Scalable Enantiomeric Separation of Dialkyl-Arylphosphine Oxides Based on Host–Guest Complexation with TADDOL-Derivatives, and their Recovery

Article abstract of DOI:10.1002/ejoc.202000035

Several dialkyl-arylphosphine oxides were prepared, and the enantioseparation of the corresponding racemates was elaborated with host–guest complexation using TADDOL-derivatives. The crystallization conditions were optimized and two separate crystallization methods, one in organic solvent, and the other in water, were found to yield five examples of phosphine oxides with enantiomeric excess values higher than 94 %. A gram scale resolution was performed, and both enantiomers of the methyl-phenyl-propyl-phosphine oxide were separated with (R,R)- or (S,S)-spiro-TADDOL. The intermolecular interactions responsible for the enantiomeric recognition between the chiral host and guest molecules were investigated by single-crystal X-ray diffractional structural determinations. The similarities in the structural patterns of a few diastereomeric crystals were checked by powder X-ray diffraction, as well. Organic solvent nanofiltration (OSN) was used as a scalable technique for the decomposition of the corresponding phosphine oxide–spiro-TADDOL molecular complexes, and for the recovery of the phosphine oxide enantiomers and resolving agents.

Full text of DOI:10.1002/ejoc.202000035

A Journal of
Accepted Article
Title: Scalable Enantiomeric Separation of Dialkyl-arylphosphine
oxides based on Host-Guest Complexation with TADDOL-
derivatives, and their Recovery
Authors: Bence Varga, Réka Herbay, György Székely, Tamás
Holczbauer, János Madarász, Béla Mátravölgyi, Elemér
Fogassy, György Keglevich, and Peter Bagi
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Eur. J. Org. Chem. 10.1002/ejoc.202000035
Link to VoR: http://dx.doi.org/10.1002/ejoc.202000035
Supported by

European Journal of Organic Chemistry
10.1002/ejoc.202000035
FULL PAPER
Scalable Enantiomeric Separation of Dialkyl-arylphosphine
oxides based on Host-Guest Complexation with TADDOL-
derivatives, and their Recovery
[a]
[a]
[b,c]
[d]
[e]
Bence Varga, Réka Herbay, György Székely, Tamás Holczbauer, János Madarász, Béla
[a]
[a]
[a]
[a]
Mátravölgyi, Elemér Fogassy, György Keglevich and Péter Bagi*
[
a]
Bence Varga, Réka Herbay, Dr. Béla Mátravölgyi, Prof. Dr. Elemér Fogassy, Prof. Dr. György Keglevich, Dr. Péter Bagi
Department of Organic Chemistry and Technology
Budapest University of Technology and Economics
Műegyetem rkp. 3., H-1111 Budapest, Hungary
E-mail: pbagi@mail.bme.hu
[b]
[c]
[d]
[e]
Dr. György Székely
Advanced Membranes and Porous Materials Center, Physical Science and Engineering Division (PSE)
King Abdullah University of Science and Technology (KAUST)
Thuwal, 23955-6900, Saudi Arabia
Dr. György Székely
Department of Chemical Engineering and Analytical Science
The University of Manchester
The Mill, Sackville Street, Manchester, M1 3BB, United Kingdom
Dr. Tamás Holczbauer
Chemical Crystallography Research Laboratory and Institute of Organic Chemistry
Research Centre for Natural Sciences
Magyar tudósok körútja 2., H-1519 Budapest, Hungary
Dr. János Madarász
Department of Inorganic and Analytical Chemistry
Budapest University of Technology and Economics
Szent Gellért tér 4., H-1111 Budapest, Hungary
Supporting information for this article is given via a link at the end of the document.
Abstract: Several dialkyl-arylphosphine oxides were prepared, and
the enantioseparation of the corresponding racemates was
elaborated with host-guest complexation using TADDOL-derivatives.
The crystallization conditions were optimized and two separate
crystallization methods, one in organic solvent, and the other in
water, were found to yield five examples of phosphine oxides with
organocatalysts is constantly increasing.[3] Thus, the preparation
of novel chiral organophosphorus scaffolds is still of great
interest.
Among the methods developed for the preparation of P-
stereogenic compounds, there is not a universal one, and every
method has its own scope and limitations.[1,4] In these methods,
the P-chiral compounds are usually prepared as bench stable
enantiomeric excess values higher than 94%.
A gram scale
resolution was performed, and both enantiomers of the methyl-
phenyl-propylphosphine oxide were separated with (R,R)- or (S,S)-
spiro-TADDOL. The intermolecular interactions responsible for the
enantiomeric recognition between the chiral host and guest
molecules were investigated by single crystal X-ray diffractional
structural determinations. The similarities in the structural patterns of
a few diastereomeric crystals were checked by powder X‐ray
diffraction, as well. Organic solvent nanofiltration (OSN) was used as
a scalable technique for the decomposition of the corresponding
phosphine oxide - spiro-TADDOL molecular complexes, and for the
recovery of the phosphine oxide enantiomers and resolving agents.
phosphine boranes or oxides, and the air sensitive P(III)-
derivatives are liberated prior to the application.[
1a,5]
The
methods developed for the preparation of P-stereogenic
enantiomers include: stereoselective syntheses using chiral
catalysts [6] bi-,[7] or monofunctional auxiliaries, stereoselective
[8]
[
9]
[10]
transformation of >P(O)H species, kinetic or dynamic kinetic
resolutions,[11] and classical resolutions (Fig. 1).[12]
Several strategies require the manipulation with air- and
moisture sensitive reagent(s) or intermediate(s), which makes
these processes less suitable for scale up and industrial use.
Many methods rely on the formation of covalent P-stereogenic
intermediates with the usage of commercially available, but often
expensive chiral auxiliaries.[7b-e] However, the recycling of those
chiral auxiliaries is sometimes neglected in those
technologies,[8a-e,11f] or even involve the loss of the auxiliary as it
Introduction
is transformed into an unreactive derivative that cannot be
The chiral organophosphorus compounds, in particular the ones
bearing a P-stereogenic center, form an interesting class, and
the preparation of P-stereogenic enantiomers became an
emerging area in the last decades. Although many privileged
P-ligands have been synthesized,[2] the number of chiral
phosphines, phosphine oxides or phosphonium salts used as
reused.[11a,11c,11d]
The classical resolution of phosphine boranes or oxides is
another alternative for the preparation of the corresponding P-
stereogenic species in optically active form.
[
1]
1
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.202000035
FULL PAPER
Figure 2. The racemic compounds (1-7) and the resolving agents [(R,R)-8 -
(R,R,R,R)-11] used in this study.
the enantiopure product, and the recycling of the rather
expensive resolving agents are the key challenges of such
resolution procedures.[15c]
The optical resolution of P-stereogenic phosphine oxides may
serve as a representative example of enantioseparations based
on host-guest complexation. The number of successful
resolution methods for these organophosphorus compounds is
limited.[16]
A few commercially available chiral acids (e.g.
camphorsulfonic acid, tartaric acid derivatives, mandelic acid)
were applicable resolving agents, but those were rather
individual examples with small scope, especially for secondary
phosphine oxides.[12b-d,17] BINOL and the TADDOL derivatives
Figure 1. Selected examples showing the main strategies for the preparation
of optically active tertiary phosphine oxides.
were found to be efficient resolving agents for many cyclic and
an acyclic P-stereogenic phosphine oxides.[
12a,b,e,f]
However, the
The optical resolution of phosphine oxides is a bench stable
process, and the potential scalability, robustness and the
recyclability of the resolving agents are the key advantages of
such procedures. One drawback of the classical resolutions is
the 50% limit in theoretical yield, which can be increased by the
recycling and racemization of the undesired enantiomer.[12f,13] In
general, resolution technologies utilize (relatively) inexpensive
and commercially available acidic or basic resolving agents. In
those instances, the ionic interactions are responsible for the
enantiomeric recognition. The recovery of the pure enantiomer
and the resolving agent can be accomplished by acid-base
extractions, which are preferred in the industry.[14] However, the
optical resolution becomes challenging for the non-acidic or non-
basic racemates. For such substrates, the resolving agent
classes capable of the formation of second order interactions
scalable recovery of the phosphine oxide enantiomers and the
resolving agents remained unresolved challenges of such
resolution procedures. Distillation can be used for this
separation, if the boiling point of the separated enantiomer is low,
but column chromatography is used in most of the instances for
the recovery of the enantiopure phosphine oxide and the
resolving agent.[12e] However, the scalability of the column
chromatography is limited. Separation of the enantiopure
products and the resolving agents can be performed using
organic solvent nanofiltration (OSN). OSN is a sustainable,
pressure-driven membrane-based technology that is capable of
distinguishing compounds between 50 and 2000 g mol-1 in
organic media.[18] OSN processes developed for chiral
molecules include enantioseparation,[19] asymmetric membrane
reactor,[20] and continuous asymmetric catalyst recycling.[21]
In our earlier study, one optically active diaryl-alkylphosphine
oxide was prepared, and those results suggested that TADDOL
derivatives may be generally applicable resolving agents for
acyclic phosphine oxides.[12f] Thus, the current study aims the
enantioseparation of a series of dialkyl-arylphosphine oxides
(
e.g. H-bonds, π-π interactions etc.) have great potential.
BINOL- and TADDOL-type compounds meet the above
mentioned criterion, and these resolving agent classes were
used for the preparation of various enantiomers having neutral
character based on host-guest complexations.[15] However,
decomposition of the diastereomers, the recovery of
(
1-7) by resolution based on host-guest complexation. The
model phosphine oxides (1-7) incorporate one phenyl- and one
methyl-group, which are common in most P-stereogenic ligands
or organocatalysts.[
2a,3c]
Moreover, the methyl-group can be
2
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.202000035
FULL PAPER
used for the construction of an ethylene bridge between two
P-chiral centers.[22] Target phosphine oxides (1-7) are
challenging, as the two alkyl groups may hinder the interactions
between the chiral host and the guest molecules. This might be
the reason why the literature lacks an efficient resolution method
for dialkyl-arylphosphine oxides,[12a] and the stereoselective
syntheses have their own scope and limitations.[
7c,d,8c,d,9a,10e,23]
TADDOL derivatives [(R,R)-8 - (R,R,R,R)-11] were chosen for
the resolution of these P-stereogenic dialkyl-arylphosphine
oxides (1-7) (Fig. 2.). The intermolecular interactions between
the chiral host and guest molecules were investigated by single
crystal or powder X‐ray crystallography to gain some insight of
the secondary interactions responsible for the enantiomeric
recognition. The applicability of the OSN was also tested for the
decomposition of the diastereomers, and for the recovery of the
enantiomers and resolving agents.
Scheme 2. Preparation of t-butyl-methyl-phenylphosphine oxide (7).
In order to prepare the t-butyl-methyl-phenylphosphine oxide (7)
phenylphosphonic dichloride (13) was reacted with t-Bu-MgCl
and Me-MgCl in consecutive reactions. The phosphine so
obtained was oxidized by
H
2
O
2
to give t-butyl-methyl-
phenylphosphine oxide (7) in a yield of 67% (Scheme 2). The t-
butyl-methyl-phenylphosphine oxide (7) could not be prepared
by the reaction of methyl methylphenylphosphinate (12a) with
t-Bu-MgCl due to the increased steric bulk of the organometallic
reagent.
Results and Discussion
Preparation of racemic dialkyl-arylphosphine oxides (1-7)
Resolution of methyl-phenyl-propylphosphine oxide (2) with
TADDOL derivatives [(R,R)-8 - (R,R,R,R)-11]
The methyl methylphenylphosphinate (12a) and ethyl
ethylphenylphosphinate (12b) were used as starting materials in
our procedure. The corresponding phosphinates (12) were
reacted with the excess of the Grignard-reagent to give
phosphine oxides (1-6) (Scheme 1). The advantage of this
method is that it directly transforms the phosphinates (12) to the
corresponding phosphine oxides (1-6) without the formation of
the phosphinic chloride intermediates.[11a] The reactions with
ethyl-, propyl- and butyl-magnesium bromides gave the
phosphine oxides (1-4) in yields of 62-88%. However, the
reaction of methyl methylphenylphosphinate (12a) with i-PrMgBr
or c-HexMgBr afforded the corresponding phosphine oxides (5
and 6) with rather diminished yields (36% and 35%,
respectively), which may be attributed to the increased steric
bulk, and the higher basicity of the Grignard-reagents causing
side reactions.
The methyl-phenyl-propylphosphine oxide (2) and spiro-
TADDOL [(R,R)-9] were considered as model compounds, for
our optimization study, as they both have “average” size within
the scope of racemic compounds (1-7) and resolving agents
[(R,R)-8 - (R,R,R,R)-11] used.
In our experiments, the racemic compound (2) and the resolving
agent [(R,R)-9] were dissolved in the corresponding hot solvent,
and the diastereomeric complexes appeared upon cooling, or by
the addition of a co-solvent. In our preliminary studies, the
amount of resolving agent [(R,R)-9] was adjusted to perform the
resolutions according to the half-equivalent method. The
temperature of the crystallization was set to 25°C, and the
crystallization time was 3h. The crystalline intermediates
separated by filtration were purified by additional
crystallization(s).
A series of solvents and solvent mixtures were tested (See
Supporting Information for the complete list of solvents).
Interestingly, the enantiomeric selectivity of the spiro-TADDOL
[
(R,R)-9] was rather low in alcohols, despite the fact that these
solvents were the most suitable ones in our previous studies
Table 1, Entry 1).[12e,f] The results were improved when the
diastereomers were precipitated by the addition of alkanes
hexane or heptane). A mixture of toluene and heptane was one
(
(
of the most suitable medium affording the (R)-methyl-phenyl-
propylphosphine oxide [(R)-2] with an ee of 96%, and in a yield
of 25% (Table 1, Entry 2). Other solvent mixtures, such as ethyl
acetate and heptane or cymene and heptane led to low yield
and ee values (Table 1, Entry 3; Supporting Information). A
series of “green solvents”, such as cymene, cyrene, cyclopentyl
methyl ether, 2-Me-THF and ethyl lactate were also tested
without success. The work of Matsumoto et. al.[24] and Tsunoda
et. al.[25] prompted us to attempt the precipitation-mediated
resolution in a mixture of methanol and water. To our delight, the
Scheme 1. Preparation of racemic phosphine oxides (1-6).
(
R)-methyl-phenyl-propylphosphine oxide [(R)-2] was obtained
with an ee of 80%, and in a yield of 59% after the first
crystallization, and these results could be increased by
3
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.202000035
FULL PAPER
Scheme 3. General procedure for the optimization of the resolution of methyl-phenyl-propylphosphine oxide (2) with spiro-TADDOL derivative [(R,R)-9].
Table 1. Resolution of methyl-phenyl-propylphosphine oxide (2) with spiro-TADDOL [(R,R)-9].
Eq. of
(R,R)-9
Diastereomeric
Complex[
Yield[c,f]
(%)
ee[d,f]
(%)
S[e,f]
(-)
Abs.
Config.
Entry
1
Solvents[a]
6×2-PrOH
b]
[g]
[
(64)
(35)
33
(0.22)
0.07
1
0.5
0.5
1
(2)∙spiro-TADDOL
2
(R)
(R)
(R)
(R)
(R)
(R)
20
(56)
(53)
96
(0.30)
0.24
2
3
5×toluene/5×heptane
2×EtOAc/10×heptane
(2)∙spiro-TADDOL
(2)∙spiro-TADDOL
25
(121)
(10)
31
(0.12)
0.20
65
(59)
(80)
98
(0.47)
0.45
4
4×MeOH/16×H
2
O
(2)∙spiro-TADDOL
(2)∙spiro-TADDOL
(2)∙spiro-TADDOL
3
46
5[h]
1
15×H
15×H
O
(79)
(88)
(0.69)
2
2
2
2
(88)
(87)
98
(0.77)
0.65
6[i]
1
O
66
[
a] Mixture of solvents for the crystallization and recrystallizations [mL of solvent/g of resolving agent]. [b] The ratio of phosphine oxide (2) and the resolving agent
1
was determined by H NMR. [c] The yield of the diastereomer was calculated based on the half of the racemic phosphine oxide (2) that is regarded to be 100% for
each enantiomer. [d] Determined by HPLC using
a chiral stationary phase. [e] Resolving capability, also known as the Fogassy parameter
[
S (-) = (Yield [%] /100)×(ee [%]/100)]. [f] The results obtained after the first crystallization are shown in parentheses, while the results obtained after purification(s)
are shown in boldface. [g] Absolute configuration of the phosphine oxide (2) was determined by X‐ray analysis (vide infra). [h] The temperature was 50°C, and the
crystallization time was 48 h. [i] The temperature was 50°C, and the crystallization time was 168 h.
recrystallization to afford (R)-2 with an ee of 98%, and in a yield
of 48%. Our efforts to improve these results with different molar
ratios of spiro-TADDOL [(R,R)-9], ratios of MeOH and water, or
with the applications of various alcohols failed (See Supporting
Information for details). Moreover, a change in the reaction
conditions often involved a change in the composition of the
diastereomer, indicating that the diastereomeric complex was
unstable in that solvent mixture. This fact might be attributed to a
competitive complexation of spiro-TADDOL [(R,R)-9] with
alcohols, which was observed in our previous studies.[12e,26]
Then, the resolution of racemic phosphine oxide 2 was also
attempted in water. In order to avoid some issues caused by the
low solubility, the resolution was accomplished at 50°C, which
gave us the best results after the first crystallization (ee: 88%,
yield: 79%; Table 1, Entry 5). Such suspension-based resolution
procedures can be found in the literature, but these results were
somewhat unexpected, as either the racemic compound (2), or
the resolving agent [(R,R)-9] is only partially soluble in water,
and the reaction mixture was heterogeneous throughout the
resolution process. However, we postulate that the hydrophilic
media forcing the two lipophilic molecules together is
responsible for the host-guest complexation, and consequently
the enantiomeric recognition. With this promising preliminary
result in hand, the reaction conditions, such as ratio of the
racemic compound (2) and the spiro-TADDOL [(R,R)-9], ratio of
the resolving agent [(R,R)-9] and water, temperature and time of
the crystallization were optimized (See Supporting Information).
It was found that crystallization at 50°C for 168 h in the presence
of 1 equivalent of spiro-TADDOL [(R,R)-9] (half-equivalent
method) gave us the best results, and the diastereomeric
complex could be purified by one recrystallization to give the
(R)-methyl-phenyl-propylphosphine oxide [(R)-2] with an ee of
98% and in a yield of 66%.
Having the suitable conditions in hand, we tested the
applicability of other TADDOL-derivatives [(R,R)-8, (R,R)-10 and
(R,R,R,R)-11] for the enantiomeric separation of methyl-phenyl-
propylphosphine oxide (2). In our solvent and parameter
optimization study, the best results (ee of 98%) were obtained in
water, and a high ee (96%) could be obtained in a mixture of
toluene and heptane (Table 1, Entries 2 and 6). Moreover, 2-
propanol or a mixture of ethyl acetate and heptane led also to
satisfactory results (Table 1, Entries 1 and 3).
4
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.202000035
FULL PAPER
Figure 3. Comparison of the resolution of methyl-phenyl-propylphosphine oxide (2) with TADDOL-derivatives [(R,R)-8 - (R,R,R,R)-11] (See Supporting
Information Supplementary Table 4 for full details; ‘no complex’ – no crystalline diastereomer was formed; ‘TADDOL’ – the resolving agent precipitated, but no
host-guest complex was formed).
Different resolving agents [(R,R)-8, (R,R)-10 and (R,R,R,R)-11]
were involved in this study to investigate whether the overall
efficiency of the resolution can be improved by changing the
structure of the resolving agent. Owing to the size-exclusion
nature of OSN, the enlargement of the resolving agents [(R,R)-
obtained in water, in a mixture of toluene-heptane or ethyl
acetate-heptane, and the conditions for crystallization were also
established. In the next step, the substrate scope of this
resolution procedure was investigated in the sphere of various
dialkyl-arylphosphine oxides (1-7). The results are summarized
in Fig. 4.
10 and (R,R,R,R)-11] was explored to increase the molecular
weight gap between the species to be separated, and to have
more ideal separation with OSN (vide infra). The results are
summarized in Fig. 3, which also contains the results obtained
with spiro-TADDOL [(R,R)-9].
Interestingly, the change in the structure of the resolving agent
has an unexpectedly significant impact on the resolution
efficiency. By changing the spiro-scaffold [(R,R)-9] to a dimethyl-
ketal [(R,R)-8] protecting group, the crystalline diastereomeric
complexes could be obtained from toluene-heptane, EtOAc-
heptane and 2-PrOH. However, the yields were low, and the
maximum ee was 87%. Changing the aromatic groups from
phenyl to naphthyl had an even more pronounced effect, namely
no diastereomers were formed with naphthyl-spiro-TADDOL
The experiments indicated that the structure of the racemic
compound (1-7) has also a significant impact on the overall
efficiency of the optical resolution. Using water as the solvent,
the (R)-enantiomer of the methyl-propyl-, ethyl-propyl- and the
butyl-methyl-phenylphosphine oxides (2-4) was prepared with an
ee of 95-98%, and in a yield of 51-66%. Interestingly, the
efficiency of the resolution was significantly lower for these
phosphine oxides (2-4), when the diastereomers were
crystallized from toluene - heptane or ethyl acetate - heptane
mixtures. This trend turned for the ethyl-methyl-, methyl-i-propyl-
and the c-hexyl-methyl-phenylphosphine oxide (1, 5 and 6) as
the enantiomers or the enantiomeric mixture could be prepared
only in the mixture of the organic solvents (ee: 31-99%, yield:
21-56%), whereas the application of water prevented the
diastereomeric complex formation, or led to racemic phosphine
oxide. Overall, the t-butyl-methyl-phenylphosphine oxide (7) was
the only derivative, which could not be separated effectively into
its enantiomers with spiro-TADDOL [(R,R)-9]. Moreover, the
maximal enantiomeric purity reached for ethyl-methyl-
phenylphosphine oxide [(R)-1] was only 31%. These results
might be attributed to steric factors, as the molecular size of the
MeEtPhP(O) (1) might be too small for sufficient interaction
[
(R,R)-10] in most of the instances. Interestingly, water was the
only solvent from which crystalline diastereomers were obtained,
but the ee and efficiency remained low (ee: 9%, S=0.04). The
enantiomeric separation was also rather low with the bis-spiro-
TADDOL [(R,R,R,R)-11], despite the fact that its structure shows
a close resemblance to the structure of spiro-TADDOL [(R,R)-9].
The maximal ee was 29% performing the crystallization of
phosphine oxide (2) with bis-spiro-TADDOL [(R,R,R,R)-11] in a
mixture of ethyl acetate and heptane. Using other solvents,
either racemic phosphine oxide (2) was obtained, or no
diastereomers were formed.
between the host and the guest molecules, which is
a
prerequisite for a good enantiomeric recognition. Whereas, the
increased steric bulk might be the reason for the inefficient
resolution of t-BuMePhP(O) (7).
On the other hand, the enantiomers of phosphine oxides 2-6
could be effectively separated by resolution with spiro-TADDOL
Resolution of dialkyl-arylphosphine oxides (1-7) with spiro-
TADDOL [(R,R)-9] under optimized conditions
In our initial studies, the spiro-TADDOL [(R,R)-9] was selected
as the most suitable resolving agent. The best results were
5
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.202000035
FULL PAPER
Figure 4. Comparison of the optical resolution of phosphine oxides (1-7) with spiro-TADDOL [(R,R)-9] under the optimized conditions.
[
(R,R)-9] regardless of the size or the degree of branching of the
crystallization and recrystallization were combined, and the
enantomeric mixture of (S)-2 was separated from the traces of
(R,R)-spiro-TADDOL [(R,R)-9]. The enantiomeric mixture of (S)-
2 (ee: 71%) so obtained was resolved with (S,S)-spiro-TADDOL
[(S,S)-9] in water, and the (S)-methyl-phenyl-propylphosphine
oxide [(S)-2] could be obtained with an ee of 98% and in a yield
of 45% after one crystallization followed by the decomposition of
the corresponding diastereomer. These results indicate that the
optimized resolution procedures for phosphine oxides 1-6 can
be accomplished on gram scale with similar efficiency. Moreover,
the other enantiomer of the target phosphine oxide (1-6) can be
obtained from the mother liquor using the (S,S) enantiomer of
spiro-TADDOL [(S,S)-9] under similar crystallization conditions.
alkyl chains, and the corresponding enantiomers were obtained
with an ee of 94-99% and in a yield of 21-65%.
The optical resolutions with spiro-TADDOL [(R,R)-9] afforded the
corresponding phosphine oxides enantiomers with (R) absolute
configuration [(R)-1 - (R)-5] in most of the instances regardless
of the crystallization parameters. Interestingly, the (S)
enantiomer could be prepared exclusively for c-hexyl-methyl-
phenylphosphine oxide [(S)-6]. Moreover, the solvent had a
great influence on the selection of the ethyl-phenyl-
propylphosphine (3) enantiomers incorporated into the
diastereomeric complexes. The (R) enantiomer [(R)-3] could be
prepared with spiro-TADDOL [(R,R)-9] in water, whereas
enantiomeric mixtures containing the (S)-3 enantiomer in excess
could be prepared in toluene - heptane and ethyl acetate -
heptane mixtures. The absolute configuration of the guest
phosphine oxides (1-7) was determined by comparing the
specific optical rotations with literature data. The absolute
configuration of methyl-phenyl-propylphosphine oxide and ethyl-
phenyl-propylphosphine oxide (2 and 3) was elucidated by X-ray
crystallography (vide infra).
Single crystal and powder X‐ray analysis of a few diastereomeric
complexes
Crystallinity of the diastereomers prepared from the
corresponding phosphine oxide (1-7) and spiro-TADDOL
[(R,R)-9] was checked by powder X‐ray diffraction. Additionally,
single crystals were successfully grown from the diastereomers
containing spiro-TADDOL [(R,R)-9] and (R)-methyl-phenyl-
Complete gram-scale optical resolution process for the
propylphosphine
oxide
[(R)-2]
or
(S)-ethyl-phenyl-
preparation
of
both
enantiomers
of
methyl-phenyl-
propylphosphine oxide [(S)-3] allowing us to investigate the
second order interactions responsible for the enantiomeric
recognition. Moreover, the direct comparison of the two crystals
gave some insight how the crystal structures and the
interactions may alter when phosphine oxides [(R)-2 and (S)-3]
with different structure and absolute configuration are present.
The single crystal X-ray analysis confirmed that the
diastereomers contained the spiro-TADDOL [(R,R)-9] and the
corresponding phosphine oxide [(R)-2 or (S)-3] enantiomer in a
1:1 ratio, which was in accordance with the NMR studies.
ORTEP style molecular structure diagram can be found in Fig. 5.
Similar unit cell parameters were measured for the two selected
crystals indicating similarities of the (R,R)-spiro-TADDOL
[(R,R)-9], the (R)-2 or (S)-3 guest molecules and hydrogen bond
propylphosphine oxide [(R)- and (S)-2] with (R,R)- and (S,S)-
spiro-TADDOL [(R,R)- and (S,S)-9]
The optimization experiments were conducted on a millimolar
scale, and as the next step, the resolution of methyl-phenyl-
propylphosphine oxide (2) with (R,R)-spiro-TADDOL [(R,R)-9]
was also accomplished on a gram scale. Besides demonstrating
the scalability, the other aim of ours was to prepare both
enantiomers of methyl-phenyl-propylphosphine oxide [(R)- or
(
S)-2] (Scheme 4). The racemic methyl-phenyl-propylphosphine
oxide (2) was resolved with (R,R)-spiro-TADDOL [(R,R)-9] in
water under the optimized conditions. After the purification and
decomposition
of
the
diastereomeric
complex
[
(R)-2]∙(spiro-TADDOL)
2
,
the (R)-enantiomer [(R)-2] was
I
systems within. The Cell Similarity Index is ( ) 0.10901.
prepared with an ee of 98% and in a yield of 72%. To prepare
the other enantiomer (S)-2, the mother liquors of the
6
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.202000035
FULL PAPER
Scheme 4. Gram-scale preparation of both enantiomers of methyl-phenyl-propylphosphine oxide [(R)- or (S)-2] on with (R,R)- and (S,S)-spiro-TADDOL [(R,R)-9
and (S,S)-9].
The density is 1.244 for the [(R)-2]∙spiro-TADDOL and 1.210 for
the [(S)-3]∙spiro-TADDOL diastereomer. The Kitaigorodsky
Packing Index (K.P.I)[27] values were also similar (67.9% and
The interactions between the (R)-2 or (S)-3 phosphine oxides
and the spiro-TADDOL [(R,R)-9] were investigated with Hirshfeld
surface by the Crystal Explorer program (See Supporting
Information for details). The Hishfeld surface fingerprints
indicated that the short and strong intermolecular interactions
(i.e. hydrogen bonds) are similar, and the main differences could
be found in the weaker interactions. More longer and weaker
contacts can be found in the crystal structure of the
[(S)-3]∙spiro-TADDOL diastereomer than in the crystal lattice of
[(R)-2]∙spiro-TADDOL. Moreover, the atom-type Hirshfeld
interpretations also showed significant dissimilarities within the
weak intermolecular interactions between specific atomic types.
The Hirshfeld fingerprint interpretation shows the difference in
atomic position and distances, but it is unable to characterize the
energies associated with these contacts. Thus, some DFT
calculations were made by the Crystal Explorer12 program on
the B3LYP/6-31G(d,p) level of theory for better understanding
the interactions between the host and guest molecules.
66.4% for the [(R)-2]∙spiro-TADDOL and [(S)-3]∙spiro-TADDOL,
respectively). However, despite these similar density and K.P.I.
values, the two crystals are not isostructural, as the guest
phosphine oxides [(R)-2 or (S)-3] have different places in the
unit cell. Moreover, two small voids were found in the crystal
structure of the [(S)-3]∙spiro-TADDOL complex, which were not
present in the crystal structure of the [(R)-2]∙spiro-TADDOL
diastereomer (Fig. 5.). These voids have a volume of ca. 2 x 33
3
Å , which size is not enough for the incorporation of smaller
3
solvent molecules (e.g. water needed 40 Å ).
Intriguingly, the absolute configuration of the phosphine oxide
guest molecules [(R)-2 or (S)-3] is the opposite, despite the
structural similarity and the same configuration of the host spiro-
TADDOL [(R,R)-9] molecule.
The crystal structures and the DFT calculations revealed several
contacts. The most relevant atomic distances, as well as the
results of the calculations are summarized in Fig. 6. The two OH
groups of the spiro-TADDOL [(R,R)-9] form an intramolecular H-
bond, which is characteristic to the TADDOL-type molecules.[
Thus, one hydroxyl group of (R,R)-9 is capable of the formation
of an intermolecular H-bridge with the corresponding phosphine
oxide enantiomer [(R)-2 or (S)-3]. The atomic distances indicate
12f]
(
2.686 Å and 2.634 Å) that this H-bridge is the strongest
intermolecular contact. CH…π interactions can also be observed,
but there are differences in the two crystal lattices. In case of the
(
S)-ethyl-phenyl-propylphosphine oxide [(S)-3],
a
2
CH …π
interaction is present, whereas a T-spape π…π contact can be
observed between the aryl groups of the methyl-phenyl-
propylphosphine oxide [(R)-2] and the spiro-TADDOL [(R,R)-9].
The interactions formed between the two phosphine oxide
Figure 5. The crystallographically independent molecules in the asymmetric
unit of the crystal a.) [(S)-3]∙spiro-TADDOL and b.) [(R)-2]∙spiro-TADDOL
without the atomic labelling. Displacement ellipsoids are drawn at the 30%
probability level for clarity. The observable voids in the crystal lattices c.) [(S)-
3]∙spiro-TADDOL and d.) [(R)-2]∙spiro-TADDOL. Two small voids in the crystal
lattice are displayed in orange colour in the case of [(S)-3]∙spiro-TADDOL.
7
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.202000035
FULL PAPER
Figure 6. A few representative atomic distances and the strongest total energies among the neighbouring molecules for the [(S)-3]∙spiro-TADDOL (a) and
(R)-2]∙spiro-TADDOL (b). The yellow highlight represents the selected phosphine oxide [(R)-2 or (S)-3] for the calculation of interaction energies.
[
molecules are also different. There is a CH… π contact between
the two MePrPhP(O) molecules [(R)-2], whereas a CH…O=P
contact can be found between the two (S)-ethyl-phenyl-
propylphosphine oxides [(S)-3].
The DFT results also confirmed that the strongest force is a
coulomb interaction between the hydrogen bridge acceptor
phosphine oxides [(R)-2 or (S)-3] and the hydrogen bond donor
orientation of sample crystal set at the measurement of powder
profile. Despite the latter observation, an indexing of the room
temperature powder pattern shows very similar, but increased
unit cell parameters to those obtained by the single crystal
measurement (See Supporting Information, Supplementary
Table 13 for details).
2
Fig. 7c also shows that the [(R)-2]∙spiro-TADDOL diastereomer
(
R,R)-spiro-TADDOL [(R,R)-9] in both crystal structures. The
prepared from aqueous media exhibited a powder XRD profile
quite different from that prepared from a mixture of toluene and
heptane (Fig. 7a and b). There are significant differences in the
positions of the diffraction peaks referring to different dominating
crystalline structures. The reason for the different powder XRD
profiles is the different stoichiometric molar ratio of the
(R)-methyl-phenyl-propylphosphine oxide [(R)-2] and the
(R,R)-spiro-TADDOL [(R,R)-9], which also results in different
orientations and interactions among the host and the guest
molecules. The enantiopure (R)-ethyl-propyl- or the (R)-butyl-
methyl-phenylphosphine oxides [(R)-3 or (R)-4] could also be
second strongest interaction is within two neighbouring
phosphine oxide moieties [(R)-2 or (S)-3] caused by the
polarization of the P=O moiety. The third strongest interaction is
observed between the other spiro-TADDOL molecule and the
corresponding methyl-propyl- and ethyl-propyl-phenyl-phosphine
oxide [(R)-2 or (S)-3]. These weak London dispersion forces are
caused by the high overlapping surfaces of the (R)-2 or (S)-3
guest and the (R,R)-9 host molecule. However, the calculations
revealed that the energetics of these interactions are rather
different in the two crystal lattices (Fig. 6.). Thus, despite the
structural similarities of the two phosphine oxides [(R)-2 or (S)-3], prepared with (R,R)-spiro-TADDOL [(R,R)-9] in water. Thus, the
significant differences were found both in the nature and the
strength of these second order interactions, which may be an
underlying reason for the different enantiopreference in case in
the two crystal structures.
host-guest complexes of [(R)-3]∙spiro-TADDOL
TADDOL were also subjected to powder XRD analysis.
Surprisingly, the X-ray diffraction profiles of the
[(R)-2]∙spiro-TADDOL and [(R)-4]∙spiro-TADDOL
2
or [(R)-4]∙spiro-
2
2
2
We were unable to grow X-ray quality crystals from the
diastereomers prepared in aqueous media. Thus, these
crystalline samples were subjected to powder X-ray
measurements along with a few crystals prepared in organic
solvents in order to seek for structural similarities between the
diastereomers prepared in aqueous media showed similar
pattern (Fig. 7c and e). Despite the different guest molecules
[(R)-2 or (R)-4], these two diastereomeric co-crystals may be
considered isomorphic, or at least isostructural ones, indicating
similar mode and spatial arrangements of binding and
interactions between the given phosphine oxides [(R)-2 or (R)-4]
and the (R,R)-spiro-TADDOL [(R,R)-9]. Interestingly, these two
diastereomeric
molecular
complexes.
First,
the
[
(R)-2]∙spiro-TADDOL diastereomer was investigated, and the
powder XRD-pattern measured at room temperature has been
compared with that of generated from the atomic coordinates
obtained by the single crystal structural determination at
temperature of 148 K (Fig. 7a and b). The series of diffraction
peak positions are in good agreement considering the
experimental profile and the corresponding simulated pattern,
which indicates that the measured single crystal was
representative for the solid state obtained by optical resolution.
Slight differences are arising from the temperature difference of
XRD-measurements resulting in extension or contraction of the
crystal lattice as a whole, and its interplanar lattice distances, as
well. There is a considerable alteration in the relative intensity of
diffraction peaks, probably because of a kind of preferred
diffraction
[(R)-3]∙spiro-TADDOL
profiles
were
different
from
that
of
2
(Fig. 7d), indicating that changing a
methyl group to ethyl group may significantly alter the
interactions within the diastereomeric complex resulting in a
different crystal structure. It is noteworthy, that an increased
peak broadening also occurred in all the samples of aqueous
origin, probably because of their lowered crystallinity or
microcrystalline feature.
8
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.202000035
FULL PAPER
Figure 7. Comparison of powder XRD profiles of
diastereomeric complexes.
a few selected
Figure 8. Evaluation of the recovery of (R)-methyl-phenyl-propylphosphine
oxide [(R)-2] and (R,R)-spiro-TADDOL [(R,R)-9] from the corresponding
diastereomer by flash chromatography (The bars show the recovery of (R)-2)
and (R,R)-9, while the area-chart show the change in eluent composition).
Comparison of the column chromatographic and nanofiltration-
enabled decomposition of diastereomeric complexes and
recovery of phosphine oxide (1-7) enantiomers and resolving
agents [(R,R)-9 - (R,R,R,R)-11]
Then, the (R)-2 phosphine oxide enantiomer could be eluted
from the column applying a DCM-MeOH or EtOAc-2-PrOH
gradient. In this manner no mixed fractions containing both (R)-2
and (R,R)-9 were obtained. The recovery of (R)-methyl-phenyl-
propylphosphine oxide [(R)-2] and (R,R)-spiro-TADDOL [(R,R)-
9] was above 96% in all cases (Fig. 8.). Considering the two
column chromatographic methods, 370 or 310 mL of solvent
was used for the decomposition of 600 mg of diastereomers, as
well as for the recovery of ca. 500 mg of the resolving agent
[(R,R)-9] and 100 mg of the (R)-methyl-phenyl-propylphosphine
oxide [(R)-2]. As 6 empty fractions were collected in both cases,
As the volatility of the phosphine oxides (1-7) is low, column
chromatography is used for the decomposition of the
corresponding diastereomeric complexes and for the recovery of
the (R)-methyl-phenyl-propylphosphine oxide [(R)-2] and
(
R,R)-spiro-TADDOL [(R,R)-9]. Several solvents were tested for
the chromatographic separation (See Supporting Information for
the complete list). Our goal was to find a solvent or solvent
system, which allows complete separation of the two
compounds [(R)-2 and (R,R)-9] without obtaining mixed fractions. one may argue that the solvent consumption can be lowered by
The DCM to DCM-MeOH mixture was our choice in our previous
ca. 50 mL using a more optimized gradient.
Organic solvent nanofiltration could serve as a scalable
studies,[
12e,f]
and this solvent system seemed suitable based on
the TLC analysis. However, we intended to choose other
solvents, which met more criteria of green chemistry and
suitability for industrial use.[28] Thus, we chose a EtOAc to
EtOAc-2-PrOH gradient elution based on our TLC tests.
For comparison, the same sample was separated under same
flash chromatographic conditions (stationary phase; flow rate;
fraction volume etc.). To our delight both solvent systems (DCM
to DCM-MeOH or EtOAc to EtOAc-2-PrOH) were appropriate for
separation with flash chromatography. The advantages of these
solvents is that the resolving agent [(R,R)-9] could be removed
completely under isocratic conditions using pure DCM or EtOAc.
technique for the decomposition of the diastereomeric
complexes, and the recovery of the phosphine oxide
enantiomers. First, the use of organic solvent nanofiltration was
explored using NP030P membrane by Nadir and methanol as
the solvent. The molecular weight cut-off (MWCO) curve of the
membrane was determined for the dialkyl-arylphosphine oxides
(1-7), and the resolving agents incorporating the spiro-TADDOL
scaffold [(R,R)-9 - (R,R,R,R)-11] (Fig. 9). In order to maximize
the rejection of the resolving agents [(R,R)-9 - (R,R,R,R)-11], the
membrane was treated with branched polyethylene glycol (PEG)
having 2,600 g mol-1 molecular weight.
9
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.202000035
FULL PAPER
Figure 9. Molecular weight cut-off (MWCO) curve for Nadir NP030P
membrane with and without treatment with branched PEG in methanol at 40
bar pressure.
The PEG treatment resulted in tighter membranes with
increased rejections (Fig. 9). This change was beneficial for the
resolving agents [(R,R)-9 - (R,R,R,R)-11], which are to be
retained by the membrane during diafiltration. The most
prominent increase in rejection from 86.2±2.7% to 100% was
obtained for the best resolving agent, spiro-TADDOL [(R,R)-9].
Fig 9. also shows that the rejection values for the dialkyl-
arylphoshine oxides (1-7) were in the range of
1
3.8±2.0%
membrane, and these values slightly increased to
1.4±1.5% – 43.9±1.8% after the PEG treatment. The low
rejection values of the phosphine oxide (1-7) compared to
complete rejection of the resolving agents [(R,R)-9
R,R,R,R)-11] were sufficient for nanofiltration-based
separation strategy.
Based on the rejection results, the [(R)-2]∙(spiro-TADDOL)
– 34.0±2.6% with the original Nadir NP030P
Figure 10. Concentration profiles during the diafiltration of the [(R)-2]∙(spiro-
TADDOL) diastereomer and the purity of the resolving agent in the retentate
2
stream (a); and the effect of (R,R)-9 rejection (see values in brackets) on the
purity of (R)-2 in the permeate stream caused by the loss of (R,R)-9 from the
retentate (b).
2
-
(
a
Conclusion
2
molecular complex was selected for the decomposition of the
diastereomer, as well as for the recovery of (R)-methyl-phenyl-
propylphosphine oxide [(R)-2] and (R,R)-spiro-TADDOL
In this paper, the resolution of a series of phosphine oxides was
investigated with TADDOL-derivatives [(R,R)-8 - (R,R,R,R)-11].
It was demonstrated, that this resolving agent class, especially
[
(R,R)-9] using the PEG-treated NP030P membrane. The
the
spiro-TADDOL
[(R,R)-9]
is
applicable
for
the
simulated diafiltration separation (lines) and the experimental
data (symbols) for processing the diastereomeric complex
enantioseparation of these challenging substrates, which have
limited capabilities to form second order interactions due to the
presence of two alkyl groups. The results showed that there is
an ideal molecular size for the sufficient enantioseparation. The
MeEtPhP(O) (1) have small molecular size, whereas the
increased steric bulk might be the reason for the inefficient
resolution of t-BuMePhP(O) (7). The enantiomers of other five
dialkyl-arylphosphine oxides (2-6) were prepared with
enantiomeric excess values greater than 94%. During the
optimization of the crystallization conditions, two separate
crystallization methods were developed, one utilized organic
solvent and the other one used water. This is one of the handful
successful examples for performing enantioseparations with
TADDOL-derivatives in aqueous media. A gram-scale resolution
was performed demonstrating the scalability of this method, and
both enantiomers of the methyl-phenyl-propylphosphine oxide
[
(R)-2]∙(spiro-TADDOL)
phosphine oxide was purged out of the system and virtually
00% purity of the spiro-TADDOL [(R,R)-9] was achieved within
less than 8 diavolumes (Fig. 10a). In this experiment, ca. 1.3 g
of (R)-2·(spiro-TADDOL) diastereomeric complex was
2
are shown in Fig. 10. The (R)-2
1
2
decomposed. As nearly complete separation of (R)-2 phosphine
oxide and spiro-TADDOL [(R,R)-9] was achieved in 6-7
diavolumes, the solvent consumption was in the range of 470-
550 mL of solvent per gram of diastereomer. The robustness of
the purification methodology was investigated through varying
the theoretical rejection of (R,R)-9 from 99.5% to 100% (Fig.
10b). As the rejection decreases, the purity of (R)-methyl-
phenyl-propylphosphine oxide [(R)-2] significantly decreases
due to the fact that the concentration of spiro-TADDOL [(R,R)-9]
in the feed stream is considerably higher. Even a small leakage
of the resolving agent through the membrane (i.e. incomplete
rejection) results in a significant contamination of (R)-2 in the
permeate stream. On the contrary, the increase in the loss of
(
(
2) were prepared with (R,R)- or (S,S)-spiro-TADDOL [(R,R)- or
S,S)-9]. The structure of a few diastereomeric complexes were
investigated by single crystal or powder X‐ray crystallography,
and the main secondary interactions responsible for the
enantiomeric recognition were also identified. Nadir NP030P
membrane was successfully employed in a diafiltration process
for the gram-scale separation of the enantiopure (R)-methyl-
(
R,R)-9 as a function of its rejection (100%  99.5%) is less
pronounced (0%  5%). These results show the importance of
obtaining complete rejection of the resolving agents, which is
easier to achieve by having large molecular weights.
1
0
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.202000035
FULL PAPER
phenyl-propylphosphine oxide [(R)-2], and the TADDOL-based
resolving agent [(R,R)-9]. The use of a branched PEG for the
preconditioning of the membrane was found to be a useful tool
to tighten the membrane and increase solute rejection. The
presented OSN methodology could be further exploited for the
recovery of other resolving agents, and the parallel isolation of
enantiopure products.
water at 50°C for 1 week to afford 0.11 g (66%) of (R)-2·(spiro-
TADDOL) with a de of 98% (Table 1, Entry 6). The (R)-methyl-phenyl-
2
propylphosphine oxide [(R)-2] was recovered from the diastereomer by
flash column chromatography (silica gel, gradient elution, ethyl acetate to
ethyl acetate – i-propyl-alcohol 80:20) to give 0.015 g (60%) of (R)-
methyl-phenyl-propylphosphine oxide [(R)-2] with an ee of 98%.
Separation of the phosphine oxides (1-7) and the resolving agents
[
(R,R)-9 - (R,R,R,R)-11] organic solvent nanofiltration
Experimental Section
The feed solution for the molecular weight cut-off (MWCO) determination
comprised a single-solute phosphine oxide (1-7) or resolving agent
[
(R,R)-9 - (R,R,R,R)-11] at 0.1 g L-1 concentration in methanol at 40 bar.
Materials and apparatus
2
The Nadir NP030P membrane with an area (A) of 52.8 cm was used
with and without treatment with a 1 g L-1 branched PEG having a
molecular weight of 2,600 g mol-1 for 24 h. The membrane were
The details of the chemicals, the instruments used in this study, as well
as the synthetic procedures for the preparation of phosphine oxides (1-7)
and resolving agents [(R,R)-8 - (R,R,R,R)-11] can be found in the
Supporting Information.
conditioned in methanol at 40 bar for 24 hours in
nanofiltration rig (Supplementary Fig. 1.) as described earlier.
a cross-flow
[
21]
A
recirculation gear pump was set at 1 L min-1 to ensured homogeneous
solute concentration and mitigate concentration polarization in the
retentate loop. The mathematical framework for the diafiltration modeling
of the concentration profiles and purities were previously reported.[29] The
Representative resolution procedures
Resolution of methyl-phenyl-propylphosphine oxide (2) with spiro-
TADDOL [(R,R)-9] using the crystallization method (Method I.)
solute rejections (R) were determined as the ratio of the permeate (C
and retentate (C ) concentrations (Eq. 1):
p
)
r
0.10 g (0.55 mmol) of Racemic methyl-phenyl-propylphosphie oxide (2)
Eq. 1
and 0.28 g (0.55 mmol) of spiro-TADDOL [(R,R)-9] were dissolved in 1.7
mL of hot i-propyl-alcohol. Colorless crystalline diastereomeric complex
The reported results are averages of two independently performed
experiments. The permeance (P) was determined as the volume of
permeate divided by the applied pressure (∆P), membrane area (A) and
time (t) (Eq. 2):
of (R)-2·(spiro-TADDOL)
2
appeared after cooling the mixture to 25°C.
After standing at 25°C for 3 hours, the crystals were separated by
filtration to give 0.21 g (64%) of (R)-2·(spiro-TADDOL)
The diastereomeric complex (R)-2·(spiro-TADDOL)
2
with a de of 35%.
was purified by
2
recrystallization according to the procedure described above from 1.7 ml
of i-propyl-alcohol to afford 0.066 g (20%) of the (R)-2·(spiro-TADDOL)
with de of 33% (Table 1, Entry 1). The (R)-methyl-phenyl-
propylphosphine oxide [(R)-2] was recovered from the diastereomer by
flash column chromatography (silica gel, gradient elution, ethyl acetate to
ethyl acetate – i-propyl-alcohol 80:20) to give 0.009 g (18 %) of (R)-
methyl-phenyl-propylphosphine oxide [(R)-2] with an ee of 33%.
Eq. 2
2
a
During the diafiltration 1.27 g of diastereomer complex (R)-2·(spiro-
TADDOL) was loaded onto the membrane in 100 mL of methanol. The
2
filtration was performed at 40 bar, and fresh solvent was continuously
added in order to compensate the permeate volume leaving the system,
keeping the system volume constant at 100 mL. The filtration was
performed up to 10 diavolumes (D, Eq. 3), which is defined as the ratio of
added solvent volume (Vadd) and system volume (Vsystem):
Resolution of methyl-phenyl-propylphosphine oxide (2) with spiro-
TADDOL [(R,R)-9] using the precipitation method (Method II.)
-
1
D = Vadd
V
system
Eq. 3
0
(
0
.051 g (0.28 mmol) of Racemic methyl-phenyl-propylphosphine oxide
2) and 0.071 g (0.14 mmol) of spiro-TADDOL [(R,R)-9] were dissolved in
.35 mL of hot toluene, and then 0.35 mL of heptane was added.
X-Ray measurements
Colorless crystalline diastereomeric complex of (R)-2·(spiro-TADDOL)
appeared immediately. After standing at 25°C for 3 hours, the crystals
were separated by filtration to give 0.054 g (56%) of (R)-2·(spiro-
TADDOL) with a de of 53%. The diastereomeric complex (R)-2·(spiro-
TADDOL) was purified by two recrystallizations according to the
procedure described above from a mixture of 0.35 mL of toluene and
Summary of crystallographic data, data collections, structure
determination and refinement for [(R)-2]∙spiro-TADDOL and [(S)-3]∙spiro-
TADDOL is listed in Supplementary Table 9. All crystallographic data
(including structure factors) for the crystal structure of [(R)-2]∙spiro-
TADDOL and [(S)-3]∙spiro-TADDOL has also been deposited with the
Cambridge Crystallographic Data Centre as supplementary publication
CCDC 1957408 and 1957409.
0.35 mL of heptane to afford 0.024 g (25%) of the (R)-2·(spiro-TADDOL)
with de of 96% (Table 1, Entry 2). The (R)-methyl-phenyl-
a
propylphosphine oxide [(R)-2] was recovered from the diastereomer by
flash column chromatography (silica gel, gradient elution, ethyl acetate to
ethyl acetate – i-propyl-alcohol 80:20) to give 0.0057 g (22%) of (R)-
methyl-phenyl-propylphosphine oxide [(R)-2] with an ee of 96%.
Acknowledgements
This work was supported by the National Research,
Development and Innovation Office - NKFIH (Grant No. OTKA
PD 116096). Tamás Holczbauer is grateful for the support of the
National Research, Development and Innovation Office-NKFIH
Resolution of methyl-phenyl-propylphosphine oxide (2) with spiro-
TADDOL [(R,R)-9] using the suspension method (Method III.)
A
suspension of 0.052
propylphosphine oxide (2) and 0.14 g (0.28 mmol) spiro-TADDOL [(R,R)-
] was stirred in 2.2 mL of water at 50°C for 1 week. The crystals were
then separated by filtration to give 0.15 (88%) of
g (0.28 mmol) of racemic methyl-phenyl-
(
Grant No. OTKA PD 128504) and the János Bolyai Research
9
Scholarship of the HAS. György Székely acknowledges the
financial support from KAUST.
g
(
(
R)-2·(spiro-TADDOL)
2
with a de of 87%. The diastereomeric complex
was purified by stirring the crystals in 2.2 mL of
R)-2·(spiro-TADDOL)
2
1
1
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.202000035
FULL PAPER
b) K. V. Rajendran, L. Kennedy, D. G. Gilheany, Eur. J. Org. Chem.
Keywords: Chiral resolution • Acyclic phosphine oxides •
Supramolecular investigation • Organic solvent nanofiltration
2010, 5642–5649; c) K. Nikitin, K. V Rajendran, H. Müller-Bunz, D. G.
Gilheany, Angew. Chem. Int. Ed. 2014, 53, 1906–1909; d) K. V.
Rajendran, K. V. Nikitin, D. G. Gilheany, J. Am. Chem. Soc. 2015, 137,
[1]
[2]
[3]
a) A. Grabulosa, P-Stereogenic Ligands in Enantioselective Catalysis,
The Royal Society Of Chemistry, Cambridge, 2010; b) M. Dutartre, J.
Bayardon, S. Jugé, Chem. Soc. Rev. 2016, 45, 5771–5794.
9
375–9381; e) K. M.-H. Lim, T. Hayashi, J. Am. Chem. Soc. 2017, 139,
122–8125; f) A. Ironmonger, M. Shipton, F. Slater, P. Szeto, M. G.
8
Unthank, F. R. Alexandre, C. Caillet, C. B. Dousson, Tetrahedron Lett.
018, 59, 2154–2156; g) A. Mohd, T. Anitha, K. R. Reddy, J. Wencel-
a) P. C. J. Kamer, P. W. N. M. Van Leeuwen, Eds. , John Wiley & Sons,
New York, 2012; b) W. Fu, W. Tang, ACS Catal. 2016, 6, 4814–4858;
c) T. Imamoto, Chem. Rec. 2016, 16, 2655–2669.
2
Delord, F. Colobert, Eur. J. Org. Chem. 2019, 7836–7841.
[
12] a) F. Toda, K. Mori, S. Zafra, I. Goldberg, J. Org. Chem. 1988, 53, 308–
a) Y. Xiao, Z. Sun, H. Guo, O. Kwon, Beilstein J. Org. Chem. 2014, 10,
312; b) J. Drabowicz, P. Łyżwa, J. Omelańczuk, K. M. Pietrusiewicz, M.
2
2
2
2
5
4
2
089–2121; b) H. Guo, Y. C. Fan, Z. Sun, Y. Wu, O. Kwon, Chem. Rev.
018, 118, 10049–10293; c) H. Ni, W.-L. Chan, Y. Lu, Chem. Rev.
018, 118, 9344–9411; d) M. Benaglia, S. Rossi, Org. Biomol. Chem.
010, 8, 3824–3830; e) S. Kotani, Chem. Pharm. Bull. 2019, 67, 519–
26; f) T. Ayad, A. Gernet, J.-L. Pirat, D. Virieux, Tetrahedron 2019, 75,
385–4418; g) S. Liu, Y. Kumatabara, S. Shirakawa, Green Chem.
016, 18, 331–341; h) A. Golandaj, A. Ahmad, D. Ramjugernath, Adv.
Mikołajczyk, Tetrahedron: Asymmetry 1999, 10, 2757–2763; c) J. Holt,
A. M. Maj, E. P. Schudde, K. M. Pietrusiewicz, L. Sieroń, W. Wieczorek,
T. Jerphagnon, I. W. C. E. Arends, U. Hanefeld, A. J. Minnaard,
Synthesis 2009, 2061–2065; d) F. A. Kortmann, M.-C. Chang, E. Otten,
E. P. A. Couzijn, M. Lutz, A. J. Minnaard, Chem. Sci. 2014, 5, 1322–
1327; e) P. Bagi, V. Ujj, M. Czugler, E. Fogassy, G. Keglevich, Dalton
Trans. 2016, 45, 1823–1842; f) P. Bagi, B. Varga, A. Szilágyi, K.
Karaghiosoff, M. Czugler, E. Fogassy, G. Keglevich, Chirality 2018, 30,
Synth. Catal. 2017, 359, 3676–3706.
[4]
[5]
[6]
J. Chrzanowski, D. Krasowska, J. Drabowicz, Heteroatom Chem. 2018,
509–522.
2
9, e21476.
D. Herault, D. H. Nguyen, D. Nuel, G. Buono, Chem. Soc. Rev. 2015,
4, 2508–2528.
a) R. Beaud, R. J. Phipps, M. J. Gaunt, J. Am. Chem. Soc. 2016, 138,
3183–13186; b) Y. S. Jang, Ł. Woźniak, J. Pedroni, N. Cramer,
[
13] G. Li, X. Wang, Y. Zhang, Z. Tan, P. DeCroos, J. C. Lorenz, X. Wei, N.
Grinberg, N. K. Yee, C. H. Senanayake, J. Org. Chem. 2017, 82, 5456–
4
5460.
[
14] a) F. Faigl, E. Fogassy, M. Nógrádi, E. Pálovics, J. Schindler,
Tetrahedron Asymmetry 2008, 19, 519–536; b) H. Lorenz, A. Seidel-
Morgenstern, Angew. Chem. Int. Ed. 2014, 53, 1218–1250; c) S. Mane,
Anal. Methods 2016, 8, 7567–7586.
1
Angew. Chem. Int. Ed. 2018, 57, 12901–12905; c) X.-T. Liu, Y.-Q.
Zhang, X.-Y. Han, S.-P. Sun, Q.-W. Zhang, J. Am. Chem. Soc. 2019,
141, 16584–16589.
[
15] a) D. Seebach, A. K. Beck, A. Heckel, Angew. Chem. Int. Ed. 2001, 40,
[
7]
a) F. Chaux, S. Frynas, H. Laureano, C. Salomon, G. Morata, M.-L.
Auclair, M. (Massoud) Stephan, R. Merdès, P. Richard, M.-J. Ondel-
Eymin, et al., C. R. Chim. 2010, 13, 1213–1226; b) Z. S. Han, N. Goyal,
M. A. Herbage, J. D. Sieber, B. Qu, Y. Xu, Z. Li, J. T. Reeves, J.-N.
Desrosiers, S. Ma, et al., J. Am. Chem. Soc. 2013, 135, 2474–2477; c)
L. Copey, L. Jean-Gérard, E. Framery, G. Pilet, V. Robert, B. Andrioletti,
Chem. - A Eur. J. 2015, 21, 9057–9061; d) A. D’Onofrio, L. Copey, L.
Jean-Gérard, C. Goux-Henry, G. Pilet, B. Andrioletti, E. Framery, Org.
Biomol. Chem. 2015, 13, 9029–9034; e) S.-G. Li, M. Yuan, F. Topic, Z.
S. Han, C. H. Senanayake, Y. S. Tsantrizos, J. Org. Chem. 2019, 84,
9
2–138; b) H. Pellissier, Tetrahedron 2008, 64, 10279–10317; c) K.
Tanaka in Comprehensive Chirality Vol. 8 (Eds.: E. M. Carreira, H.
Yamamoto), Elsevier, Amsterdam, 2012 pp. 63-90; d) B. Yang, H. Xie,
K. Ran, Y. Gan, Lett. Org. Chem. 2017, 15, 10–14.
[
[
[
[
16] P. Bagi, R. Herbay, in Organophosphorus Chemistry - Novel
Developments (Ed.: G. Keglevich), De Gruyter, Berlin, 2018, pp. 66–90.
17] J. Meisenheimer, J. Casper, M. Höring, W. Lauter, L. Lichtenstadt, W.
Samuel, J. Liebig. Ann. Chem. 1926, 449, 213–248.
18] G. Szekely, M. F. Jimenez-Solomon, P. Marchetti, J. F. Kim, A. G.
Livingston, Green Chem. 2014, 16, 4440–4473.
7
291–7302.
a) O. Korpiun, R. A. Lewis, J. Chickos, K. Mislow, J. Am. Chem. Soc.
968, 90, 4842–4846; b) U. Schmidt, B. Riedl, H. Griesser, C. Fitz,
19] a) N. F. Ghazali, F. C. Ferreira, A. J. P. White, A. G. Livingston,
Tetrahedron: Asymmetry 2006, 17, 1846–1852; b) N. F. Ghazali, F. C.
Ferreira, A. J. P. White, A. G. Livingston, Desalination 2006, 199, 398–
[8]
1
Synthesis 1991, 655–657; c) N. G. Andersen, P. D. Ramsden, D. Che,
M. Parvez, B. A. Keay, Org. Lett. 1999, 1, 2009–2011; d) N. G.
Andersen, P. D. Ramsden, D. Q. Che, M. Parvez, B. A. Keay, J. Org.
Chem. 2001, 66, 7478–7486; e) M. Oliana, F. King, P. N. Horton, M. B.
Hursthouse, K. K. (Mimi) Hii, J. Org. Chem. 2006, 71, 2472–2479; f) H.
Adams, R. C. Collins, S. Jones, C. J. A. Warner, Org. Lett. 2011, 13,
4
00; c) I. Sereewatthanawut, F. C. Ferreira, N. F. Ghazali, A. G.
Livingston, AIChE J. 2010, 56, 893–904.
[
20] C. Didaskalou, J. Kupai, L. Cseri, J. Barabas, E. Vass, T. Holtzl, G.
Szekely, ACS Catal. 2018, 8, 7430–7438.
[
21] P. Kisszekelyi, A. Alammar, J. Kupai, P. Huszthy, J. Barabas, T. Holtzl,
L. Szente, C. Bawn, R. Adams, G. Szekely, J. Catal. 2019, 371, 255–
6576–6579.
2
61.
22] B. Zupančič, B. Mohar, M. Stephan, Adv. Synth. Catal. 2008, 350,
024–2032.
23] K. M. Pietrusiewicz, M. Zablocka, J. Monkiewicz, J. Org. Chem. 1984,
9, 1522–1526.
24] a) K. Matsumoto, T. Okamoto, K. Otsuka, Bull. Chem. Soc. Jpn. 2004,
7, 2051–2056; b) K. Matsumoto, K. Otsuka, T. Okamoto, H. Mogi,
[
9]
a) R. K. Haynes, T.-L. Au-Yeung, W.-K. Chan, W.-L. Lam, Z.-Y. Li, L.-L.
Yeung, A. S. C. Chan, P. Li, M. Koen, C. R. Mitchell, et al., Eur. J. Org.
Chem. 2000, 3205–3216; b) Q. Xu, C.-Q. Zhao, L.-B. Han, J. Am.
Chem. Soc. 2008, 130, 12648–12655; c) A. J. Bloomfield, S. B. Herzon,
Org. Lett. 2012, 14, 4370–4373; d) S.-Z. Nie, Z.-Y. Zhou, J.-P. Wang, H.
Yan, J.-H. Wen, J.-J. Ye, Y.-Y. Cui, C.-Q. Zhao, J. Org. Chem. 2017, 82,
[
[
[
2
4
7
9
425–9434; e) Z. S. Han, H. Wu, Y. Xu, Y. Zhang, B. Qu, Z. Li, D. R.
Caldwell, K. R. Fandrick, L. Zhang, F. Roschangar, et al., Org. Lett.
017, 19, 1796–1799; f) J. Chrzanowski, D. Krasowska, M. Urbaniak, L.
Synlett 2007, 729–732.
[
25] a) H. Kaku, A. Nakamaru, M. Inai, T. Nishii, M. Horikawa, T. Tsunoda,
Tetrahedron 2010, 66, 9450–9455; b) H. Kaku, T. Imai, R. Kondo, S.
Mamba, Y. Watanabe, M. Inai, T. Nishii, M. Horikawa, T. Tsunoda, Eur.
J. Org. Chem. 2013, 8208–8213; c) H. Kaku, M. Ito, M. Horikawa, T.
Tsunoda, Tetrahedron 2018, 74, 124–129.
2
Sieroń, P. Pokora-Sobczak, O. M. Demchuk, J. Drabowicz, Eur. J. Org.
Chem. 2018, 4614–4627.
[
10] a) P. Kiełbasiński, R. Źurawiński, K. M. Pietrusiewicz, M. Zabłocka, M.
Mikołajczyk, Tetrahedron Lett. 1994, 35, 7081–7084; b) K. Shioji, Y.
Ueno, Y. Kurauchi, K. Okuma, Tetrahedron Lett. 2001, 42, 6569–6571;
c) S. Kaczmarczyk, M. Kwiatkowska, L. Madalińska, A. Barbachowska,
M. Rachwalski, J. Błaszczyk, L. Sieroń, P. Kiełbasiński, Adv. Synth.
Catal. 2011, 353, 2446–2454; d) Y. Zhang, H. He, Q. Wang, Q. Cai,
Tetrahedron Lett. 2016, 57, 5308–5311; e) H. Fernández-Pérez, A.
Vidal-Ferran, Org. Lett. 2019, 21, 7019–7023.
[
26] P. Bagi, A. Fekete, M. Kállay, D. Hessz, M. Kubinyi, T. Holczbauer, M.
Czugler, E. Fogassy, G. Keglevich, Heteroatom Chem. 2015, 26, 79–90.
27] a) M. A. Spackman, D. Jayatilaka, CrystEngComm 2009, 11, 19–32; b)
S. P. Thomas, P. R. Spackman, D. Jayatilaka, M. A. Spackman, J.
Chem. Theory Comput. 2018, 14, 1614–1623.
[
[
28] a) R. K. Henderson, C. Jiménez-González, D. J. C. Constable, S. R.
Alston, G. G. A. Inglis, G. Fisher, J. Sherwood, S. P. Binks, A. D.
Curzons, Green Chem. 2011, 13, 854; b) F. P. Byrne, S. Jin, G.
[
11] a) E. Bergin, C. T. O’Connor, S. B. Robinson, E. M. McGarrigle, C. P.
O’Mahony, D. G. Gilheany, J. Am. Chem. Soc. 2007, 129, 9566–9567;
1
2
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.202000035
FULL PAPER
Paggiola, T. H. M. Petchey, J. H. Clark, T. J. Farmer, A. J. Hunt, C.
Robert McElroy, J. Sherwood, Sustain. Chem. Process. 2016, 4, 7.
[29] G. Székely, J. Bandarra, W. Heggie, B. Sellergren, F. C. Ferreira, J.
Membr. Sci. 2011, 381, 21–33.
1
3
This article is protected by copyright. All rights reserved.

European Journal of Organic Chemistry
10.1002/ejoc.202000035
FULL PAPER
Table of Contents
The enantioseparation of dialkyl-arylphosphine oxides was elaborated, and five phosphine oxides were prepared with ee higher than
4%. It was shown that these resolutions can be performed on gram scale, and can give both enantiomers of a phosphine oxide. It
9
was demonstrated that organic solvent nanofiltration can serve as an alternative method of column chromatography for the recovery
of the P-oxide enantiomers and the resolving agents.
Twitter username: @peter_bagi
Key Topic: Optical Resolution
1
4
This article is protected by copyright. All rights reserved.