported by the Director, Office of Science,
Office of Basic Energy Sciences, of the U.S. Department of
Energy under Contract No. DE-AC02-05CH11231. We also
thank Dr. Alexander Filatov (University of Chicago) for
obtaining X-ray structures of 5 and 6 and Dr. Antoni Jurkiewicz
(University of Chicago) and Dr. C. Jin Qin (University of
Chicago) for help with NMR spectroscopy and mass
spectrometry, respectively. Financial support for this work
came from The Scripps Research Institute, the University of
Chicago, and the Belgian American Education Foundation
(fellowship to F.T.S.).
REFERENCES
■
(1) Solomons, T. W. G.; Fryhle, C. B.; Snyder, S. A. Organic
Chemistry, 12th ed.; Wiley, 2016; pp 1124.
(2) For recent, selected examples with tertiary systems, see:
(a) Dudnik, A. S.; Fu, G. C. J. Am. Chem. Soc. 2012, 134, 10693.
(b) Zultanski, S. L.; Fu, G. C. J. Am. Chem. Soc. 2013, 135, 624.
(c) Zhao, C.; Jia, X.; Wang, X.; Gong, H. J. Am. Chem. Soc. 2014, 136,
17645. (d) Wang, X.; Wang, S.; Xue, W.; Gong, H. J. Am. Chem. Soc.
2015, 137, 11562. (e) Atack, T. C.; Cook, S. P. J. Am. Chem. Soc. 2016,
138, 6139. (f) Peacock, D. M.; Roos, C. B.; Hartwig, J. F. ACS Cent.
Sci. 2016, 2, 647.
(3) (a) Whitmore, F. C.; Johnston, F. J. Am. Chem. Soc. 1933, 55,
5020. (b) Schmerling, L. J. Am. Chem. Soc. 1946, 68, 195. (c) Dewar,
M. J. S.; Fahey, R. C. J. Am. Chem. Soc. 1963, 85, 2245. (d) Brown, H.
C.; Rei, M.-H. J. Org. Chem. 1966, 31, 1090. (e) Stille, J. K.;
Sonnenberg, F. M.; Kinstle, T. H. J. Am. Chem. Soc. 1966, 88, 4922.
(f) Fahey, R. C.; McPherson, C. A. J. Am. Chem. Soc. 1971, 93, 2445.
(g) Becker, K. B.; Grob, C. A. Synthesis 1973, 1973, 789. (h) Becker,
K. B.; Grob, C. A. Helv. Chim. Acta 1973, 56, 2723. (i) Frøyen, P.;
Skramstad, J. Synth. Commun. 1994, 24, 1871.
(4) (a) Kennedy, J. P.; Sivaram, S. J. Org. Chem. 1973, 38, 2262.
(b) Landini, D.; Rolla, F. J. Org. Chem. 1980, 45, 3527. (c) Khurana, J.
M.; Tetenyi, P.; Kodomari, M.; Steven, S. Indian J. Chem. 1988, 27B,
1129. (d) Kropp, P. J.; Daus, K. A.; Crawford, S. D.; Tubergen, M. W.;
Kepler, K. D.; Craig, S. L.; Wilson, V. P. J. Am. Chem. Soc. 1990, 112,
7433. (e) Alper, H.; Huang, Y. Organometallics 1991, 10, 1665.
(f) Delaude, L.; Laszlo, P. Tetrahedron Lett. 1991, 32, 3705. (g) Kropp,
P. J.; Daus, K. A.; Tubergen, M. W.; Kepler, K. D.; Wilson, V. P.;
Craig, S. L.; Baillargeon, M. M.; Breton, G. W. J. Am. Chem. Soc. 1993,
115, 3071. (h) Boudjouk, P.; Kim, B.-K.; Han, B.-H. Synth. Commun.
1996, 26, 3479. (i) de Mattos, M. C. S.; Sanseverino, A. M. Synth.
CONCLUSION
■
A new group of pre-reagents is reported that, when hydrolyzed,
are efficient in converting a range of alkenes into their
hydrohalogenated counterparts, with a simple procedure using
D2O-saturated CH3NO2 affording the means to obtain their
deuterated congeners with seemingly complete (i.e., >95%)
isotopic labeling. Their critical value is their ease of use and
effectiveness on polyunsaturated substrates where careful
control of stoichiometry is required. Extensive mechanistic
studies point to the likely active species, suggesting that there is
complexity beyond simple HCl and HBr generation that
accounts for the outcomes observed. Further potential
applications with other functional groups are the subject of
current investigations, as are the development of additional
reagents based on related designs.39
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.6b12653.
■
S
G
DOI: 10.1021/jacs.6b12653
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Commun. 2000, 30, 1975. (j) Yadav, V. K.; Babu, K. G. Eur. J. Org.
Chem. 2005, 2005, 452.
(16) Ionic salts of R3P with molecular halogen and SbCl5 are known:
(a) Wiley, G. A.; Stine, W. R. Tetrahedron Lett. 1967, 8, 2321.
(5) Indirect, but powerful, methods that reduce vinyl chlorides and
other rearrangement processes for alkyl bromides have also arisen. For
example: (a) Grigg, R. D.; Van Hoveln, R.; Schomaker, J. M. J. Am.
Chem. Soc. 2012, 134, 16131. (b) Grigg, R. D.; Rigoli, J. W.; Van
Hoveln, R.; Neale, S.; Schomaker, J. M. Chem. - Eur. J. 2012, 18, 9391.
(c) Iwasaki, K.; Wan, K. K.; Oppedisano, A.; Crossley, S. W. M.;
Shenvi, R. A. J. Am. Chem. Soc. 2014, 136, 1300. (d) King, S. M.; Ma,
X.; Herzon, S. B. J. Am. Chem. Soc. 2014, 136, 6884. (e) Van Hoveln,
R.; Schmid, S. C.; Tretbar, M.; Buttke, C. T.; Schomaker, J. M. Chem.
Sci. 2014, 5, 4763. (f) Van Hoveln, R.; Hudson, B. M.; Wedler, H. B.;
Bates, D. M.; Le Gros, G.; Tantillo, D. J.; Schomaker, J. M. J. Am.
Chem. Soc. 2015, 137, 5346.
(b) Hartke, J.; Akgun, E. Chem. Ber. 1979, 112, 2436.
̈
(17) For such complexes, see: (a) Ellermann, V. J.; Thierling, M. Z.
Anorg. Allg. Chem. 1975, 411, 15. For more recent references to the
growing body of literature indicating the existence of phosphine-
phosphenium coordination complexes, see: (b) Burford, N.; Herbert,
D. E.; Ragogna, P. J.; McDonald, R.; Ferguson, M. J. J. Am. Chem. Soc.
2004, 126, 17067 and references cited therein. (c) Carpenter, Y.;
Burford, N.; Lumsden, M. D.; McDonald, R. Inorg. Chem. 2011, 50,
3342 and references cited therein. In these papers, an arrow would be
drawn from the non-charged phosphine to the halophosphonium
species to indicate that coordination:
(6) Alternatively, it is also possible to indirectly form such products
through a hydrometalation/halogen trap sequence with a variety of
́
metals. For example: (a) Brown, H. C.; Rathke, M. W.; Rogic, M. M. J.
(18) See the Supporting Information section for the full range of
counterions attempted. Efforts to make corresponding fluorinated or
iodinated counterparts failed in terms of appropriate alkene reactivity.
(19) As denoted in the Supporting Information, some Lewis acids
screened afforded protocyclization as either the major and/or
substantial product. At present, since many of these reactions gave
product mixtures, not single adducts, it is challenging to provide a
global basis to understand the differential reactivity. One hypothesis,
however, is the overall ability of each Lewis acid to dissociate once in
solution and afford free halide ions that could readily make HCl or
HBr, with greater content of those species facilitating the
protocyclization reaction. For a recent paper describing the
observation of similar Lewis acid dependence in either protocyclization
or hydrochlorination, see: Li, S.; Chiu, P. Tetrahedron Lett. 2008, 49,
1741.
Am. Chem. Soc. 1968, 90, 5038. (b) Brown, H. C.; Lane, C. F. J. Am.
Chem. Soc. 1970, 92, 6660. (c) Brown, H. C.; Lane, C. F. J. Am. Chem.
Soc. 1970, 92, 7212. (d) Lane, C. F.; Brown, H. C. J. Organomet. Chem.
1971, 26, C51. (e) Lane, C. F. J. Organomet. Chem. 1971, 31, 421.
(f) Sato, F.; Mori, Y.; Sato, M. Chem. Lett. 1978, 7, 833. (g) Makabe,
H.; Negishi, E. Eur. J. Org. Chem. 1999, 1999, 969. (h) Gagneur, S.;
Makabe, H.; Negishi, E. Tetrahedron Lett. 2001, 42, 785. (i) Podhajsky,
S. M.; Sigman, M. S. Organometallics 2007, 26, 5680.
(7) For other syntheses of tertiary and secondary halides through C−
H functionalization approaches, see the following for chloride
additions: (a) Kundu, R.; Ball, Z. T. Org. Lett. 2010, 12, 2460.
(b) Liu, W.; Groves, J. T. J. Am. Chem. Soc. 2010, 132, 12847. (c) Qin,
Q.; Yu, S. Org. Lett. 2015, 17, 1894. (d) Quinn, R. K.; Konst, Z. A.;
̈
Michalak, S. E.; Schmidt, Y.; Szklarski, A. R.; Flores, A. R.; Nam, S.;
Horne, D. A.; Vanderwal, C. D.; Alexanian, E. J. J. Am. Chem. Soc.
2016, 138, 696. (f) Wang, Y.; Li, G.-X.; Yang, G.; He, G.; Chen, G.
Chem. Sci. 2016, 7, 2679. For bromide additions: (g) Schmidt, V. A.;
Quinn, R. K.; Brusoe, A. T.; Alexanian, E. J. J. Am. Chem. Soc. 2014,
136, 14389. See also ref 7b,f.
(20) The Carreira method (ref 8) worked especially well for several
of these same, monoalkene-containing substrates.
(21) The one exception to this trend was with a special class of
monosubstituted terminal alkenes in the form of styrenes. For
example, 4-tert-butylstyrene and 4-chlorostyrene did react with
reagents 5 and 6 but typically did so with incomplete conversion as
well as concomitant benzyl chloride hydrolysis, rendering these
substrates nonviable overall.
(22) The main functional groups that were not tolerated with our
reagents were free alcohols and silyl protecting groups, presumably
due to a pathway similar to alcohol activation by the Hendrickson
reagent: Hendrickson, J. B.; Schwartzman, S. M. Tetrahedron Lett.
1975, 16, 277.
(8) Gaspar, B.; Carreira, E. M. Angew. Chem., Int. Ed. 2008, 47, 5758.
(9) For subsequent papers that extend these findings in additional
directions, see: (a) Leggans, E. K.; Barker, T. J.; Duncan, K. K.; Boger,
D. L. Org. Lett. 2012, 14, 1428. (b) Ma, X.; Herzon, S. B. Chem. Sci.
2015, 6, 6250.
(10) Mayo, F. R.; Walling, C. Chem. Rev. 1940, 27, 351.
(11) Sanseverino, A. M.; de Mattos, M. C. S. J. Braz. Chem. Soc. 2001,
12, 685.
(12) Typically, DCl or DBr is made from the reaction of D2O and a
species such as PCl3, TiCl4, PBr3, or BBr3, with distillation, trapping,
titration, and/or weighing necessary for controlled use: (a) Brown, H.
C.; Groot, C. J. Am. Chem. Soc. 1942, 64, 2223. (b) Skell, P. S.; Allen,
R. G. J. Am. Chem. Soc. 1959, 81, 5383. (c) Dewar, J. M. S.; Fahey, R.
C. J. Am. Chem. Soc. 1963, 85, 2245. (d) Brown, H. C.; Liu, K.-T. J.
Am. Chem. Soc. 1975, 97, 600. (e) Hassner, A.; Fibiger, R. F. Synthesis
1984, 1984, 960. An alternate approach for DBr addition is
deuterioboration followed by bromination: (f) Han, B.; Ph.D. Thesis,
Rochester Institute of Technology, 1991.
(23) (a) Julia, M.; Roy, P. Tetrahedron 1986, 42, 4991. (b) Demotie,
A.; Fairlamb, I. J. S.; Radford, S. K. Tetrahedron Lett. 2003, 44, 4539.
(24) Intriguingly, additional equivalents of reagent 14 or further
stirring did not lead to the dihydrobrominated adduct from product
46. We propose that the discrepancy in the reactivity of these two
reagents with 46 and 48 might result from the higher Lewis acidity
and/or electrophilicity of Ti(IV) versus Hf(IV). This phenomenon,
however, is unique to the hydrobromination process as the chlorine
counterpart of the Ti(IV) reagent (i.e., 12) did not provide
dihydrochlorinated product.
(25) D2O-saturated CH3NO2 was prepared by the following
procedure: CH3NO2 (10 mL) was added under argon to a flask
containing CaH2, and the resultant slurry was heated at 60 °C for 2
min and then was allowed to cool to 23 °C. It was then distilled, with a
total of 5−6 mL of CH3NO2 collected. After cooling to 23 °C, the
distillate was dried over freshly activated 4 Å molecular sieves (flame-
dried under high vacuum). It was then transferred into a flame-dried
flask under argon, D2O (0.2 mL) was added, and the biphasic mixture
was stirred for 5 min at 23 °C. At this time, stirring was stopped, and
the CH3NO2 was taken out carefully via syringe so as not to disturb
the D2O droplets in the original flask. It was then added into another
flame-dried flask, and the same process of D2O addition and syringe
removal was repeated twice to obtain D2O-saturated CH3NO2.
(13) Snyder, S. A.; Treitler, D. S.; Brucks, A. P. J. Am. Chem. Soc.
2010, 132, 14303.
(14) For the related Br+ and I+ sources, see: (a) Snyder, S. A.;
Treitler, D. S. Angew. Chem., Int. Ed. 2009, 48, 7899. (b) Snyder, S. A.;
Treitler, D. S.; Brucks, A. P.; Sattler, W. J. Am. Chem. Soc. 2011, 133,
15898. (c) Snyder, S. A.; Brucks, A. P.; Treitler, D. S.; Moga, I. J. Am.
Chem. Soc. 2012, 134, 17714. For a recent paper developing active
bromonium species with catalytic amounts of sulfur ligands, see:
(d) Ke, Z.; Tan, C. K.; Chen, F.; Yeung, Y.-Y. J. Am. Chem. Soc. 2014,
136, 5627.
(15) For its relative reactivity, see: Ashtekar, K. D.; Marzijarani, N. S.;
Jaganathan, A.; Holmes, D.; Jackson, J. E.; Borhan, B. J. Am. Chem. Soc.
2014, 136, 13355.
H
DOI: 10.1021/jacs.6b12653
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(26) Preparation of H2O-saturated CH3NO2 in a similar way as
described above afforded a water content measurement of >30000
ppm based on Karl−Fischer titration (i.e., out of the limit of
detection).
(39) For a recent example where we developed competent RS+
sources for polyene cyclizations using related components, see:
Schevenels, F. T.; Shen, M.; Snyder, S. A. Org. Lett. 2017, 19, 2.
(27) Interestingly, compound 53 possesses two chiral centers
following deuteriobromination; based on the 1H and 13C NMR
obtained, however, only a single diastereomer is formed (noting that
that CH2D signal overlaps other protons). This result might suggest a
potential directing effect from the OBz group within the initial
substrate.
(28) Banerjee, A. S.; Engel, R.; Axelrad, G. Phosphorus Sulfur Relat.
Elem. 1983, 15, 15.
(29) Although we presume that 13 and 14 can form active species
similar to those from the corresponding chlorine reagents, we did not
note the formation of a similar vinylbromide product using substrate
56; instead, as denoted in Figure 3, we obtained product 45 with
reagent 14.
(30) The structure of 5 has previously been observed by X-ray: Byers,
H. L.; Dillon, K. B.; Goeta, A. E. Inorg. Chim. Acta 2003, 344, 239.
This material was prepared differently, using PCl3 instead of Cl2 gas,
suggesting it might be possible to avoid gaseous reagents if desired to
prepare 5 and 6
(31) From crystal 5, in the [SbCl5]2− portion the Sb atom with five
neighboring Cl has bond distances of 2.369, 2.617, 2.623, 2.630, and
2.611 Å along with bond angles that are all around 90 degrees. From
crystal 6, in the [Sb2Cl8]2− portion the Sb atom with 5 neighboring Cl
has bond distances of 2.368, 2.407, 2.476, 2.813, and 2.954 Å and bond
angles close to 90 degrees with some distortions. Similar Sb(III)
anions have been observed in many compounds. For selected
examples reporting [SbCl5]2−, see ref 30 and: (a) Webster, M.;
Keats, S. J. Chem. Soc. A 1971, 298. (b) Bujak, M.; Angel, R. J. J. Phys.
Chem. B 2006, 110, 10322. For selected examples reporting
[Sb2Cl8]2−, see: (c) Willey, G. R.; Palin, J.; Lakin, M. T.; Alcock, N.
W. Transition Met. Chem. 1994, 19, 187. (d) House, D. A.; Browning,
J.; Pipal, J. R.; Robinson, W. T. Inorg. Chim. Acta 1999, 292, 73. In
−
addition, polymeric [SbCl4 ]n has also been reported: (e) Bujak, M.;
Zaleski, J. J. Mol. Struct. 2003, 647, 121.
(32) Razak, I. A.; Usman, A.; Fun, H.-K.; Yamin, B. M.; Kasim, N. A.
M. Acta Crystallogr., Sect. C: Cryst. Struct. Commun. 2002, 58, m225.
(33) Carey, F. A.; Sundberg, R. J. Advanced Organic Chemistry, 5th
ed., Part B: Reactions and Synthesis, Springer, 2007; p 291.
(34) Although that designation does not include the solvent,
CH3NO2 does play a unique role as other solvents failed. For another
case of this phenomenon, see: Dryzhakov, M.; Hellal, M.; Wolf, E.;
Falk, F. C.; Moran, J. J. Am. Chem. Soc. 2015, 137, 9555.
(35) For the full range of studies of different complexes, conditions,
control experiments, and yields, see the Supporting Information.
(36) Yamamoto, H.; Futatsugi, K. Angew. Chem., Int. Ed. 2005, 44,
1924.
(37) As aptly noted by a reviewer, the suspected hydrolysis of the
year-old reagent in the solid phase may not be equivalent to hydrolysis
of the reagent in solution. Indeed, different hydrolyzed products may
result, and therefore, nonfully equivalent reactivity may result. The fact
that long-term stored reagent is effective, however, is an important
finding.
(38) Attempts to directly crystallize 12, 13, and 14 from the reaction
in a fashion similar to that for 5 and 6 led only to amorphous powders
with 1,2-dichloroethane or 1,1,2,2-tetrachloroethane as solvents.
Recrystallizations of the isolated powders of 12, 13, and 14 in an
attempt to obtain their hydrolyzed products failed with all tested
solvents (CH3NO2, CH3CN, CHCl3, CCl4, C6H6, C6H5Cl, DMSO,
DMF, THF, Et2O, EtOAc, MeOH, EtOH, ethylene glycol, dioxane,
and acetone). No signals were observed by IR (ATR and film) for the
five synthesized reagents. MS showed the presence of dppeO2 for 5
and 6, while a mixture of dppe, dppeO, and dppeO2 was detected for
12, 13 and 14. NMR studies in CDCl3, CD2Cl2, and CD3NO2 clearly
showed the hydrolyzed species for 5 and 6. By contrast, 12, 13, and 14
were insoluble in CDCl3.
I
DOI: 10.1021/jacs.6b12653
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX