Multinuclear magnetic resonance spectroscopy and density function theory calculations for the identification of the equilibrium species in THF solutions of organometallic complexes suitable as electrolyte solutions for rechargeable Mg batteries

Article abstract of DOI:10.1021/om300865a

We present a multinuclear nuclear magnetic resonance (NMR) and density functional theory (DFT) study of electrolyte solutions based on organometallic complexes with aromatic ligands. These solutions constitute a unique electrolyte family possessing a wide electrochemical window, making them suitable for rechargeable magnesium batteries. In our previous study we identified equilibrium species in the solutions based on a combination of Raman spectroscopy and single-crystal XRD analyses,(1) and herein we extend our studies to include multinuclear NMR analyses. These solutions are comprised of the metathesis reaction products of MgCl_2-xPh_x and AlCl_3-yPh_y in various proportions, in THF. In principle, these reactions involve the exchange of ligands between the magnesium and the aluminum based compounds, forming ionic species and neutral molecules, such as Mg₂Cl₃⁺·6THF, MgCl₂· 4THF and AlCl_4-yPh_y^- (y = 0-4). The identification of the solution phase species from the spectroscopic results is supported by spectral analyses of specially synthesized reference compounds and DFT quantum-mechanical calculations. The current approach reveals new aspects about the NMR shift of the organometallic complexes and, in particular, facilitates differentiation between ionic and neutral species.

Full text of DOI:10.1021/om300865a

Article
pubs.acs.org/Organometallics
Multinuclear Magnetic Resonance Spectroscopy and Density
Function Theory Calculations for the Identification of the Equilibrium
Species in THF Solutions of Organometallic Complexes Suitable As
Electrolyte Solutions for Rechargeable Mg Batteries
Nir Pour, Yossi Gofer, Dan T. Major, Keren Keinan-Adamsky, Hugo E. Gottlieb, and Doron Aurbach*
Department of Chemistry and the Lise Meitner-Minerva Center of Computational Quantum Chemistry, Bar-Ilan University,
Ramat-Gan 52900, Israel
S
* Supporting Information
ABSTRACT: We present a multinuclear nuclear magnetic
resonance (NMR) and density functional theory (DFT) study
of electrolyte solutions based on organometallic complexes
with aromatic ligands. These solutions constitute a unique
electrolyte family possessing a wide electrochemical window,
making them suitable for rechargeable magnesium batteries. In
our previous study we identified equilibrium species in the
solutions based on a combination of Raman spectroscopy and
single-crystal XRD analyses,1 and herein we extend our studies
to include multinuclear NMR analyses. These solutions are
comprised of the metathesis reaction products of MgCl2−xPhx and AlCl3−yPhy in various proportions, in THF. In principle, these
reactions involve the exchange of ligands between the magnesium and the aluminum based compounds, forming ionic species
+
and neutral molecules, such as Mg2Cl3 ·6THF, MgCl2·4THF and AlCl4−yPhy− (y = 0−4). The identification of the solution phase
species from the spectroscopic results is supported by spectral analyses of specially synthesized reference compounds and DFT
quantum-mechanical calculations. The current approach reveals new aspects about the NMR shift of the organometallic
complexes and, in particular, facilitates differentiation between ionic and neutral species.
ligands between the magnesium and the aluminum based
compounds, forming ionic species and neutral molecules, such
INTRODUCTION
■
In the early 2000s, we presented the first prototype of a
rechargeable magnesium battery.2 This battery was realized
largely because of the development of novel complex
organometallic solutions of magnesium salts of the nominal
formula RxMgCl2−x R′yAlCl4−y− (R, R′ = n-butyl and/or ethyl, x
= 0−2, y = 0−4) in THF. The standard solution was selected to
be the product of the metathesis reaction between MgBu2 and
AlCl2Et at a ratio of 1:2 forming, technically, Mg(AlCl2BuEt)2.
This solution displayed favorable electrochemical character-
istics, with an electrochemical stability window of ∼2.2 V (vs
Mg), specific conductivity of a few 10−3 Ω−1 cm−1, and
magnesium electrodeposition at 100% Coulombic efficiency.
Nonetheless, such organometallic solutions suffer from a
limited stability window, which hampers the possibility to
couple the Mg anode with high voltage cathodes. It was shown
that the weakest link that limits the oxidation stability is a
carbon−aluminum bond that breaks electrochemically through
a β-H-elimination.3,4 Thus, we theorized that replacing the
aliphatic ligands with phenyl, for which β-hydrogen elimination
is impossible, will lead to significantly more robust organo-
metallic complexes.
+
as Mg2Cl3 ·6THF, MgCl2·4THF and AlCl4−yPhy− (y = 0−4).
These electrolytes will henceforth be referred to as “APC”. The
electrochemical stability of these solutions was measured by
cyclic voltammetry and found to display a window of ∼3.3 V
(vs Mg),1 proving that the basic hypothesis is correct.
In a previous paper1 we presented a comprehensive analytical
work, identifying nearly all the species existing in these
solutions. In the current article we focus on multinuclear
NMR studies in conjunction with DFT calculations, to
complement the previous analyses of the organometallic
complexes in THF. This combined approach allows identi-
fication of the entire range of species existing in the solutions
(with the exception of trace compounds). In particular, we
successfully differentiate between charged and neutral mole-
cules. The excellent agreement between the calculated spectra
and the trends obtained empirically resulted in the complete
determination of the solutions composition.
This new family of solutions is comprised of the metathesis
reaction products of MgCl2−xPhx and AlCl3−yPhy in different
proportions, in THF. These reactions involve the exchange of
Received: September 7, 2012
© XXXX American Chemical Society
A
dx.doi.org/10.1021/om300865a | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Several important results are presented in Table 1. (1) For
each molecule (Al-based and Mg-based) the ipso carbon
chemical shift is the largest, followed by the ortho carbon. (2)
The meta and para carbons are always shifted significantly less
than the ipso and ortho ones and show much smaller variability.
(3) The ipso carbon shifts for the ionic species increase with
increasing phenyl to Cl content ratio: (AlCl3Ph−) 146.0 <
RESULTS AND DISCUSSION
■
Since the multinuclear NMR analysis provides vast amounts of
information, we found it logical to divide it into two sections,
namely, 13C and 27Al NMR. The integration of the data gained
from 13C and 27Al NMR provides detailed information on the
composition of the Al-based compounds. Thus, the current
work, which identifies Al-based compounds, in combination
with our previous identification of the remaining species
employing Raman spectroscopy, forms a complete picture of
the composition of this novel family of solutions.
−
−
−
(AlCl2Ph2 ) 147.5 < (AlClPh3 ) 153.2 < (AlPh4 ) 156.4. (4) A
similar trend is also observed for the ortho carbons, albeit with
significantly smaller values. The more organic ligands (i.e., less
Cl), the higher the chemical shift: (AlCl3Ph−) 134.0 <
DFT Calculated and Measured 13C NMR. Previous 13C,
1H, 27Al and 25Mg NMR analyses revealed that the information
extractable is restricted to the Al coordination number, i.e., the
number of organic vs inorganic ligands.4−7 In fact, very little
−
−
−
(AlCl2Ph2 ) 134.3 < (AlClPh3 ) 134.7 < (AlPh4 ) 134.9. (5)
The ipso and the ortho shifts for the neutral molecules also
followed the same trend, with a stronger effect for the ipso
position. (6) The opposite trend is obtained for the chemical
shifts of the para carbons, and a similar, although much weaker
trend, for the meta carbons.
1
could be gained from 13C, H and 25Mg NMR analyses. One
important limitation of the 27Al NMR spectra is that the Cl and
THF ligands affect the chemical shift in a similar manner. Thus,
with the widths of the typical spectra it was impossible to
differentiate between several pairs of molecules possessing the
same number of phenyl groups, yet a different THF and Cl
The results for MgClPh are also presented in Table 1. Its
purpose was both for the validation of the correlation between
the calculated spectra and the empirical data, and to serve as an
identifier for phenyl ligand containing species in case they exist
in the solution. The data show important features: The whole
set of peaks appear with appreciably higher chemical shifts, the
ipso carbon having the highest shift and the para the lowest.
Compared to the equivalent carbons on the Al species, the para
and meta carbon shifts are very close (difference of 2−3 ppm).
However the ortho and ipso shift is much higher, as expected on
the basis of the higher electropositivity of Mg and the 13C
NMR date reported in the literature for organo-phenyl
compounds.8
−
composition, e.g., AlPh3·THF and AlClPh3 . In such cases, the
different types of composition reflect neutral (THF) vs ionic
molecules (Cl). Interestingly, the DFT calculations results for
the 13C NMR revealed coherent chemical-shift differences
between such pairs of neutral and ionic species, which allowed,
for the first time, to distinguish between them.
As was described in the previous account, correct selection of
the basis-set and the reliability of the calculated molecule
structures and their reconstructed spectra were ascertained by
comparison between empiric and calculated data for selected
pure compounds.
Table 1 below displays all the phenyl-ligand bearing Al and
Mg species that can possibly exist in APC-type solutions and
The DFT calculated 13C NMR data for the above species
suggests that a reliable database compiling the related
experimental data may be utilized for spectroscopic identi-
fication of the various species in the family of APC solutions.
Specifically, the chemical shifts obtained from the calculations
suggest that reliable identification of the relevant solution
species can be realized by using a proper library of the ipso and
para NMR peak shifts as they prove to be species specific. This
feature adds cross-measurement certainty to Raman spectros-
copy, as described in the previous paper. Moreover, the ipso
chemical shifts complement the capability, unattained by
Raman spectroscopy, to differentiate between the “equivalent”
neutral versus anionic pairs (neutral and anionic species
containing the same number of phenyl ligands).
Table 1. Calculated 13C NMR Peak Positions of ipso, ortho,
meta and para Carbons of the Phenyl in the Al and Mg
Species Using B3LYP/6-31+G(d,p)
carbon [ppm (vs TMS)]
molecule
ipso
ortho
meta
para
AlCl3Ph−
146.0
136.3
147.5
140.5
153.2
144.3
156.4
165.8
134.0
134.6
134.3
134.4
134.7
135.7
134.9
142.0
124.0
124.1
122.6
124.1
122.6
124.1
122.5
127.0
125.0
126.2
123.2
125.0
121.7
124.8
120.6
126.0
AlCl2Ph·THF
−
AlCl2Ph2
AlClPh2·THF
−
Comparison of the para, meta and ortho carbon shifts for the
ionic and neutral molecules with identical number of phenyl
groups reveals that their chemical shifts are nearly indistin-
guishable. For example, the chemical shift for the ortho carbon
AlClPh3
AlPh3·THF
−
AlPh4
MgClPh
−
in AlCl2Ph2 is 134.3 ppm, whereas for the corresponding
“equivalent” neutral molecule, AlClPh2·THF, the shift is 134.4
ppm. This phenomenon has been demonstrated previously5
and is rationalized on the basis of the similar effect THF and Cl
have on the NMR chemical shifts of the phenyl para, meta and
ortho carbons. In contrast, the ipso carbon shifts exhibit unique
and significant differences between the neutral and ionic
“equivalent” pairs. The shift for AlPh3·THF is 144.3 ppm, while
their corresponding 13C NMR chemical shift obtained from the
DFT calculations.
The phenyl ligands possess four sets of carbons that can be
distinguished by NMR. Each of the carbons is affected
differently by the Al core. The complete set of data for each
of these carbons is listed below.
The peaks for phenylmagnesium chloride are also added.
This molecule is the only organomagnesium species ever
detected in our solutions under the limitation of our NMR and
Raman spectroscopies. It is used to estimate the reliability of
the DFT calculations of chemical shifts for our compounds.
The calculated TMS 13C NMR peak was used as a reference
(see Experimental Section).
−
for its “equivalent” ionic pair, AlClPh3 , it is 153.2 ppm: a
−
difference of 8.9 ppm. For the other two pairs, AlCl2Ph2 /
AlClPh2·THF and AlCl3Ph−/AlCl2Ph·THF, the shift differ-
ences are 7.0 and 9.7 ppm, respectively.
As displayed by the DFT calculation, each of the para, meta,
ortho and ipso carbons has a range of chemical shifts centered
around a unique value. These shifts are very informative as they
B
dx.doi.org/10.1021/om300865a | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
depend in a consistent manner on the nature of the molecule
(ionic, neutral, number of phenyl groups, etc.). Experimental
13C NMR limitations preclude the use of this tool to
differentiate between the different molecules through the para
and meta peaks. Fortunately, the spectral features for the ortho
and ipso carbons were obtained with sufficient quality to allow
distinguishing between the complexes.
Table 2. Calculated (B3LYP/6-31+G(d,p)) and Measured
13C NMR Comparison of Phenyl Carbons in MgClPh (Shift
vs C3 of THF)
MgClPh 13C NMR [ppm (vs TMS)]
Cposition
calc
ipso
ortho
meta
para
165.8
170.0
4.2
142.0
140.6
−1.4
127.0
125.5
−1.5
126.0
123.9
−2.1
We measured the 13C NMR spectra for a wide variety of
THF solutions containing either single, known compounds or
the reaction products of MgClPh with AlCl3−xPhx The
empirical shifts obtained followed the expected trend (from
the DFT calculated spectra): the higher the number of phenyl
ligands on the aluminum, the larger the chemical shift
(downfield).
measured
Δ
was also obtained in all the comparisons between the 13C NMR
spectra obtained from the DFT and the measured ones (vide
infra).
MgPh2 has frequently been used as a reactant in electrolyte
synthesis.1 the 13C spectrum for this molecule was indistin-
guishable from the one obtained for MgClPh (see Figure S1,
Supporting Information). Thus, NMR measurements are not
practical for differentiation between MgClPh and MgPh2, and
only elemental analysis achieved clear identification. (Raman
measurements were also inadequate in this respect because of
strong fluorescence interfering with the analysis).
Because of the lack of essential literature data on the spectral
properties of many of the relevant molecules, we first compiled
a spectral database by measuring the multinuclear NMR spectra
for solutions containing known compounds. The empirical data
was also correlated with the DFT calculated spectra. In a few
cases, where it was possible, pure compound solutions in THF
were measured for the library. In other cases, where a great
excess of a specific molecule was unequivocally identified by
other spectroscopies (27Al NMR, Raman, SC-XRD), the 13C
spectrum was also measured for the same purpose. In all other
cases the measured spectrum had to be carefully analyzed, as it
contained an equilibrium mixture of several molecules at an
unknown molar ratio.
Figure 2 displays the 13C NMR spectrum for AlPh3·THF.
This is one of the two molecules we were able to obtain as a
Figure 1 presents the spectrum obtained in the spectral
region between 120−170 ppm. In this region there are number
Figure 2. 13C NMR spectrum of 0.25 M AlPh3·THF in THF (shift vs
TMS).
Figure 1. 13C NMR spectrum of 0.5 M MgClPh in THF (shift vs
TMS).
single component in THF and thus was valuable for the
compilation of a reliable spectral database. Raman spectroscopy
for the solution proved that this solution contains only one Al-
based specie, and 1H NMR proved that this molecule is
of significant peaks that are related to the different carbons of
the phenyl group. The peaks at 123.9, 125.5, 141, and 170 ppm
correspond to the para, meta, ortho and ipso position,
respectively. Table 2 presents a comparison between the 13C
NMR chemical shifts obtained with DFT and the measured
ones for MgClPh. In both cases we used the TMS 13C NMR
peak position as the reference for the comparison (see
Experimental Section).
Table 2 shows that once an appropriate reference data point
is selected, there is a very good correlation between the DFT
results and the measured ones. The computed trend is in
reasonable agreement with the observed trend, and the
chemical shifts are of similar magnitude. Such a correlation
1
−
AlPh3·THF (and not AlClPh3 ).
The spectrum of this molecule was very valuable as it was the
first verified neutral molecule in of the APC family to be
identified.
Figure 2 presents the 120−160 ppm region and a
magnification of the 150 ppm region (insert). There are a
number of significant peaks that relate to the different carbons
on the phenyl group. The peaks at 126.9, 127.1, 138.2, and
147.9 ppm, correspond to the para, meta, ortho and ipso
carbons, respectively. As with the case of MgClPh, a constantly
increasing shift was also observed in the DFT calculation
presented in Table 1.
C
dx.doi.org/10.1021/om300865a | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Figure 3. 13C NMR spectrum of APC 1−4 (0.25 M AlCl3 and 1 M MgClPh) in THF. Insert: Magnified 160−175 ppm region (shift vs TMS).
Figure 4. 13C NMR spectrum of 0.4 M APC (0.4 M AlCl3 and 0.8 M MgClPh) in THF (shift vs TMS).
The spectra presented below were measured in various
solutions, in which only partial molecular identifications were
obtained previously from Raman spectroscopy. In addition to
the partial knowledge regarding the solution chemistry, all these
solutions are expected to include multiple equilibrium species.
Because of the lack of resolution and the richness in peaks, in
most cases it was not possible to differentiate between the para
and meta carbon peaks of the different species. Therefore, our
analysis has to focus on the ipso and ortho carbons.
Unfortunately, the ipso carbon signal was usually very weak
and in some cases unobservable. Figure 3 displays the spectrum
of APC 1−4; this solution was synthesized from 0.25 M AlCl3
and 1 M MgClPh as described in the Experimental Section.
D
dx.doi.org/10.1021/om300865a | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Figure 5. 13C NMR spectrum of 0.4 M APC + 0.8 M LiCl (0.4 M AlCl3 and 0.8 M MgClPh + 0.8 M LiCl) in THF (shift vs TMS).
From the Raman spectrum of this solution we identified the
main Al-species in this solution to be AlPh4 .
basis of the understanding gained to date4,10 and on the DFT
results, these peaks can be assigned as follows: 136.8 ppm to
−
−
−
As one can see, this spectrum is much richer in peaks, as
expected from solution containing several species in equili-
brium. Some of the species in this case, however, have very low
concentration. Figure 3 presents the 120−170 ppm region and
magnifications of two important regions. In this spectrum we
can see the peaks already presented in Figure 1, related to
MgClPh (at ∼123, ∼125.5, ∼141, and 170 ppm). The presence
of MgClPh and its associated peaks is expected, as this solution
was synthesized with an excess of MgClPh. Apart for the
features associated with this already identified molecule, four
new peaks are clearly seen, at 124.0, 125.4, 139.2, and 159.0
ppm. On the basis of the information obtained from Raman
spectroscopy, we ascribe these peaks to the para, meta, ortho
AlCl3Ph−, 137.5 ppm to AlCl2Ph2 , 138.2 ppm to AlClPh3
(see Figure 2) and 139.2 ppm to AlPh4− (see Figure 3). On the
basis of the ipso carbon features, which will be thoroughly
discussed later, and supported by ionic conductivity measure-
ments, we conclude that these species are all ionic. In Figure 4
one can also observe the 6-fold spin coupled splitting of the
−
AlPh4 peak at 160 ppm. Although weak, but still discernible,
one can identify the ipso carbon peaks of AlClPh3− at 155 ppm,
and the ipso carbon peaks for AlCl2Ph2− and AlCl3Ph− at ∼153
ppm.
Figure 5 presents the 13C NMR spectrum for 0.4 M APC
solution containing 0.8 M LiCl (LiCl is added as electro-
chemical ionic additive1). On the basis of Raman analysis, this
−
−
and ipso carbons of AlPh4 , respectively. A weak peak is also
solution was determined to contain AlClPh3 that is the main
observed at 128.4 ppm. This peak corresponds to benzene.9
Benzene is a common breakdown product of these organo-
metallic molecules; therefore, the presence of its traces in these
solutions is not unexpected. A unique feature in the spectrum is
the location of the ipso carbon, at 160 ppm. Because of its direct
bond to 27Al, it clearly displays the 6-fold splitting, resulting
Al-based species. Figure 5 displays the measured 13C NMR
spectrum of 0.4 M APC solution + 0.8 M LiCl in THF.
In the 140 ppm region, the same 4 peaks presented in Figure
4 are also observed. However, in this solution, the peak at 138.2
ppm (A) is much stronger than the other three. This peak is
−
assigned to the ortho carbon for AlClPh3 . From the peak
−
from the spin coupling with the aluminum nucleus. For AlPh4
integration it is evident that this molecule exists in excess.
Comparison between this spectrum and that for AlPh3·THF in
Figure 2 reveals the difference between the chemical shifts for
the “equivalent” ionic and the neutral molecules. A unique and
significant spectral difference is seen, as predicted from the
DFT calculations, in the ipso carbon chemical shift, in the 150
ppm region. From Figure 2, it is found that the ipso carbon
peak for AlPh3·THF is at 147.9 ppm, whereas the “equivalent”
these splittings are particularly visible because of the high
symmetry of the molecule. This splitting is not seen in the ipso
−
−
carbon features of the other molecules (AlClPh3 , AlCl2Ph2
etc.). Because of lower symmetry of these species, the peaks are
too weak and broad to show the fine splitting details.
Figure 4 presents the 13C NMR for 0.4 M APC solution (this
solution was synthesized from 0.4 M AlCl3 and 0.8 M MgClPh
in THF). From previous Raman spectroscopy, it was
established that this solution contains a wide range of Al
species.1
When examining the peaks associated with the ortho carbons,
in the 135−140 ppm region, it can be determined that there are
4 different species in the solution (marked A to D). On the
−
anion, AlClPh3 , shows the ipso carbon peak at 155.6 ppm. The
experimental difference in position between these two peaks is
about 9 ppm, close to the DFT results that predicted a
difference of 10 ppm. These results complete the identification
of the variety of species present in APC-type solutions by 13C
NMR, including the possibility to distinguish between neutral
E
dx.doi.org/10.1021/om300865a | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
and ionic molecules. Further corroboration of the peak
assignments for the spectra in Figures 3, 4, and 5 are given
in the Figures S2−S4 (Supporting Information).
ipso of the ionic molecules and ipso of the “equivalent” neutral
molecules) obtained from the DFT calculations and the 13C
NMR measurements.
From Table 4 it can clearly be seen that the DFT calculations
reproduce fairly well the 13C NMR shifts for the phenyl Mg and
Al organometallic compounds. In some cases, the difference
between the calculated values and the measured are as low as
2.6 ppm.
Two practical conclusions can be derived from meticulous
examination of the empirical 13C NMR spectra for the APC-
family of organometallic complex solutions: (1) The peak
position of the ipso carbons can be utilized to distinguish
between neutral and ionic molecules, otherwise equivalent both
in 27Al NMR and Raman spectroscopy. (2) The peaks position
of the ortho carbon can be used to determine the number of
phenyl ligands bound to the aluminum.
DFT Calculated and Empirical 27Al NMR. Previous
studies of a homologue series of aliphatic organometallic
complexes,10 and related work,11−14 have shown that 27Al NMR
contains invaluable structural information. In many cases it can
be used to shed light on structural and chemical aspects of
aluminum-based molecules. This is despite the quadrupolar
27Al nucleus,6,7 which causes peak broadening. It has been
shown,15 as a rule of thumb, that the coordination number of
the aluminum core atom is the primary influencing parameter
on the chemical shift. To a lesser degree, but still significantly,
the ligand’s nature influences the chemical shift in an additive
and predictable manner: while inorganic ligands, like Cl−, drive
the chemical shift upfield, organic ligands drive the chemical
shift downfield. In this respect, it was previously shown that
highly polar donor ligands, such as THF, influence the chemical
shift in an identical manner to Cl−. These phenomenological
features combined with DFT calculations and Raman spec-
troscopy facilitate complete identification of all species in these
complex, multicomponent organometallic equilibrium solu-
tions.
Table 3 summarizes the measured 13C NMR peaks positions
of the ipso and ortho carbons as obtained from Figures 3, 4, 5
Table 3. Measured 13C NMR Peaks Positions of ipso and
ortho Carbons of the Phenyl in the Al and Mg Species
carbon [ppm (vs TMS)]
molecule
ortho
ipso
MgClPh
MgPh2
140.6
140.6
139.2
138.2
137.3
136.6
n/a
170.0
170.0
159.0
155.6
152.9
152.9
n/a
−
AlPh4
−
AlClPh3
−
AlCl2Ph2
AlCl3Ph−
neutral
AlPh3·THF
138.2
147.9
In contrast, the 25Mg NMR spectra are much less informative
and usually too broad and weak to be of help in structural
determination. In addition, the 25Mg NMR spectra demonstrate
much smaller chemical shift variations.
and Figures S2−S4 (Supporting Information). This is used in
the following tables for the comparison between the measured
and the calculated peaks positions.
The DFT results presented at the beginning of the paper
predicted that the more phenyl groups attached to the Al, the
higher the ppm value of the ipso and ortho carbon. Moreover,
the DFT calculations predicted that the corresponding
chemical shifts for the organomagnesium species will be
downfield. It can be seen from Table 3 that the trends in the
measured chemical shifts for all the above organometallic
compounds follow those predicted by the DFT calculations.
In order to highlight the correlation between the DFT
calculations results and the measured spectra, Table 4 compares
the chemical shift (vs TMS) of a specific carbon type (ortho,
Table 5 lists the 27Al NMR shifts obtained from the DFT
calculations (B3LYP/6-31+G(d,p)) for the various molecules,
as described before. The peak for AlCl3THF was taken as a
reference.
Table 5. Calculated 27Al NMR Chemical Shifts for the Al
Species Using B3LYP/6-31+G(d,p)
DFT calculated 27Al NMR shift
molecule
chemical shift [ppm]
AlCl3·THF
AlCl4
AlCl3Ph−
0.0
0.2
−
Table 4. Calculated (B3LYP/6-31+G(d,p)) and Measured
13C NMR Chemical Shifts (vs TMS) of ipso and ortho
Carbons of the Al-Species
17.5
19.4
32.3
36.4
43.9
52.1
28.3
AlCl2Ph·THF
−
AlCl2 Ph2
AlClPh2·THF
ionic molecules: ortho carbon
−
AlClPh3
−
−
−
AlPh4
AlClPh3
AlCl2Ph2
AlCl3Ph−
AlPh3·THF
−
calculated
measured
Δ
134.7
139.2
4.5
134.7
138.2
3.5
134.3
137.3
3
134
136.6
2.6
AlPh4
ionic molecules: ipso carbon
−
−
−
The results of the DFT calculations listed in Table 5 show a
clear trend in the chemical shifts. As was shown before, the
chemical shift depends on the Ph:Cl− ratio for all the
tetracoordinated molecules. Moreover, THF as a ligand
influences the chemical shift in a qualitatively similar way to
the inorganic ligand, but quantitatively different. As a general
rule, it can be said that the higher the Ph:Cl− ratio, the higher
the chemical shift for these tetracoordinated organometallic Al
based coordination species.
AlPh4
AlClPh3
AlCl2Ph2
AlCl3Ph−
calculated
measured
Δ
155.4
159
3.6
153.2
155.6
2.4
147.5
152.9
5.4
146.0
152.9
6.9
neutral molecules: ipso carbon
AlPh3·THF
AlClPh2·THF
AlCl2Ph·THF
calculated
measured
144.3
147.9
140.5
n/a
136.3
n/a
F
dx.doi.org/10.1021/om300865a | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
−
Interestingly, AlPh4 is an outlier in this respect. This
molecule’s 27Al NMR peak appears at a significantly lower ppm
than expected on the basis of the constantly increasing chemical
shift with increasing number of phenyl ligands. This
phenomenon is still not fully understood. However, since this
phenomenon is both predicted by the DFT calculations and
obtained in the NMR measurement, it suggests that it reflects
subtle changes in the electronic structure of the tetracoordi-
nated, all organic ligand, molecule. It is important to note here
that the very same phenomenon was revealed for the series of
organo-halo aliphatic (ethyl, butyl) compounds, the DCC
family of electrolytes.15
Also, the differences in the chemical shifts obtained for
“equivalent” neutral ionic pairs, show a constant trend: the
chemical shift for the neutral molecule is usually higher than
that of its “equivalent” anionic. For example, the 27Al peak for
AlCl3Ph− is 17.5 ppm, while its neutral “equivalent”,
AlCl2−Ph·THF is at 19.4 ppm. The pair AlCl3·THF and
AlCl4 is an exception to this trend, although the differences
in the chemical shifts are exceedingly small. As will be shown
later, most of the trends obtained above from the DFT
calculations were observed in the measured spectra for the
molecules that could be synthesized.
−
+
Figure 7. Solid-state 27Al NMR spectrum of the AlPh4 Mg2Cl3 .
All the other 27Al NMR spectra display wider and weaker
peaks, as presented in Figure 8, below.
The spectrum of AlPh3·THF in THF in Figure 8a exhibits a
single peak at 142.3 ppm with a weak, but noticeable shoulder
to the right, presumably associated with traces of the reactant
AlCl3. Figure 8b displays the spectrum obtained from the APC
2−1 solution. This solution was shown, on the basis of Raman
and 13C NMR (Figure S3, Supporting Information), to contain
Figure 6 displays two 27Al NMR spectra, (a) for 0.25 M AlCl3
in THF and (b) APC 1−4.
a mixture of AlCl3Ph−, AlCl4 and AlCl3·THF. The peak for
−
AlCl3·THF can clearly be seen in the 60 ppm region.6 The 27Al
−
NMR for AlCl4 is reported to be located at around 102−108
ppm,16 and thus it can be inferred that the wide peak at 102.5
ppm is a superposition of the two peaks associated with
AlCl3Ph− (∼107.0 ppm, Figure S5, Supporting Information)
−
and AlCl4 (∼100.0 ppm, Figure S5, Supporting Information).
Figure 8c displays the spectrum obtained from 0.25 M APC,
the standard working electrochemical solution. It was shown
previously by Raman and 13C NMR to contain a very wide
array of species in equilibrium.1 This spectrum is displayed in
order to emphasize the difficulty in the analysis of these
complex organometallic compounds solutions. Figure 8
illustrates the need for the combined work of multinuclear
NMR and Raman spectroscopy, with the aid of high quality
DFT calculations for the reliable peak associations. As can be
seen in Figure 8, apart from two clearly identifiable peaks, the
spectrum displays very wide and smeared overlapping features
associated with a variety of species. Additional spectra are
displayed in Figures S5 and S6 (Supporting Information)
supporting the remaining peak assignments.
Figure 6. 27Al NMR spectra of (a) 0.25 M AlCl3 in THF and (b) APC
1−4 (AlPh4 Mg2Cl3 in THF).
−
+
0.25 M AlCl3 in THF contains mainly the tetracoordinated
AlCl3·THF and its measured NMR peak is at 60.4 ppm (Figure
6a).6 Figure 6b, presents the 27Al NMR spectrum for APC 1−4
solution. The 13C NMR spectrum for the same solution was
presented in Figure 3. The main Al species in this solution was
Table 6 summarizes and compares all the measured and DFT
calculated 27Al NMR peaks for the above organometallic
compounds. For simplicity the NMR shifts in Table 6 are
referenced to AlCl3.
−
proven to be AlPh4 . Thus the peak at 131.9 ppm is assigned to
−
−
the AlPh4 . Both AlPh4 and AlCl3·THF display sharp and
strong peaks. Because of its high symmetry, AlPh4− displays the
sharpest and highest peak among all the current 27Al NMR
spectra.
Interestingly, despite the relatively large difference between
the calculated and the measured chemical shifts, the same
general trend can be observed: more organic ligands attached to
the Al results in a higher chemical shift. Also, the exception to
−
Further confirmation of the existence of AlPh4 in the APC
−
1−4 solution came from a complementary measurement of
solid state 27Al NMR. As was described in a previous paper,1
crystals precipitated from APC 1−4 solutions were identified as
the trend related to the AlPh4 chemical shift is also predicted
by the DFT calculations. Increasing the basis set with the
B3LYP functional does not significantly change the chemical
shifts. However, slightly better agreement with experiment is
obtained with the mPW1PW91 functional with a larger basis
set.17,18 We note that calculating 27Al chemical shifts is
challenging for modern DFT methods,19 although the errors
−
+
AlPh4 Mg2Cl3 by SC-XRD. Figure 7 below presents the solid
state 27Al NMR spectrum of these crystals.
As in the case of the solution spectra, solid state 27Al NMR
for AlPh4− shows very sharp peaks owing to its high symmetry.
G
dx.doi.org/10.1021/om300865a | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Figure 8. 27Al NMR spectra of (a) 0.25 M AlPh3·THF, (b) APC 2−1 (0.25 M AlCl3 + 0.125 M MgClPh), and (c) 0.25 M APC (0.25 M AlCl3 + 0.5
M MgClPh). All spectra were measured in THF.
Table 6. Calculated and Measured 27Al NMR Comparison of
the Al-Species
EXPERIMENTAL SECTION
■
All the syntheses and electrochemical measurements were carried out
under pure argon atmosphere (99.999%) in a M. Braun, Inc., glovebox
(less than 1 ppm of water and oxygen).
27Al NMR (ppm vs AlCl3)
AlCl4
0.2
AlCl3Ph− AlCl2Ph2 AlPh3·THF AlPh4
−
−
−
Chemicals and Syntheses. All electrolyte solutions were
synthesized according to the same basic procedure: First, AlCl3
(Aldrich, anhydrous, 99.999%) in THF (Aldrich, 99.9%, less than 20
ppm water) solution was prepared in the desired concentration by very
slow addition to the vigorously stirred solvent. Then, this solution was
added dropwise to a predetermined quantity of 2 M phenylmagnesium
chloride (MgClPh) solution in THF (Aldrich). Both reactions are very
exothermic. The resulting solution was stirred for additional 16 h or
more (at room temperature).
calculated
a
17.5
21.8
47.0
32.3
35.3
60.0
52.1
n/a
82.3
28.3
n/a
72.0
(ppm)
calculated
6.9
b
(ppm)
measured
(ppm)
42−48
a
b
B3LYP/6-31+G(d,p). mPW1PW91/Aug-cc-pVTZ.
The family of electrolytes containing phenyl ligands in various
proportions are referred to in this paper as “APC” (all phenyl
complex). Electrolyte concentrations are reported in terms of the
aluminum-based species concentration. If not indicated otherwise, the
AlCl3 to MgClPh ratio is 1:2. Thus, a 0.25 M APC solution is
synthesized by reacting equal quantities of 0.25 M AlCl3 and 0.5 M
MgClPh in THF. The standard electrolyte, to which others are
compared, is 0.25 M APC, as it exhibits favorable electrochemical
properties. In cases where solutions were prepared with AlCl3 and
MgClPh at ratios other than 1:2, they are denoted as APC m:n
solutions (m and n are the relative amounts of AlCl3 and MgClPh,
respectively). Unless otherwise noticed, the AlCl3 concentration that
was used in all of the solutions was 0.25 M.
herein are somewhat larger than what is typically observed for
27Al species. We also computed the 1J(27Al,13C) coupling
−
constants for AlPh4 . The computed value using mPWPW91/
6-31+G(d,p) is 74.9 Hz (while the measured value from our
experiments was 96.0 Hz), comparable to the computed value
for Al(t-Bu)3 of 76.3 Hz.20 In spite of the significant differences
between the calculated and the empirical results for the
chemical shifts, the importance of the 27Al NMR analysis to
fully unravel the existing species of the metallic−organic
solutions is demonstrated.
CONCLUSIONS
Synthesis of AlPh3·THF Reference Compound.21 A 2 M THF
solution of phenylmagnesium chloride (22.5 mmol) was slowly added
to a 1.2 M solution of AlCl3 (1.0 g, 7.5 mmol) in THF (6.25 mL) at 0
°C. The solution was then stirred for 14 h at rt, after which the solvent
was removed under reduced pressure. The solid residue (mainly
MgCl2) was separated and removed by toluene extraction and
centrifugation. The white powder left after toluene removal contained
mainly AlPh3·THF with less than 10% (molar) MgCl2. NMR indicated
a Ph:THF ratio of 3:1.1
■
In this paper we completed the identification of the equilibrium
species in the APC family of electrolyte solutions formed by
reacting MgClPh and AlCl3 in THF. Multinuclear NMR
combined with Raman spectroscopy, elemental analysis and
DFT calculations were utilized to carry out this complex
endeavor. Both the Al and Mg based species, neutral and ionic,
were identified, and their spectral features were reported. It was
shown that for the complete elucidation of the various mixtures
of equilibrium compounds, the combined use of 13C, 27Al NMR
and Raman spectroscopic measurements is crucial. Each of the
spectroscopic techniques is used for the identification of
particular pieces of the complete picture, and only the
combined work can lead to complete analysis. NMR, like
Raman spectroscopy, is especially suited for the task of
identifying the equilibrium solution species of these organo-
metallic compounds under identical conditions to those
exercised during the electrochemical measurements. The
analysis of the APC electrolytic system is particularly important,
as it is one of the most advanced electrolyte solutions for
rechargeable magnesium batteries with operating voltage of up
to 3 V. The spectroscopic methods and the spectral database
obtained in this work can be valuable for the analysis of similar
complexes and organometallic species in other fields, such as
organic synthesis and polymerization catalysis.
MgPh2 syntheses followed well-documented procedures.22
Nuclear Magnetic Resonance. NMR spectra of the solutions
were measured with a Bruker DMX-600 spectrometer. Solutions of
different concentrations and ratios of reagents in THF were measured
in 8 mm NMR tubes without a deuterium lock. 13C chemical shifts are
reported in the figures relative to TMS. 27Al chemical shifts are
reported relative to external references: a solution of AlCl3 in H2O. All
the measurements were performed at room temperature (25 2 °C).
DFT Computations. All aluminum core geometry optimizations
were performed at the B3LYP/6-31+G* level of theory.23−25
Magnesium core geometry optimizations were done at the B3LYP/
CEP-31+G* level in the gas-phase and in solution. The solvent (THF,
ε = 7.58) was taken into account via the IEFPCM model.26
Theoretical NMR spectra were calculated at the B3LYP/6-31+G(d,p)
level in solution, as well as with the mPW1PW91 functional27 in
conjunction with the Aug-cc-pVTZ basis set.28,29 For comparison
between the calculated and the measured 13C NMR, the peak position
of the carbon in TMS (192.981 ppm as obtained from the DFT
H
dx.doi.org/10.1021/om300865a | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
calculation) was used as a reference. All calculations were performed
using Gaussian 09 package.30
(27) Adamo, C.; Barone, V. J. Chem. Phys. 1998, 108, 664−675.
(28) Dunning, T. H., Jr. J. Chem. Phys. 1989, 90, 1007−1023.
(29) Woon, D. E.; Dunning, T. H., Jr. J. Chem. Phys. 1993, 98, 1358−
1371.
ASSOCIATED CONTENT
■
(30) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci,
B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H.
P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.;
Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.;
Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin,
K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.;
Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega,
N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.;
Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.;
Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.;
Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.;
S
* Supporting Information
Supporting Figures S1−S6. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*Tel: 972-35318317. E-mail: Doron.Aurbach@biu.ac.il.
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
̈
Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.;
Partial support for the work was obtained from the I-SAEF
(Israel−US Strategic Alternative Energy Foundation) and from
Pellion, Inc., USA.
Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09,
Revision A.02; Gaussian, Inc.: Wallingford, CT, 2009.
REFERENCES
■
(1) Pour, N.; Gofer, Y.; Major, D. T.; Aurbach, D. J. Am. Chem. Soc.
2011, 133 (16), 6270−6278.
(2) Aurbach, D.; Lu, Z.; Schecter, A.; Gofer, Y.; Gizbar, H.;
Turgeman, R.; Cohen, Y.; Moshkovitz, M.; Levi, E. Nature 2000, 407,
724.
(3) Aurbach, D.; Gizbar, H.; Schechter, A.; Chusid, O.; Gottlieb, H.
E.; Gofer, Y.; Goldberg, I. J. Electrochem. Soc. 2002, 149, A115.
(4) Gofer, Y.; Chusid, O.; Gizbar, H.; Vestfrid, Y.; Gottlieb, H. E.;
Marks, V.; Aurbach, D. Electrochem. Solid-State Lett. 2006, 9, A257.
(5) Gizbar, H. The Study of Non Aqueous Electrochemical Systems of
Magnesium. Ph.D. Thesis, Bar-Ilan University, 2006.
(6) Mizrahi, O. Electrolyte Solutions for Rechargeable Magnesium
Batteries. M.Sc. Thesis, Bar-Ilan University, 2007.
(7) Ashby, E. C. Q. Rev. Chem. Soc. 1967, 21, 259.
(8) Sakamoto, S.; Imamoto, T.; Yamaguchi, K. Org. Lett. 2001, 3,
1793.
(9) Kalinowski, H. O.; Berger, S.; Braun, S. Carbon-13 NMR
Spectroscopy; Wiley: New York, 1988.
(10) Vestfries, Y.; Chusid, O.; Goffer, Y.; Aped, P.; Aurbach, D.
Organometallics 2007, 26, 3130.
(11) Benn, R. J. Organomet. Chem. 1987, 333, 169−180.
(12) Benn, R. J. Organomet. Chem. 1987, 333, 155−168.
(13) Benn, R. Angew. Chem., Int. Ed. Engl. 1986, 25, 861−881.
(14) Benn, R. Angew. Chem., Int. Ed. Engl. 1983, 22, 779−780.
(15) Gizbar, H.; Vestfrid, Y.; Chusid, O.; Gofer, Y.; Gottlieb, H. E.;
Marks, V.; Aurbach, D. Organometallics 2004, 23, 3826−3831.
(16) Cerny, Z.; Machacek, J.; Casensky, B.; Kriz, O.; Tuck, D. G.
Inorg. Chim. Acta 2000, 300−300, 556.
(17) Wiberg, K. B. J. Comput. Chem. 1999, 20, 1299−1303.
(18) Weitman, M.; Lerman, L.; Nudelman, A.; Major, D. T.; Gottlieb,
H. E. Tetrahedron 2010, 66, 1465−1471.
(19) Wang, X.; Wang, C.; Zhao, H. Int. J. Mol. Sci. 2012, 13, 15420−
15446.
(20) Wrackmeyer, B.; Klimkina, E. V. Z. Naturforscher 2008, 63b,
923−928.
(21) Chen, C. R.; Gau, H. M. Acta Crystallogr. 2008, E64, m1381.
(22) Wotiz, J. H.; Hollongsworth, C. A.; Dessy, R. E. J. Am. Chem.
Soc. 1956, 78, 1221.
(23) Becke, A. D. J. Chem. Phys. 1993, 98, 5648−5652.
(24) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785−789.
(25) Hehre, W. J.; Radom, L.; Schleyer, P. v. R.; Pople, J. A. Ab Initio
Molecular Orbital Theory; John Wiley & Sons: New York, 1986.
(26) (a) Cance, M. T.; Mennucci, B.; Tomasi, J. J. Chem. Phys. 1997,
107, 3032. (b) Mennucci, B.; Tomasi, J. J. Chem. Phys. 1997, I, 5151.
(c) Mennucci, E.; Cances, M. T.; Tomasi, J. J. Phys. Chem. B 1997,
101, 10506. (d) Tomasi, J.; Mennucci, B.; Cances, E. J. Mol. Struct.:
THEOCHEM 1999, 464, 211.
I
dx.doi.org/10.1021/om300865a | Organometallics XXXX, XXX, XXX−XXX