Conformations of monoylidic diester triphenylphosphonium ylides

Article abstract of DOI:10.1016/j.molstruc.2012.08.051

In ylidic triphenylphosphonium carboxylic esters the ester oxygen can be oriented towards (syn) or away (anti) from phosphorus, but except for small ylidic ester groups, e.g., Me, the anti conformer is dominant. With suitable crystals conformations are established by X-ray crystallography, but HF and DFT computations, with NMR and IR spectroscopy, are useful methods. Bulky ylidic or nonylidic groups strongly favor the anti conformer and even with small carboxylic groups, e.g. ethoxy, anti conformers are preferred in solution and are dominant in the crystal. The balance of attractive interactions between anionoid oxygen and cationoid phosphorus and nonbonding interactions, controls conformations, as indicated by evidence from NMR and IR spectroscopy, HF and DFT calculations, and X-ray observations.

Full text of DOI:10.1016/j.molstruc.2012.08.051

Journal of Molecular Structure 1034 (2013) 51–56
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Structure
journal homepage: www.elsevier.com/locate/molstruc
Conformations of monoylidic diester triphenylphosphonium ylides
a
a
b
c
Fernando Castañeda a, , Paul Silva , Cristina Acuña , M. Teresa Garland , Clifford A. Bunton
⇑
a Departamento de Química Orgánica y Fisicoquímica, Facultad de Ciencias Químicas y Farmacéuticas, Universidad de Chile, Santiago-1, Casilla 233, Santiago, Chile
b Departamento de Física, Facultad de Ciencias Físicas y Matemáticas, Universidad de Chile, Casilla 487-3, Santiago, Chile
c Department of Chemistry and Biochemistry, University of California, Santa Barbara, CA 93106, USA
h i g h l i g h t s
"
"
"
Conformations were determined by spectroscopy, X-ray, HF, BLYP and B3LYP methods.
Anti conformers are preferred in solution and are dominant in the crystal.
Conformations are controlled by electric and non-bonding steric interactions.
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 27 July 2012
Received in revised form 28 August 2012
Accepted 30 August 2012
Available online 8 September 2012
In ylidic triphenylphosphonium carboxylic esters the ester oxygen can be oriented towards (syn) or away
(anti) from phosphorus, but except for small ylidic ester groups, e.g., Me, the anti conformer is dominant.
With suitable crystals conformations are established by X-ray crystallography, but HF and DFT computa-
tions, with NMR and IR spectroscopy, are useful methods. Bulky ylidic or nonylidic groups strongly favor
the anti conformer and even with small carboxylic groups, e.g. ethoxy, anti conformers are preferred in
solution and are dominant in the crystal. The balance of attractive interactions between anionoid oxygen
and cationoid phosphorus and nonbonding interactions, controls conformations, as indicated by evidence
from NMR and IR spectroscopy, HF and DFT calculations, and X-ray observations.
Keywords:
Triphenylphosphonium ylides
Molecular structures
NMR and IR spectroscopy
HF and DFT calculations
X-ray crystallography
Ó 2012 Elsevier B.V. All rights reserved.
1. Introduction
R0 a simple alkyl or aryl group the syn conformer is preferred,
and in solution the syn–anti ratio is modestly sensitive to R0 [2]. Ef-
Triphenylphosphonium ylides with one aliphatic keto or ester
group can have syn or anti conformations, defined by orientation
of the ylidic acyl group relative to phosphorus, Scheme 1, with
the acyl oxygen as syn, toward, or anti, away from, phosphorus.
There is electronic delocalization with partial double bonding be-
tween the ylidic and acyl carbons, and partial single bonding to
phosphorus, allowing syn–anti conversions. Coulombic interaction
between the anionoid acyl oxygen and cationoid phosphorus
should favor the syn conformer as in mono keto ylides [1–3].
In some monomethyl ester ylides [2] 1, R = CH3 and R0 is hydro-
gen, an alkyl or alkyl substituted group, or an aryl group, with
favorable interactions between anionoid acyl oxygen and cationoid
phosphorus in a syn ylide (Scheme 1). The crystalline monoester 1,
with R = CH3 and R0 = H, has the expected syn conformation, consis-
tent with computation, but the 1H NMR spectrum shows that in
CDCl3 there is the syn and the minor anti conformer [2,4,5]. With
fects of R0 on the syn–anti ratio can involve steric and electronic ef-
fects and conformer conversion is slow on the NMR time scale at
room temperature [2].
There are exceptions to the preferred syn conformation in some
ylidic monoesters, and in an extensive compilation of X-ray evi-
dence, Aitken et al. [6] noted that for some monoylidic esters anti
conformers are preferred. For example, Cameron et al. [7] showed
that the crystalline monoylidic diester 2 (Scheme 2) with R = CH3
and R0 = t-Bu has the anti conformation and the nonylidic ester
group is linked to the ylidic carbon by a CH2 tether group so here
conformational control is apparently due to steric rather than elec-
tronic effects. The conformation in solution was not examined in
detail, but spectral and NMR evidence indicate that only one con-
former is present. For this and some other monoylides, [6,7] bulky
nonylidic groups are controlling the conformation of the ylidic es-
ter group in the crystal and probably also in solution, although they
are not bonded to the ylidic carbon. Small nonylidic ester groups
do not control conformations, but the crystalline monoester 3 with
R and R0 = CH2CH3, has the anti conformation [8a] and it was exam-
⇑
Corresponding author. Tel.: +56 2 978 2856; fax: +56 2 978 2868.
E-mail address: fcastane@ciq.uchile.cl (F. Castañeda).
0022-2860/$ - see front matter Ó 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.molstruc.2012.08.051

52
F. Castañeda et al. / Journal of Molecular Structure 1034 (2013) 51–56
RO
C
O
O
'
C
'
R
R
C
O
R
C
PPh3
PPh3
syn
anti
Chart 1. Atomic numbering for calculated and observed monoylidic diester
structures. aRef. [7]. bRef [8b].
O-
C
RO
C
Positions of bonds and angles for conformers are designated as
in X-ray crystallography [8a]. Computed distances and angles are
given to the second decimal places and angles to first whole num-
bers but for the crystal structures full numerical values are given in
References 7 and 8a. The sum of the bond angles at the acyl car-
bons is close to the 360° for a planar carbon and strong delocaliza-
tion over the ylidic system. Natural population analysis (NPA) for
syn and anti 3 (B3LYP/6-31G(d)) and for anti ylide 5 (B3LYP/6-
31G(d) and BLYP/6-31G(d)) are in Table 1. Acyl stretching frequen-
cies were estimated with unconstrained structures, and frequen-
cies are rounded off to the nearest whole number. Calculated and
observed bond lengths and angles for 2, 3, 4 and 6 are in Tables
5–8 (Supplementary data).
'
R
C
'
R
-
O
O
C
R
+PPh3
+PPh3
Scheme 1. Syn and anti conformers of monoylidic methyl esters 1. R = CH3; R0 = H,
alkyl, aryl [2,5].
ined in solution, together with other ylidic monoesters with differ-
ent sized alkoxy groups, Scheme 2. The atom numbering of the
ylides examined in this work, Chart 1, follows that applied to the
diethyl ylide [8].
In triphenylphosphonium ylides electronic delocalization favors
zwitterionic structures with cationoid phosphorus, anionoid oxy-
gen, and an approximately planar ylidic group [3,9], Scheme 1.
Conformer conversion involves loss of ylidic resonance and syn
and anti conformers in solution, can generally be identified by
NMR spectroscopy and we used it and IR spectroscopy in examin-
ing conformations in solution, Scheme 2. Bond lengths and angles
were estimated by calculations with HF and DFT methods, which
also predicted acyl stretching frequencies, and geometries were
compared with observed values for ylides with known crystal
structures. In identifying ylides with two alkyl groups the alkyl
symbol of the ylidic ester group, R, is given first, followed by that
of the nonylidic group, R0, as for simple monoylides [2,4].
2.2. IR spectroscopy
The IR spectra were examined in KBr disks or in CHCl3 solution
on a Bruker IFS56 FT or on Leitz III-C spectrometers and the signal
maxima were identified by the OPUS deconvolution method with
resolution at 4 cmꢀ1. CHCl3 was treated with alumina before use.
There are two strong signals assigned to the ylidic and nonylidic
ester C@O groups which are well separated from low frequency
signals. It appeared that signals of only the anti conformer are
strong in both KBr disks and CHCl3 where we had hoped to see anti
and syn signals. The combination of IR acyl stretching frequencies
and ab initio methods are useful in identification of anti conforma-
tions of ylidic esters [11]. Predicted signals from the BLYP method
and experimental IR acyl stretching frequencies for diester ylides
2–8 are in Table 2.
2. Experimental
2.1. Computation
Computations for structural optimization were made with
Wavefunction (Windows), ’06 or ‘08 software and optimized with
HF/6-31G(d) and DFT methods [10] and with reasonable agree-
ment between these structures and those from X-ray crystallogra-
phy, BLYP and B3LYP, gives slightly longer acyl bonds than in the
crystal but for angles, results are acceptable.
2.3. NMR spectroscopy
The 1H and 13C NMR signals were examined on a Bruker DRX
300 or Varian Inova 500 spectrometers in acid free CDCl3 and are
referred to TMS or 85% H3PO4. The 1H NMR signal is useful in char-
acterizing anti conformations of novel monoylidic diesters and in
establishing orientations of CH3 alkoxy groups relative to triphen-
ylphosphonium groups by observation of p-shielding [12]. Interac-
tions between a terminal alkyl group and the face of a phenyl
O
R
O
C
C
'
R
CH2 CO2
'
R
Table 1
RO
C
CH2 CO2
O
C
Natural population analyses for syn and anti mono ylidic diethyl diester 3,
P Ph3
Ph3P@C(CO2Et)ACH2ACO2Et,
and
anti
monoylidic
diester
5,
P Ph3
Ph3P@C(CO2AC(CH3)3)ACH2ACO2CH3.
Syn ylide 3
Anti ylide 3
Anti ylide 5
anti
syn
P1
C1
C2
O2
O1
C7
O5
O6
1.720
ꢀ0.762
0.762
ꢀ0.696
ꢀ0.577
0.847
1.712
ꢀ0.781
0.775
ꢀ0.655
ꢀ0.605
0.857
1.712 (1.647)
ꢀ0.777 (ꢀ0.734)
0.778 (0.726)
ꢀ0.665 (ꢀ0.627)
ꢀ0.612 (ꢀ0.586)
0.854 (0.806)
- O
C
R
O
C
'
R
'
CH2 CO2
CH2 CO2R
- O
C
R O
C
ꢀ0.608
ꢀ0.633
ꢀ0.633 (ꢀ0.602)
+P Ph3
+P Ph3
ꢀ0.572
ꢀ0.560
ꢀ0.548 (ꢀ0.520)
Scheme 2. Syn and anti conformers of monoylidic diesters. R = alkyl. R0 = alkyl,
R = R0; R – R0.
Atoms are numbered as in Chart 1. Fractional charges are from B3LYP/6-31G(d) and
in parentheses BLYP/6-31G(d).

F. Castañeda et al. / Journal of Molecular Structure 1034 (2013) 51–56
53
Table 2
In a general method for synthesis of monoylidic diesters, 4 and
5, a solution of alkyl 2-bromoacetate (12 mmol; alkyl = ethyl or
methyl) in dry ethyl acetate (5 mL) was added dropwise to a solu-
tion of alkyl 2-(triphenylphosphoranylidene) acetate (24 mmol; al-
kyl = ethyl or t-butyl) in dry ethyl acetate (100 mL) under an inert
atmosphere at 25 °C. After 10 min of stirring the mixture was kept
at 40–45 °C for 4 h forming a white solid which was removed by
suction filtration and washed with dry ethyl acetate. This solid is
the triphenylphosphonium salt (Ph3P+ACH2ACO2R Brꢀ; R = Et, t-
Bu), formed by transylidation [12b]. After filtration the solvent
was removed by rotary evaporation giving a solid which was
recrystallized from ethyl acetate–hexane (1:1).
The monoylidic diesters isomers 7 and 8 were made as in the
above general procedure but with benzene as solvent at 25 °C for
8 h. Ylide 7 was made by transylidation from Ph3P@CHACO2Et
and t-butyl 2-bromoacetate and ylide 8 was made from Ph3-
P@CHACO2-t-Bu and ethyl 2-bromoacetate.
Predicted and observed IR acyl stretching frequencies of monoylidic diester ylides 2–
8.
Predicte (cmꢀ1
)
BLYP
Observeda (cmꢀ1
)
Anti ethyl ethyl, 3 [8b]
Syn ethyl ethyl, 3
Anti ethyl methyl, 4
Anti ethyl, methyl ethyl, 6 [8b]
Syn ethyl methyl ethyl, 6
Anti methyl t-butyl, 2 [7]
Anti t-butyl methyl, 5
Anti ethyl t-butyl, 7
1654
1593
1655
1644
1588
1665
1635
1655
1711
1724
1730
1685
1720
1692
1708
1712
1620
1730
1630
1631
1730
1724
1620b
1630
1628
1622
1730b
1730
1731
1730
Anti t-butyl ethyl, 8
a
In a KBr disk unless specified.
In CHCl3 [7].
b
group may be stabilizing [13]. Selective 1H and 13C NNR data for
ylides 2–8 are in Table 3.
Synthetic and spectroscopic details for ylides 3–8 and physical
properties, are given as Supplementary data.
2.4. Elemental analyses
Ylides formal names are noted as Supplementary data, but gi-
ven their length, abbreviated simple names and numbers are used
in the main text.
Elemental analyses were made on a Fison EA 1108 analyzer. The
purity of novel compounds 4, 5, 7 and 8 for spectral work was
shown by elemental microanalysis (Supplementary data).
3. Results and discussion
2.5. Materials
3.1. Structural geometries
– The methyl, t-butyl diester 2 was made by reaction of monoes-
ter Ph3P@CHACO2CH3 with t-butyl 2-bromoacetate [7].
– The diethyl diester 3 was made from Ph3P@CHACO2CH2ACH3
and ethyl 2-bromoacetate by transylidation [8b].
– The diethyl 3-methyl-2-triphenylphosphoranylidene succinate
6 was made by methylating the lithium salt of diethyl ester 3
[8b].
Bond lengths and angles estimated by HF and DFT methods (Ta-
bles 5–8 in Supplementary data) were compared with X-ray crys-
tallographic values, when available. Computations are for anti–syn
conformers, although for some ylides only one conformer is de-
tected in the crystal or in solution. Geometries of the ylidic group
in anti and syn conformers from the B3LYP/6-31G(d) method are
Table 3
1H and 13C NMR selective data for monoylidic diester ylides 2–8 showing anti–syn conformations.
Et–Et 3 [8b]
Et–Me 4
Et–Me–Et 6 [8b]
Me–t-Bu 2 [7]
t-Bu–Me 5
Et–t-Bu 7
t-Bu–Et 8
Anti
Syn
Anti
Syn
Anti
Syn
Anti
Anti
Anti
Anti
1H NMR
O1ACH2ACH3
0.38t
1.94H
J = 7.1
0.46t
2.05H
J = 7.1
1.19t
0.95H
J = 7.1
0.47t
1.7H
J = 7
0.45t
3H
J = 7.1
O1ACH3
3.34s
3H
1.00–1.14
3t; 4.06H
1.10–1.35
3t; 7.3H
O6ACH2ACH3
O6ACH3
1.09t
3H
J = 7.1
3.50s
2.05H
3.53s
4.09H
3.48s
3H
O1AC(CH3)3
O6AC(CH3)3
0.96s
9H
0.94s
9H
1.32s
9H
1.25s
9H
C10HAR00
2.92d
1.41H
J = 17.4
2.82d
0.65H
J = 17.7
3.00d
1.36H
J = 17.3
2.88d
0.68H
J = 17.1
2.07–2.83
m; 1,1H
2.89d
2H
J = 17
2.90d
2H
J = 17
2.90d
2H
J = 17.3
2.95d
2H
J = 17.4
13C NMR
C2O2A
170.2d
J = 12.9
170.9d
J = 18.8
170.2d
J = 12.7
170.8d
J = 18.6
171.3
s
170.5
s
172.2
s
170.6
s
C7O2A
175.2
s
175.4
s
175.1
s
175.5
s
175.1
s
175.2
s
175.5
s
175.4
s
Ylides and atoms are numbered as in Chart 1. 1H and 13C NMR in CDCl3. At 25 °C chemical shifts referenced to TMS or external 85% H3PO4. 1H coupling constants J in Hz.
31PA13C coupling constants in Hz with 1H decoupling.

54
F. Castañeda et al. / Journal of Molecular Structure 1034 (2013) 51–56
compared with X-ray geometries in Table 4a, and in Table 4b com-
putational evidence on conformations of ylides which do not form
crystals suitable for use of X-ray are compared with those from
NMR spectroscopy in CDCl3.
The simplicity of the signals of the nonylidic ester groups shows
that in solution there is free rotation of this ester group on the NMR
time scale.
In the methyl derivative 6, the 1H signals of the terminal methyl
hydrogens of the two ethyl ester groups and of the methyl hydro-
gens on the tether group are readily characterized (Table 3) and in
solution the anti: syn ratio of 57:43 indicates that the anti con-
former preference is less than in the other ylides, Table 3.
3.2. NMR spectroscopy
The 1H or 13C, NMR spectra, of diethyl, and ethyl methyl mono-
ylidic diesters 3 and 4 in CDCl3 show that they are anti–syn mix-
tures while X-ray crystallography shows that crystalline 3 has
the anti conformation [8a] Unlike the methyl ester ylides [2] anti
conformers are dominant in the crystal and in solution, as shown
3.3. Acyl stretching frequencies and IR spectra
The known syn conformation of the crystalline monomethyl
derivative, 1b [5,6] and the anti conformations of diethyl and
methyl t-butyl derivatives 3 and 2 [7,8a] test computational meth-
ods and the use of NMR and IR spectroscopy. Stretching frequen-
cies of ylidic acyl groups are sensitive to conformation and in
comparisons of observed and predicted values the latter are cor-
rected with method sensitive Scale Factors, SF, as surveyed by Scott
and Radom [14]. For BLYP/6-31G(d) SF = 0.9945, which can be ne-
glected within the accuracy of measurements. The literature SF va-
lue for HF/6-31G(d) of 0.8953 [14] overpredicts ylidic acyl
stretching frequencies and fits were obtained with SF = 0.834 and
0.866 for ylidic ketone and ester acyl groups, respectively [12d].
Estimated frequencies with B3LYP/6-31G(d), SF = 0.9614, are con-
sistentIy high [11,15]. The calculations are for molecules in the
gas phase at absolute zero, but inclusion of an empirical solvent
correction has IittIe effect on the predictions.
The higher frequency signals of the nonylidic ester groups do
not provide useful information on ylidic conformations because
of free rotation about the tether group. Acyl stretching frequencies
estimated in KBr disks for anti conformers are compared with cal-
culated values in Table 2. The generalIy used computations of acyl
lR frequencies are unreliable in predicting relative signal strength
[11] and for stabilized syn–anti dialkyl ester ylides, with known
conformations, computed and observed anti signals are always
the stronger [10b].
by p-shielding of NMR signals of the terminal hydrogens of the yli-
dic alkoxy groups. The 1H NMR spectrum of the t-butyl methyl
monoylidic diester 5 in CDCI3 is that of the anti conformer with
no indication of a minor syn conformer (Table 3). The methyl t-bu-
tyl monoylidic diester 2, the isomer of 5, is the anti conformer in
the crystal and probably also in solution, although NMR signals
were not assigned [7]. Snyder et al. [2a] reported that large alkyl
substituents at ylidic carbon atoms, in monoylidic methyl ester 1
(Scheme 1), favor a relatively high population of anti conformers.
These conclusions are consistent with NMR evidence on effects of
alkyl group size on syn–anti ratios in solution.
The dialkyl derivatives 3 and 4 are conformer mixtures in CDCl3
in an approximate 2:1 ratio and assignments are in Table 3 (the
conformation for crystalline 3 is anti [8a]). The 1H doublets at
2.8–3.0 ppm in 3, and 4 are of the ylidic CH2 tether group, with sig-
nificant 31P coupling, and the 1H chemical shifts in conformers, 3
and 4, are governed by different interactions in the ylidic moieties,
as in other monoylides. The 1H doublets of the ylidic CH2 tether
group are similar for any conformer of 2–8 and the bulky ylides
2, 5, 7 and 8 are assigned as anti conformers. Compound 7, Table 3,
is another example of a dominant anti conformer in solution. The
1H methyl of the ylidic ester showed
p-shielding at 0.45 ppm and
a normal resonance for the 1H t-butyl group at 1.25 ppm. Moreover
anti conformer 7 (Table 3) showed only one 1H doublet at 2.90 ppm
for the CH2 tether group and only two 13C signals at 172.2 and
175.5 ppm for the two ester groups indicating the existence of only
one conformer.
The combination of IR spectra of an ylidic ester group and NMR
spectroscopy is useful in distinguishing between single and mixed
conformers in solution. The IR C@O stretching signals are well sep-
arated from other signals and predicted frequencies from HF and
Table 4
Conformations from computation, X-ray crystallography or NMR evidence for monoylidic esters.
Ylide
Calculated geometrya
X-ray geometry
P@C
(a) Computation with and without X-ray
CAC
C@O
Torsion angle
P@C
CAC
C@O
Torsion angle
1a Me syn
anti
1.71
1.71
1.73
1.73
1.72
1.73
1.42
1.43
1.42
1.47
1.43
1.42
1.24
1.22
1.24
1.23
1.22
1.24
2.2
1.70
1.72
1.72
1.39
1.40
1.42
1.23
1.23
1.22
2.3b
176
169
4.2
169
8.2
3: Et, Et anti
syn
177c
177d
2: Me, t-Bu anti
syn
Ylide
P@C
CAC
C@O
Torsion angle
NMR evidence
(b) Conformations and 1H NMR in CDCl3 (without X-ray)
1 Me
syn
anti
anti
syn
anti
syn
1.72
1.72
1.74
1.74
1.72
1.73
1.43
1.43
1.44
1.44
1.43
1.43
1.23
1.24
1.24
1.25
1.3
169f
8.4f
Anti:syn = 67:33
Anti:syn = 57:43
Anti:syn = 100:0
6: Et, Me, Et
5: t-Bu, Me
164e
4.6e
169f
8.4f
1.26
fB3LYP/6.31G(d) unless specified.
a
B3LYP/6-31G(d), length in Å, angles in °.
Aitken [6].
[8a].
Cameron [7].
b
c
d
e
BLYP/6-31G(d).

F. Castañeda et al. / Journal of Molecular Structure 1034 (2013) 51–56
55
DFT methods, are in reasonable agreement with observed values,
but the weakness of syn acyl signals limits the utility of this meth-
od in examining ratios of mixed conformers. Infrared spectroscopy
complements other conformational evidence but is only useful in
estimating concentrations when spectra of the individual compo-
nents can be examined independently.
A treatment when X-ray is not available is shown in Table 4b.
Estimation of torsion angles, with bond strengths and angles,
establishes conformation and with NMR evidence establishes con-
formation ratios. We follow this approach for several of the ylides
which were examined in more detail.
3.6. Conformation control
3.3.1. IR spectra of acyl groups
Observed acyl stretching frequencies of ylidic and nonylidic acyl
groups are in reasonable agreement with predicted values for anti
ylides 2 and 3 (Table 2). There are problems in treating IR spectra
of the diethyl and ethyl methyl compounds 3 and 4 which from
NMR in CDCl3, are syn and anti mixtures, although only the anti
diethyl ylide 3 is present in the crystal [8a]. This solvent effect
on conformation was observed with the simpler monoylide 1a
[1a,16]. The strong IR observed signals of the anti ylidic acyl groups
in 2, 3 and 6 (Table 2) are 1620, 1620 and 1631 cmꢀ1 respectively,
as expected for spectra in KBr disks or CHCl3. The X-ray spectra of
some crystalline anti–syn diylidic diester triphenylphosphonium
ylides had shown that the syn acyl oxygen is located between the
edges of two phosphonium phenyl groups and is close enough to
interact with the phenyl hydrogens [14,17] which could affect IR
signals of syn acyl groups in solution. These types of interactions
could be responsible for the weak signals of syn acyl groups in 3,
4 and 6 but we cannot test this speculation by X-ray data for 3 with
its anti conformer in the crystal. Calculated geometries of syn acyl
groups in stabilized syn–anti diester ylides are consistent with X-
ray, NMR and IR evidence regarding locations of acyl groups
[11,18].
Conformations in the crystal, are sensitive to molecular pack-
ing and syn–anti ratios may differ from those in solution, as for
the methyl ester 1, R0 = H [5], and the diethyl ylide 3 [8a], Ta-
ble 6, but the monoylidic diester 2 with a bulky nonylidic t-butyl
ester group is a single anti conformer in the crystal and probably
also in solution, [7]. The monoylidic diesters 5 and 8 with an yli-
dic t-butyl ester group and a small nonylidic methyl or ethyl es-
ter groups, and could not be examined by X-ray crystallography
but NMR and IR spectroscopy, with computations, indicate that
it also has the anti conformation. The monoylidic diesters with
smaller ester groups, 3 and 4 are mixtures of conformers in
CDCl3 with anti–syn ca. 2:1, consistent with earlier evidence that
single conformations in the crystal and in solution are sensitive
to bulky ester groups, [7], and are affected by the ylidic ester
groups to different extents in the crystal and in solution, Table 4,
[6]. Consistently bulky ylidic or nonylidic alkoxy groups favor
the anti conformation and ylides with small alkoxy groups are
generally mixtures in solution, although usually not necessarily
so in the crystal [7,8a].
Evidence on stabilized keto ester and diester ylides with
known structures illustrates the factors controlling conforma-
tional differences [12b,c]. Steric factors affect conformations,
for example, in stabilized keto ester ylides the smaller keto
group is syn and the ester group is anti, and in diesters the smal-
ler ester group is typically syn and the larger group is anti,
although crystalline anti–anti conformers are observed with
some small alkoxy groups [20]. Bulky alkoxy groups, e.g., iso pro-
poxy or t-butoxy, apparently do not interact unfavorably with
the triphenylphosphonium group [18].
In crystalline diethyl ester, 3, the absence of a syn conformer
could be ascribed to molecular packing, which would not control
conformations in solution, while 2, 5, 7 and 8 with their t-butyl
groups have the anti conformation, regardless of positions of this
bulky group. Ylides 4 and 3 with small methoxy or ethoxy groups
are syn–anti mixtures in solution, as are the methyl monoester
ylides, e.g., 1 [2].
3.4. Partial atomic charges
In stabilized ylides, keto ester and diester groups are linked
electronically through ylidic resonance, but this link is absent in
the monoylidic mono and some diesters where steric effects are
important. For an anti alkyl ester group orientation of a terminal al-
kyl hydrogen towards triphenylphosphorus may be modestly
attractive [13] and Coulombic interactions between syn acyl
groups and phosphorus do not control conformations, which in
solution and in the crystal are related to sizes of substituent groups
[6,7]. Molecular packing, with formation of the more compact anti
conformer, should be important in the crystal, and affect confor-
mation, but for dilute solutions it is necessary to consider other
interactions sensitive to alkyl group size, noting that electronic
interactions are limited by a CH2 tether group .
In monoylidic diesters 2–8 in solution free rotation about the
CH2 tether group [21] should allow the nonylidic ester alkoxy
group to interact sterically with the ylidic ester alkoxy group
in a syn conformer. Interactions between the alkyl groups were
examined for hypothetical syn conformers of some of the
monoylides in Chart 2 (Supplementary data) with assumed rota-
tions of the ylidic alkoxy group and the nonylidic ester group
bound to the CH2 tether group. Structures of hypothetical syn
conformers from the B3LYP/6-31G(d) calculations were modeled
by allowing free rotation of the nonylidic ester groups without
changes in bond lengths. The interference between the hypothet-
ical ylidic and nonylidic groups are shown as Supplementary
material (Chart 2). For the hypothetical syn conformer of methyl
t-butyl ylide, 2, this rotation brings the t-butyl group unfavor-
ably close to the ylidic methoxy group, but for the anti con-
former the only near contact would be with the small acyl
Natural Population Analysis (NPA) was used to estimate local
charges for known structures or those from computation
[10a,12d,19]. The numerical significance of these charges is uncer-
tain, but they indicate factors affecting ylidic conformations. For
example, the acyl oxygen in ylidic esters has a significant negative
charge and should interact favorably with cationoid phosphorus.
However, NPA analysis indicates some anionoid character in the
alkoxy oxygen and Coulombic interactions between phosphorus
and the acyl oxygen do not control formation of syn monoylidic es-
ters in all conditions and steric factors cannot be ignored. Examples
of partial charges are given in Table 1 where the dominant anti
conformer 5 showed NPA values for O1 = ꢀ0.612; O2 = ꢀ0.665;
and P1 = 1.712 from B3LYP/6-31G(d). Assignments of NMR signals
of chemically similar groups are consistent with NPA values.
3.5. Examples of conformer identification
Provided that suitable crystals can be isolated X-ray crystallog-
raphy identifies conformations and results can be compared with
those from other methods, e.g., computation. Table 4a shows a
simple comparison of conformations from computation and X-
ray crystallography. In Table 4, ethyl, ethyl anti monoylidic 3
shows bond lengths (Å), calculated from computation:
P1@C1 = 1.73; C1AC2 = 1.42 and C2@O2 = 1.24, compared with those
observed from X-ray crystallography of 1.72; 1.40 and 1.23 respec-
tively, and torsion angles (°) from computation of 169 and 177
from X-ray crystallography.

56
F. Castañeda et al. / Journal of Molecular Structure 1034 (2013) 51–56
oxygen, and should not be so energetically unfavorable. Similarly
for the hypothetical syn t-butyl methyl ylide 5 interference be-
tween the ylidic t-butyl alkoxy group and the nonylidic methoxy
group could destabilize a syn conformer. Carboxylic ester groups
typically have the preferred Z conformation [21] and in crystal-
line syn ylidic esters the alkyl group are generally oriented away
from a nonylidic group, but in solution rotational barriers are
week [21] and rotation of the alkoxy group in the ylidic ester al-
lows it in a syn conformer to be oriented at some time in an E
conformation and toward the nonylidic ester group, with unfa-
vorable inter-alkyl contact. Similar considerations regarding
interactions between ylidic, and nonylidic groups apply to other
monoylidic aliphatic esters, but should be unimportant when
hydrogen is attached to the ylidic carbon, as in 1. These steric
effects are important in syn–anti diester ylides with different es-
ter groups where the bulkier alkoxy group has the anti confor-
mation [18]. Unfavorable interactions between alkoxy groups
could be present in hypothetical aliphatic syn–syn diesters
which, so far as we know, have not been observed in solution
or in the crystal.
Structures of hypothetical syn conformers of the t-butyl deriva-
tives 2 and 5 [6,7] were simulated by HF and DFT methods with
rotation about the bonds to the ylidic and nonylidic ester groups.
This approach was also applied to the minor syn conformer of
the monoylidic diethyl diester 3 and examples of probable unfa-
vorable encounters between ylidic and nonylidic alkyl groups are
shown as Supplementary data. These figures with rotations about
bonds involving the CH2 tether and nonylidic ester group and the
ylidic ester group in a syn conformer have no quantitative signifi-
cance but simply show how the ester groups can come into desta-
bilizing contact in solution. For anti ylides OAH interactions
between the small ester acyl oxygen and nonylidic ester alkyl
groups should not be unfavorable. These postulated interactions
in syn esters should not apply to keto ylides which have syn orien-
tations [12b,22].
Acknowledgements
The authors thank Ximena Ramírez for her valuable work in
production of this paper and are grateful to CEPEDEQ, Facultad
de Ciencias Químicas y Farmacéuticas, Universidad de Chile, for
instrumental facilities.
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.molstruc.2012.
08.051.
References
[1] (a) L.F. Wilson, J.C. Tebby, J. Chem. Soc. Perkin Trans. 2 (1972) 31;
(b) L.F. Wilson, J.C. Tebby, Tetrahedron Lett. (1970) 3769.
[2] (a) H.I. Zeliger, J.P. Snyder, H.J. Bestmann, Tetrahedron Lett. (1969) 2199;
(b) D.M. Crouse, A.T. Wehman, E.E. Schweizer, Chem. Commun. (1968) 866;;
(c) H.J. Bestmann, J.F.M. Oth, R. Merényi, H. Weitkamp, Tetrahedron Lett.
(1966) 3355;
(d) H.I. Zeliger, J.P. Snyder, H.F. Bestmann, Tetrahedron Lett. (1970) 3313.
[3] S.M. Bachrach, C.I. Nitsche, in: F.R. Hartley (Ed.), The Chemistry of
Organophosphorus Compounds, vol. 3, Wiley, New York, 1994, p. 273
(Chapter 4).
[4] (a) T.A. Albright, M.D. Gordon, W.J. Freeman, E.E. Schweizer, J. Am. Chem. Soc.
98 (1976) 6249;
(b) G.A. Gray, J. Am. Chem. Soc. 95 (1973) 7736;
(c) F.J. Randall, A.W. Johnson, Tetrahedron Lett. (1968) 2841;
(d) R. Navarro, E.P. Urriolabeitia, J. Chem. Soc. Dalton Trans. (1999) 4111;
(e) A. Lledós, J.J. Carbó, E.P. Urriolabeitia, Inorg. Chem. 40 (2001) 4913.
[5] P.D. Boyle, Acta Crystallogr. Sect. C52 (1996) 2859.
[6] R.A. Aitken, N. Karodia, P. Lighfoot, J. Chem. Soc. Perkin Trans. 2 (2000) 333.
[7] A.F. Cameron, F.D. Duncanson, A.A. Freer, V.W. Armstrong, R. Ramage, J. Chem.
Soc. Perkin Trans. 2 (1975) 1030.
[8] (a) F. Castañeda, P. Silva, C.A. Bunton, M.T. Garland, R.F. Baggio, Acta
Crystallogr. C67 (2011) 319;
(b) R. Bacaloglu, A. Blaskó, C.A. Bunton, G. Cerichelli, F. Castañeda, E. Rivera, J.
Chem. Soc. Perkin Trans. 2 (1995) 965.
[9] S.M. Bachrach, J. Org. Chem. 57 (1992) 4367.
[10] (a) W.J. Hehre, A Guide to Molecular Mechanics and Quantum Chemical
Calculations, Wavefunctions, Inc., Irvine, 2003 (Chapter 16).;
(b) C.J. Cramer, Essentials of Quantum Chemistry, Wiley, New York, 2002
(Chapter 9).
[11] (a) C.A. Bunton, P. Silva, F. Castañeda, J. Mol. Struct. 921 (2009) 6;
(b) C.A. Bunton, F. Castañeda, J. Mol. Struct. 936 (2009) 132.
[12] (a) A.D. Abell, B.M. Clark, W.T. Robinson, Aust. J. Chem. 42 (1989) 1161;
(b) F. Castañeda, C.A. Terraza, C.A. Bunton, N.D. Gillit, M.T. Garland,
Phosphorus, Sulfur, Silicon 178 (2003) 1973;
The presence of a CH3 substituent on the tether group in 6 de-
creases the anti preference, relative to that in the other monoylides
(Table 3), and reduction of the rotational mobility of the nonylidic
ester group should decrease interference in a syn conformer.
4. Conclusions
(c) F. Castañeda, C.A. Terraza, M.T. Garland, C.A. Bunton, R.F. Baggio, Acta
Crystallogr. C57 (2001) 180;
In solution and the crystal the syn conformer is dominant
only in simple monoester triphenylphosphonium ylides with
hydrogen or small alkyl groups on the ylidic carbon, consistent
with interactions between anionoid acyl oxygen and cationoid
phosphorus. However, anti conformers are dominant when
nonylidic carboxylic ester groups are linked to the ylidic carbon
(d) F. Castañeda, P. Silva, M.T. Garland, A. Shirazi, C.A. Bunton, Phosphorus
Sulfur Silicon 184 (2009) 19.
[13] (a) M. Nishio, Y. Umezawa, M. Hirota, V. Takeuchi, Tetrahedron 51 (1995) 8665;
(b) M. Nishio, M. Hirota, Tetrahedron 45 (1989) 7201.
[14] A.P. Scott, L. Random, J. Phys. Chem. 100 (1996) 16502.
[15] F. Castañeda, P. Silva, M.T. Garland, A. Shirazzi, C.A. Bunton, J. Mol. Struct. 1004
(2011) 284.
[16] J.P. Snyder, Tetrahedron Lett. (1971) 215.
[17] G.R Desiraju, T. Steiner, The Weak Hidrogen Bond in Structural Chemistry and
Biology, Oxford University Press, 2001 (Chapter 3).
[18] (a) F. Castañeda, P. Silva, M.T. Garland, A. Shirazzi, C.A. Bunton, Phosphorus
Sulfur Silicon 184 (2009) 2152;
(b) F. Castañeda, P. Silva, C.A. Bunton, M.T. Garland, R.F. Baggio, Acta
Crystallogr. C64 (2008) 405.
[19] (a) A.E. Reid, R.B. Weinstock, F. Weinholl, J. Chem. Phys 83 (1985) 735;
(b) R.F.W. Bader, Chem. Rev. 91 (1991) 893.
by
a CH2 tether group which limits electronic effects and,
depending on sizes of the alkoxy groups, allows unfavorable
interactions between nonylidic and ylidic ester groups in syn
conformers. Conformations in solution are conveniently deter-
mined by examination of 1H and 13C NMR spectra and in the
crystal by X-ray crystallography. With large ester groups confor-
mations are the same in the crystal and in solution, but with
smaller groups there may be one conformer in the crystal which
is only dominant in solution. Molecular packing should be
important in the crystal, while the extent of preference for some
anti conformers in solution depends on sizes of the ester groups.
Computed bond lengths and angles of these ylides are similar to
values from X-ray crystallography. Stretching vibrations of ylidic
acyl groups in IR spectra are sensitive to conformation and com-
puted frequencies for anti groups are higher than for syn and
intensities of the latter are very weak, possibly due to interac-
tions with the triphenylphosphonium group.
[20] (a) F. Castañeda, C. Aliaga, C.A. Bunton, M.T. Garland, R. Baggio, Acta
Crystallogr. C61 (2005) 496;
(b) F. Castañeda, C. Jullian, C.A. Bunton, M.T. Garland, R. Baggio, Acta
Crystallogr. C63 (2007) 267;
(c) F. Castañeda, C. Aliaga, M.T. Garland, J. Hu, N.D. Gillitt, C.A. Bunton,
Phosphorus Sulfur Silicon 182 (2007) 151.
[21] (a) E.I. Eliel, S.H. Wilen, Stereochemistry of Organic Compounds, Wiley, New
York, 1994. pp. 635 and 641.;
(b) T.B. Grindley, Tetrahedron Lett. 23 (1982) 1757;
(c) C.E. Blom, H.H. Günthard, Chem. Phys. Lett. 84 (1981) 267;
(d) K.B. Wiberg, K.E. Laidig, J. Am. Chem. Soc. 100 (1987) 5935.
[22] F. Castañeda, G.I. Recabarren, J. Hu, N.D. Gillit, C.A. Bunton, Phosphorus Sulfur
Silicon 178 (2003) 2505.