Phosphoranimines containing cationic N-imidazolinium moieties

Article abstract of DOI:10.1016/j.ica.2017.05.075

Three new monomeric phosphoranimines (PAs; R′[sbnd]N[dbnd]PR₃) were synthesized by the Staudinger route using 2-azido-1,3-dimethylimidazolinium hexafluorophosphate [ADMImPF] and various phosphines to form their respective PA salts of triphenylphosphine (1), tri(n-octyl)phosphine (2), and tris(2,2,2-trifluoroethoxy)phosphite (3). These PA salts have a cationic imidazolinium moiety attached to nitrogen. Interestingly, 2 is a room temperature ionic liquid (RTIL) with a melting point around ?5?°C and a viscosity of 740cP at 25.5?°C. 1 and 2 show stability to moisture exposure and atmosphere for weeks whereas 3 is not stable toward atmospheric moisture undergoing decomposition to form 2-amino-1,3-dimethylimidazolinium phosphate (4). Thermal analyses reveal that 1 and 2 are stable to 90?°C under nitrogen, but 3 undergoes degradation by exposure to moisture and/or elevated temperature. All compounds exhibit similar multinuclear NMR chemical shifts as compared to their corresponding phosphine oxide counterparts. However, their respective X-ray structure determinations indicate that 2 has an open structure with the octyl groups that avoid the ionic charges, whereas 1 and 3 have tighter structural packing of the hexafluorophosphate to their pendant groups.

Full text of DOI:10.1016/j.ica.2017.05.075

Inorganica Chimica Acta 466 (2017) 254–265
Contents lists available at ScienceDirect
Inorganica Chimica Acta
journal homepage: www.elsevier.com/locate/ica
Research paper
Phosphoranimines containing cationic N-imidazolinium moieties
John R. Klaehn a, , Harry W. Rollins a, Joshua S. McNally a, Navamoney Arulsamy b, Eric J. Dufek c
⇑
a Biological and Chemical Processing, Idaho National Laboratory, P.O. Box 1625, Idaho Falls, ID 83415-3732, USA
b Department of Chemistry, University of Wyoming, 1000 E. University Ave, Dept 3838, Laramie, WY 82071, USA
c Energy Storage and Advanced Vehicles, Idaho National Laboratory, P.O. Box 1625, Idaho Falls, ID 83415, USA
a r t i c l e i n f o
a b s t r a c t
Three new monomeric phosphoranimines (PAs; R0AN@PR3) were synthesized by the Staudinger route
using 2-azido-1,3-dimethylimidazolinium hexafluorophosphate [ADMImPF] and various phosphines to
form their respective PA salts of triphenylphosphine (1), tri(n-octyl)phosphine (2), and tris(2,2,2-trifluo-
roethoxy)phosphite (3). These PA salts have a cationic imidazolinium moiety attached to nitrogen.
Interestingly, 2 is a room temperature ionic liquid (RTIL) with a melting point around ꢀ5 °C and a viscos-
ity of 740cP at 25.5 °C. 1 and 2 show stability to moisture exposure and atmosphere for weeks whereas 3
is not stable toward atmospheric moisture undergoing decomposition to form 2-amino-1,3-dimethylim-
idazolinium phosphate (4). Thermal analyses reveal that 1 and 2 are stable to 90 °C under nitrogen, but 3
undergoes degradation by exposure to moisture and/or elevated temperature. All compounds exhibit
similar multinuclear NMR chemical shifts as compared to their corresponding phosphine oxide counter-
parts. However, their respective X-ray structure determinations indicate that 2 has an open structure
with the octyl groups that avoid the ionic charges, whereas 1 and 3 have tighter structural packing of
the hexafluorophosphate to their pendant groups.
Article history:
Received 20 April 2017
Accepted 31 May 2017
Available online 1 June 2017
Keywords:
Synthesis
Characterization
Phosphoranimine
X-ray crystal structure
Room temperature ionic liquid
Ó 2017 Published by Elsevier B.V.
1. Introduction
terized; however, the resulting ionic PAs are not stable towards
atmospheric moisture and will tend to degrade. Other PA com-
Phosphoranimines (PAs; Fig. 1) are small molecule phosphorus-
nitrogen compounds that are common building blocks to make
polyphosphazenes or cyclic phosphazenes [1–4]. These materials
have unique properties that range in a variety of membrane, bio-
logical, and electronic applications [5–8], and recently PAs were
shown that they could be used as an additive battery electrolyte
in lithium-ion batteries [9]. Traditional PA chemistry has been used
to generate polymeric phosphazenes, where the PAs contain a
trimethylsilyl-group at nitrogen (Fig. 1) and different phosphorus
pendant groups, such as directly bound alkyl/aryl groups and a
good leaving group (chloro-, trifluoroethoxy- or phenoxy-)
[3,7,10–12]. Generally, heating these PAs to 180–200 °C results in
condensation polymerization, and recent literature has shown that
some of their oligomeric intermediates can be isolated as ionic
moieties even at ambient temperatures in the presence of Lewis
acid catalysts (e.g., PCl5) [13–16]. This polymerization has been
studied to determine the reactive intermediate compound (ionic
PA) that sustains the living polymerization at ambient tempera-
tures [15,16]. Several different ionic PA intermediates were charac-
pounds with a reactive imide ligand (hydrogen, halogen or
organosilane) can be transformed into a reactive intermediate that
builds larger PA structured compounds and metal complexes, such
as the aza-Wittig reagents [17–23]. Interestingly, recent research
has shown similar ionic and non-ionic PA compounds can be used
for polymerization and depolymerization catalysts/co-catalysts of
various organic condensation polymers, e.g. depolymerization of
polyesters using sterically hindered, strong PA Lewis bases
[24–26,21].
Most PAs without a P@N-P backbone (only P@N) structure are
made by two common synthetic strategies: Neilson method or
Staudinger reaction [3,11,27–30]. The Neilson method is the pre-
ferred synthetic route in making PA precursors, as it avoids the
use of hazardous organic azides, as well as the creation of
unwanted by-products through P@N cyclization, and limits phos-
phite or PA rearrangement products, such as alkyl phosphorami-
dates ((RO)2P(@O)NH2) [11,30,31]. The disadvantage of the
Neilson method is that it requires several reaction steps followed
by vacuum distillation to achieve a pure PA product. Also, because
the method produces products that tend to polymerize to yield
polyphosphazenes, the Neilson method is limited to producing
moisture sensitive PAs with organosilanes at nitrogen. As the
⇑
Corresponding author.
E-mail address: john.klaehn@inl.gov (J.R. Klaehn).
http://dx.doi.org/10.1016/j.ica.2017.05.075
0020-1693/Ó 2017 Published by Elsevier B.V.

J.R. Klaehn et al. / Inorganica Chimica Acta 466 (2017) 254–265
255
R
P
4-dimethylaminopyridine [39–42]; however, their PA products
are not commonly stable under atmospheric conditions for
extended periods. Also, there are a limited number of usable organic
azide compounds, and the syntheses of PA compounds with a catio-
nic pendant group at nitrogen have not been explored by the Stau-
dinger method. In the literature, atmospherically stable PA
compounds are formed as either (large) sterically encumbered
alkyl/aromatic PA compounds [11,43,44] or as various forms of ionic
aromatic structures of Ph3P@N+@PPh3 Xꢀ ([PPN]Cl) with halogen
anions [20,42,45], giving the [PPN]Cl structure its stability and its
charge distribution across the compound. Hence, the reported com-
pounds 1–3 have the ability to distribute the charge across the PA
structure through electron resonance structures (Fig. 3), which
can enhance their stability for extended periods.
R'''
N
R''
R'
R, R' R'' = alkyl, aryl, alkoxy, aryloxy
R''' = alkyl, aryl, organosilane
Fig. 1. N-organophosphoranimine (PA) structure.
oldest synthetic route to PAs, the Staudinger route comprises of a
single step reaction between an azide and phosphine where a vari-
ety of N and P pendant groups can be utilized to make unique PA
compounds.
Over the past decade, new interest in the Staudinger reaction
has re-emerged through ‘‘click” chemistries, such as azide fluores-
cent markers on biological materials or attachments to polymer
pendant groups [32–35]. Several new organic azides from com-
mercial sources not only can be utilized for the Staudinger 3+2
‘‘click” reactions, but they can also be used with phosphines to
make new PAs. In some cases, these organic azides are stable as
purified solids, e.g. 2-Azido-1,3-dimethylimidazolinium hexafluo-
rophosphate [ADMImPF] [36–38]. ADMImPF can be used directly
in the Staudinger reaction without needing to employ rigorous
safety precautions during the synthesis. The added azide stability
is essential to the Staudinger method and introduces new PA
compounds which are not easily accessible by the Neilson method.
We report three cationic N-imidazolinium PA compounds that
were synthesized from the reactions of triphenylphosphine (TPP),
trioctylphosphine (TOP) or tris(2,2,2-trifluoroethoxy)phosphite
(TTFP) with ADMImPF. All isolated PA products were characterized
by multinuclear NMR and single crystal X-ray diffraction, and the
thermal properties and reactivity towards the atmosphere and
moisture were also characterized.
ADMImPF is a commercial product that is available as a white
powder. It is stable to brief exposure to the atmosphere at room
temperature. Thus, it is relatively simple to handle in the labora-
tory and a good candidate for the Staudinger reaction. Even though,
ADMImPF is not completely soluble in toluene (as reported in the
literature), the reported phosphines were able to react with the
azide in toluene without issues. In addition, toluene allowed for
easier workup and has a useful temperature range for these
reactions.
All three new PAs form a yellow colored intermediate during
reaction. 1 and 2 show distinct color changes that dissipated over
time as the azide reacted with the phosphine. It is interesting that
all three reactions do not proceed at room temperature do not
immediately react as the phosphine was added, but further heating
was required to yield PA products, such as 2 and 3. All PA products
were not completely soluble in toluene and tend to precipitate or
oil out of solution upon cooling to room temperature. Also, these
compounds do not show water solubility. The toluene solution
was decanted or pipetted away from the products, where they
were washed and isolated from hexane solutions. The PAs that
are solids could be purified further by common recrystallization
methods. In the instance of synthesizing compound 2, isolation
was carried out using hexanes, and the compound remained as
an oily liquid at room temperature after solvent removal. This pro-
duct was placed in a freezer (ꢀ30 °C), where prismatic crystals
were generated. All of the reported PA products were isolated
and analyzed by multi-nuclear NMR, thermal analysis, and sin-
gle-crystal X-ray diffraction.
2. Results and discussion
A series of monomeric PA salt products (Scheme 1 and Fig. 2)
with an organophosphine (TPP, TOP or TTFP) and ADMImPF were
synthesized by the Staudinger method: P,P,P-triphenylphosphoran-
iminyl-N-1,3-dimethylimidazolinium hexafluorophosphate [1], P,P,
P-trioctylphosphoraniminyl-N-1,3-dimethylimidazolinium
hex-
In Table 1, the multi-nuclear P-31 NMR spectra of 1–3 show
that CF3CH2OA groups contribute more electron density to phos-
phorus than the phenyl or octyl groups. When comparing com-
pounds 1–3 to their phosphine oxide precursors [46], all
compounds follow similar trends in the chemical shifts; however,
PAs 1–3 have slightly more electron density at phosphorus result-
ing in a 12–17 ppm upfield shift. No differences in chemical shifts
are observed among the P-31 NMR spectra of the PFꢀ6 counter ion
to compounds 1–3, which would be expected due to the over-
afluorophosphate [2], and P,P,P-tris(2,2,2-trifluoroethoxy)phospho-
raniminyl-N-1,3-dimethylimidazolinium hexafluorophosphate [3].
These PA compounds can be a solid or liquid at room temperature,
depending on the pendant groups at phosphorus. PA 2 was the most
interesting due to it being an atmospherically stable room temper-
ature ionic liquid (RTIL) with a viscosity of 740cP (25.5 °C). Previous
PA compounds by the Neilson and Staudinger synthetic methods
focused on attaching or creating ionic moieties at the phosphorus
position, such as imidazolium halides, organophosphines, and
CH3
CH3
PF6
PF6
R
R
N
N
toluene
P
N
P
+
N3
R
R
R
ambient to reflux
-N2
R
N
N
CH3
CH3
R = Phenyl, Octyl, OCH2CF3
R = Phenyl (1), Octyl (2),
(3)
OCH2CF3
Scheme 1. Staudinger reaction of 2-azido-1,3-dimethylimidazolinium hexafluorophosphate (ADMImPF) and phosphines to form PA compounds 1–3.

256
J.R. Klaehn et al. / Inorganica Chimica Acta 466 (2017) 254–265
PF
6
PF
6
PF
6
CF
3
N
N
N
O
N
P
N
P
N
P
O
N
N
CF
N
3
O
CF
3
P,P,P-tris(2,2,2-trifluoroethoxy)-
phosphoraniminyl-N-1,3-
dimethylimidazolinium
P,P,P-tri-n-octylphosphoraniminyl-N-
1,3-dimethylimidazolinium
P,P,P-triphenylphosphoraniminyl-N-
1,3-dimethylimidazolinium
hexafluorophosphate (1)
hexafluorophosphate (2)
hexafluorophosphate (3)
Fig. 2. PA structures of compounds 1–3.
PFa
6
PF6
R1
R1
N
N
N
P
R2
N
P
R2
N
N
R3
R3
Fig. 3. Potential ionic resonance structures of the imidazolinium PA compounds.
return of the scan. There is a slight difference in the melting point
between the two DSC runs, but the crystallization showed the
same temperature. PA 2 is the RTIL that showed a small broad
melting point at ꢀ5 °C that remained after the first heating cycle.
The interesting point is that the compound does not show a
crystallization point on the cooling return of the DSC, but it
does solidify during the 30-min hold at ꢀ80 °C, which might
flash-freeze the compound (amorphous solid structure) to give a
large ꢀ65 °C melting point. The inflection points prior to the melt-
ing point could be due to some variation in the crystalline and
amorphous solid structures of 2. Data for PA 3 indicates that the
compound decomposed upon exposure of the atmosphere. When
using the melting point apparatus, crystals of 3 show a melting
point at 97–98 °C; however, the DSC indicates two different com-
pounds were present that had melting points at ꢀ45 °C and at
95 °C. The two melting points indicate that 3 is not stable under
atmospheric conditions or in thermocycling. Scheme 2 illustrates
a possible degradation pathway that yields the imidazolinium
hexafluorophosphate and the organophosphate(s). The resulting
imidazolinium salt, 2-amino-1,3-dimethylimidazolinium hexafluo-
rophosphate (4), was later identified by X-ray crystallography (see
crystallographic data). The CF3CH2OA groups are known as good
leaving groups for the PA condensation polymerization; therefore,
it is reasonable to assume that decomposition of PA 3 initiates with
loss of trifluoroethoxide. In addition, the X-ray structure of PA 3
could not be well resolved as it crystallizes out as clusters of thin
plates, which did not yield the highest quality crystals. Conse-
quently, the structural parameters are less accurately determined.
The data, however, can be considered as reasonable approximation
of the parameters. Crystallographic data collection and refinement
parameters are given in Table 2, and selected structural data are
collected in Table 3.
Table 1
P-31{1H} NMR Chemical Shifts (d in ppm) of Ionic PAs 1–3 and their respective P-31
{1H} NMR Chemical Shifts of phosphine oxides (relative to 85% H3PO4).
R@
OCH2CF3 [3]
Ph [1]
n-Octyl [2]
Ionic PA1
ꢀ14.8 (s)
ꢀ2.8 (s)
13.0 (s)
27.0 (s)
31.7 (s)
48.5 (s)
R3P@O2
1
2
This work.
Ref. [46].
whelming influence of local fluorines. Since 3 was not found to
have long term stability under atmospheric conditions, exposure
testing was only performed on 1 and 2. These exposure tests were
conducted in moisture conditions (ꢁ10 wt% water in d8-THF;
Figs. 4–7), and they were analyzed by P-31 and F-19 NMR to see
if any degradation products of the PA or PF6 were produced over
a period of several weeks. Over this time period, no detectable
aqueous decomposition products, which are phosphine oxides
from 1 and 2 and hydrofluoric acid formation from PFꢀ6 , are
observed in 1 for either the P-31 or F-19 NMR spectra after 14 days.
But, 2 showed a small amount of phosphine oxide that increased
from 0.6% (initial) to 2.3% (29 days) by integration in P-31 NMR
(upper left in Fig. 6), and no observable changes in the F-19 after
30 days. These NMR spectra show that water exposure is not an
immediate issue with 1 and 2 under these conditions; however,
further studies will have to be performed to understand if these
compounds can be used in more aggressive conditions (tempera-
ture, acidic pH, etc.). Overall, the combined results of exposure to
water and atmospheric conditions indicate that some of these ionic
PA products can be exposed for long periods without decomposing
at ambient conditions. This is important for RTILs, such as 2, which
could be used in aqueous/organic solvent conditions.
In contrast, the crystallographic structures of 1 and 2 are well
refined. These structures together with that of PA 3 provide evi-
dence that ion pairing is observed with the imidazolinium nitro-
gens or phosphorus (see Fig. 3), but this might be due to crystal
packing. From Fig. 3, there is a possibility of a resonance structure
in these PA compounds where the positive charge can be
distributed from the imidazolinium nitrogens to phosphorus.
Determination of melting points and/or degradation tempera-
tures for these ionic PA compounds was performed using differen-
tial scanning calorimetry (DSC; see Figs. 8–10). All thermal scans
were cooled down to ꢀ80 °C and held for 30 min to ensure that
the compounds were solids and at thermal equilibrium prior to
determining their melting points. PA 1 showed a melting point at
about 185 °C and crystallization at about 145 °C upon the cooling

J.R. Klaehn et al. / Inorganica Chimica Acta 466 (2017) 254–265
257
Fig. 4. P-31{1H} NMR Spectra of 1 and PF6 (upper left) for 14 days.
Fig. 5. F-19{1H} NMR Spectra of 1 and HF (upper right) for 14 days.
The P-31 NMR spectra (Fig. 6) show an electron poor phosphorus
center containing the octyl groups (2), and the phosphorus center
becomes more electron rich when phenyl (1) and alkoxy groups (3)
are bound to phosphorus. However, when comparing the PA elec-
tron density between their P-31 NMR spectra and the location of
the PFꢀ6 anion to PA to in the X-ray structures (Figs. 11–16), there
is no clear spectral trend with the location of the PFꢀ6 anion and
the PA phosphorus center. Even from the X-ray structural data,
the ionic bond distances between the nitrogen atoms of the imida-
zolinium cations and PFꢀ6 anions are at 3.456(1), 3.163(4) and 3.273
(9) Å, respectively in 1–3. The PFꢀ6 anions are within 0.3 Å in prox-
imity, indicating all have similar ionic interactions. As shown in

258
J.R. Klaehn et al. / Inorganica Chimica Acta 466 (2017) 254–265
Fig. 6. P-31{1H} NMR Spectra of 2, (Oct)3P@O (upper left) and HF (lower right) for 29 days.
Fig. 7. F-19{1H} NMR Spectra of 2 and HF (upper right) for 29 days.
Figs. 8, 10 and 12, the tight packing in 2 and 3 are also reflected in
their closest interionic contacts. The other remarkable feature in 1
is the close proximity of the PFꢀ6 anion to one of the phenyl groups
and imidazolinium moiety with the associated Cꢂ ꢂ ꢂP (C20ꢂ ꢂ ꢂP2)
interatomic distances of 3.121(2) Å. The most interesting feature
of crystal packing in 2 is that the three octyl groups are staggered
and spread out laterally away from the ionic moieties. This open
crystal packing among the octyl groups (between 8 and 13 Å) prob-
ably allows 2 to have a low melting point as observed in DSC mea-
surements. Similar to the packing in 2, compound 3 also exhibits
tight packing of ions. The three C-N bond distances in all three
PA’s are comparable and fall between typical CAN single and dou-

J.R. Klaehn et al. / Inorganica Chimica Acta 466 (2017) 254–265
259
Fig. 8. DSC Trace of PA 1, two thermal cycles from ꢀ80 to 250 °C.
Fig. 9. DSC Trace of PA 2, two thermal cycles from ꢀ80 to 95 °C.
Fig. 10. DSC Trace of PA 3, two thermal cycles from ꢀ80 to 95 °C.

260
J.R. Klaehn et al. / Inorganica Chimica Acta 466 (2017) 254–265
PF
6
CF
3
CF
O
3
PF
6
2
N
O
N
O
P
R
H O
2
O
N
P
+
NH
O
O
CF
N
3
N
O
CF
3
CF
3
3
4
R = H, CF CH
3 2
Determined by X-ray
Crystallography
Scheme 2. Decomposition products of compound 3 after water exposure.
Table 2
PA 1–3 crystallographic data.
Compound
1
2
3
4
Chemical formula
T, K
C
150
23H25F6N3P2
C29H60F6N3P2
150
C11H16F15N3O3P2
150
C5H12F6N3P
150
k, Å
space group
0.71073
P21/c
0.71073
P21/c
0.71073
P21/c
0.71073
P21/n
a, Å
b, Å
c, Å
10.2614(2)
17.1198(3)
14.2207(3)
108.6855(10)
2366.52(8)
4
1.458
0.248
0.416
0.579
14.01775)
13.4151(5)
19.2683(7)
103.594(2)
3521.9(2)
4
1.182
0.177
0.0463
0.1165
9.3742(15)
24.605(4)
10.7300(16)
115.470(9)
2234(6)
4
1.740
0.338
0.1600
0.2114
8.1882(3)
11.3949(4)
11.2514(4)
102.920(2)
1023.22(6)
4
1.682
0.329
0.0467
0.1394
b, °
V, Å3
Z
Dcalc, Mg mꢀ3
l
, mmꢀ1
R1[I > 2
wR2[I > 2
r
(I)]a
(I)]b
r
P
P
P
P
a
b
2
R1 = ||Fo ꢀ |Fc||/ |Fo|; wR2 = { [w(Fo ꢀ Fc2)2]/ w(Fo2)2]}1/2
.
Table 3
Selected bond distances (Å), bond angles (°) and torsion angles (°) for 1–4.
Compound
P1AN1
N1AC1
C1AN2
C1AN3
PAC or PAO
PAC or PAO
PAC or PAO
N1AC1AN2
N1AC1AN3
N2AC1AN3
P1AN1AC1
P1AN1AC1AN2
P1AN1AC1AN3
1
2
3
4
–
1.5789(9)
1.3100(13)
1.3434(14)
1.3569(13)
1.7977(11)
1.8001(11)
1.8024(11)
130.33(10)
120.11(10)
109.56(9)
141.53(8)
3.6(2)
1.5890(13)
1.302(2)
1.342(2)
1.501(6)
1.323(8)
1.316(9)
1.316(9)
1.551(6)
1.549(5)
1.599(6)
121.1(8)
126.9(7)
112.0(6)
144.5(6)
ꢀ118.8(11)
64.2(14)
1.325(2)
1.327(2)
1.324(2)
–
–
–
123.99(12)
124.42(13)
111.58(12)
–
–
–
1.357(2)
1.7998(16)
1.7981(16)
1.8037(16)
122.63(14)
128.13(14)
109.12(12)
140.42(12)
ꢀ135.2(2)
49.1(3)
ꢀ175.7(1)
ble bond distances (Table 2). The double bond across all three PA
imidazoliniums is also nearly planar indicating possible delocaliza-
tion of a double bond across the N2AC1AN3 units and potentially
the lone pair on N1. The 2-amino-1,3-dimethylimidazolinium
cation present in 4 also exhibits similar double bond delocalization
over the CN3 unit (Table 2) and close proximity with the anion
(Figs. 17 and 18). In addition, the CAN bond distances and NACAN
bond angles in the cation are similar to the corresponding dis-
tances in PA’s 1–3 indicating that the P@N double bond present
in the latter compounds do not significantly impact the nature of
bonding within the dimethylimidazolinium moiety.
From Table 4, the bond distances of a phosphorus-nitrogen
bonds in an assortment of compounds can range from a PAN single
bond at 1.78 Å to a P@N double bond at 1.54 Å [7,39–42,47]. Also in
Table 4, various PA bond distances and bond angles measured in
other studies are given along with the PA compounds 1–3 which
are reported here; however, compound 4 is a decomposition pro-
duct of 3. Both ionic and neutral PAs included in the table display
a variety of phosphonium (PAN) and phosphoranimine (P@N)
Fig. 11. View of the PA cation, 1 and the PFꢀ6 anion. Hydrogen atoms are omitted.
The broken lines represent the closest interionic contacts. The Cꢂ ꢂ ꢂF and Nꢂ ꢂ ꢂF
distances are 3.121(1) and 3.456(1) Å, respectively.

J.R. Klaehn et al. / Inorganica Chimica Acta 466 (2017) 254–265
261
Fig. 12. Packing in the crystals of PA 1 as viewed along the b-axis. Hydrogen atoms are omitted and all atoms are unlabeled for clarity.
angles from 144° to 166° and Me3SiAN@PPh3 at 138°. This sug-
gests that the imidazolinium group is interacting RAN@P bond to
give 1–3 larger bond angles. However, Irmie et al.’s PA compounds
show smaller bond angles at around 125°, which is caused by the
substituted nitrogen and forcing the RAN@P to be bent. Again,
these bond angles and bond distances could be influenced by crys-
tal packing, but there are some trends that are different from other
PA compounds.
3. Conclusions
Three new PA salts were formed by the Staudinger reaction
using an ionic azide, ADMImPF. The reaction is straightforward in
making the respective ionic PA products; however, these PA prod-
ucts tend not to be soluble in toluene, hexanes or water. It is inter-
esting that the isolated PA 1–2 salts are not extremely reactive to
air or moisture; however, the isolated PA 3 salt decomposes upon
exposure to the atmosphere. The remarkable feature of these com-
pounds is that 2 is an ionic liquid at room temperature. From P-31
and F-19 NMR, the spectra show that PAs 1–2 are stable towards
water laden organic solvents for extended periods without notice-
able degradation products. In addition, thermal cycling of 1–3 by
DSC shows that 1 and 2 are stable to 90 °C, but 3 decomposes into
its corresponding phosphine oxide and DMImPF6 after one thermal
cycle. The X-ray structure data of 1–3 reveal that the PFꢀ6 anion
resides nearby the imidazolinium functional group. Also from the
X-ray, crystals of 2 have an open packing structure which can
explain its RTIL behavior (m.p. ꢀ5 °C). When comparing the X-
ray structures of other known ionic PA compounds, it can be seen
that the imidazolinum interacts with the P@N bond resulting in
larger than 140° bond angle at R0AN@P. Overall, 1–3 are easily syn-
thesized to form N-substituted ionic PA compounds without noted
complications commonly found using the Staudinger reaction.
Fig. 13. View of the cation and anions in PA 2. Hydrogen atoms and the second set
of sites for the four disordered F atoms of the anion are not shown. The broken line
represents the shortest interionic distance. The Nꢂ ꢂ ꢂF (N2ꢂ ꢂ ꢂF1) distance is 3.163
(4) Å.
bond distances and angles. The interesting point is that the average
P@N distance for phosphonium PA compounds [42] (Irmie) is
about 1.63 Å and 1–2 are about 1.585 Å; however, the P@N dis-
tance for phosphonium PA compounds [39–41] (Huynh) and 3
have the shortest bond distances of ꢁ1.50 Å. In all cases, the phos-
phonium PA salt P@N bond distances suggest a bond character
between a single and double bond. The closest representative
single bond phosphonium bond is with Irmie et al.’s compounds
[42]. The bond distance differences between the ionic and neutral
forms of PA compounds are not substantial; however they reside
closer to a double bond than a single bond in most cases. Therefore,
it suggests that the positive charge remains delocalized rather than
localized on a single atom.
4. Experimental section
4.1. Materials and general procedures
The following reagents were obtained from commercial sources
and used without further purification: 2-Azido-1,3-dimethylimida-
zolinium hexafluorophosphate (ADMImPF; Aldrich), triph-
enylphosphine (TPP; Aldrich), tris(2,2,2-trifluoroethoxy)phosphite
(TTFP; Strem) and tri-n-octylphosphine (TOP; Strem). House deion-
ized (DI) water was purified by a Millipore 302 system that
achieved an 18 MOhm rating and used directly. Anhydrous sol-
vents, such as toluene, tetrahydrofuran (THF) and hexane were
The RAN@P bond angles on these compounds show interesting
trends, and the Huynh et. al.’s PAs and PAs 1–3 have a large bond
angle that is greater than 140° (more linear). The Me3Si-group con-
tributes back
p-orbital overlap to give a linear RAN@P bond angle,
which is shown with Huynh et. al.’s PA compounds with large bond

262
J.R. Klaehn et al. / Inorganica Chimica Acta 466 (2017) 254–265
Fig. 14. Packing in the crystals of PA 2, a as viewed along the b-axis. Hydrogen atoms and the second set of sites for the four disordered F atoms of the anion are not shown
and all atoms are unlabeled.
obtained from Aldrich and used as received. Proton, 13C{1H}, 19F
{1H} and 31P{1H} Nuclear Magnetic Resonance (NMR) spectra were
recorded on a Bruker Ascend Avance III 600 MHz spectrometer.
MNOVA NMR software package (v10.0.2) was used for multiple
NMR spectra arrangements [51].
4.2. 19F{1H} and 31P{1H} NMR moisture exposure experiments
NMR-based moisture exposure experiments were conducted on
PA compounds, 1 and 2, over a 29 day time frame. The solutions
were made with 0.2–0.25 g of PA in 1 mL of d8-THF (Cambridge
Isotopes) containing 10 wt% water (by solvent) in each NMR sam-
ple tube. A polypropylene cap was used on the NMR tubes for the
duration of these experiments. 19F{1H} and 31P{1H} NMR spectra
were periodically recorded to observed spectral changes in the
hexafluorophosphate and the PA compounds.
4.3. Differential Scanning Calorimetry (DSC) data
Thermal analysis was obtained using a TA Instruments (New-
Castle, DE) model Q200 Differential Scanning Calorimeter (DSC).
Fig. 15. View of the cation and anion in PA 3. Hydrogen atoms and the second set of
For data collection, Tzero aluminum hermetic pans were loaded
sites for the four disordered F atoms of the anion are not shown. The broken line
with ꢁ10 mg of sample and the temperature was cycled between
represents the shortest interionic contact. The NꢂꢂꢂF (N2ꢂꢂꢂF16) distance is 3.273(9)
Å.
ꢀ80 and 250 °C at a rate of 5 °C, under a purge gas (N2) with a flow

J.R. Klaehn et al. / Inorganica Chimica Acta 466 (2017) 254–265
263
Fig. 16. Packing in the crystals of PA 3, a as viewed along the b-axis. Hydrogen atoms and the second set of sites for the four disordered F atoms of the anion are not shown. All
atoms are unlabeled for clarity.
Fig. 17. View of the cation and anion in 4. Hydrogen atoms and the second set of
Fig. 18. Packing in the crystals of 4, as viewed along the a-axis. Hydrogen atoms
sites for the four disordered F atoms of the anion are not shown. The broken line
and the second set of sites for the four disordered F atoms of the anion are not
represents the shortest interionic contact. The NꢂꢂꢂF (N1ꢂꢂꢂF3) distance is 2.841(6) Å.
shown. All atoms are unlabeled for clarity.
rate of 50.0 mL/min. Before cycling, the temperature was held at
ꢀ80 °C to ensure complete equilbration.
a distance of 5.13 cm from the crystal during the data collection. A
series of narrow frames of data were collected with a scan width of
0.5° in
x or / and an exposure time of 20 s per frame. The frames
4.4. Crystallographic data
were integrated with the Bruker SAINT Software package1 using a
narrow-frame integration algorithm. The data were corrected
for absorption effects by the multi-scan method (SADABS).
Crystallographic data collection parameters and refinement data
are collected below in Table 1. Structures were solved by the direct
methods using the Bruker SHELXTL (V. 2014.11-0) Software
Package [52].
The X-ray diffraction data for 1–4 were measured at 150 K on a
Bruker SMART APEX II CCD area detector system equipped with a
graphite monochromator and a Mo Ka fine-focus sealed tube oper-
ated at 1.5 kW power (50 kV, 30 mA). Colorless crystals were glued
to a glass fiber using Paratone N oil and the detector was placed at

264
J.R. Klaehn et al. / Inorganica Chimica Acta 466 (2017) 254–265
Table 4
Bond distances and bond angles from X-ray crystal determination data on various phosphoranimines along with PAs 1–3.
Phosphoranimine
Bond distances (Å)
C-P
Bond angle (°)
P@NAR0
P@N
NAR0
[1]a
[2]a
[3]a
1.800(1)
1.800(2)
1.566(6)
1.808
1.818
1.795
1.784
1.800
1.800
–
1.579(1)
1.589(1)
1.501(6)
1.603
1.547
1.598
1.628
1.621
1.646
1.489
1.337(1))
1.334(2)
1.318(9)
1.329
1.692
–
–
–
–
1.690
1.722
1.694
1671
1.696
141.53(8)
140.42(12)
144.5(6)
130.4
138.4
133.1
124.2
129.1
121.3
166.0
b
PhAN@PPh3
c
Me3SiAN@PPh3
[Ph3P@N@PPh3]+Clꢀd
[iPrAN(H)@PPh3]+ Brꢀe
[tBuAN(H)@PPh3]+ Brꢀ
e
[tBuAN(Me)@PPh3]+ Iꢀe
[Me3SiN@P(OCH2CF3)2 DMAP]+ BF4ꢀf
[Me3SiN@P(OCH2CF3)2 PEt3]+ Brꢀg
[Me3SiN@P(OCH2CF3)2 PMe3]+ Brꢀg
[Me3SiN@PMe2 PEt3]+ CF3SO3ꢀg
[Me3SiN@PMe2 PMe3]+ Brꢀg
–
–
1.492
1.479
1.508
1.533
158.1
160.9
154.0
144.0
1.793
1.792
a
b
c
d
e
f
This work.
Ref. [48].
Ref. [49].
Ref. [50].
Ref. [42].
Ref. [40].
Ref. [41].
g
All non-hydrogen atoms were located in successive Fourier
(s) 2.78; 13C NMR d (d6-Acetone) = (s) 133.7, (d, JPC = 10.6 Hz)
132.4, (d, JPC = 12.1 Hz) 129.7, (d, JPC = 107.2 Hz) 128.3, (s) 47.0,
(s) 33.2].
maps and refined anisotropically. The PFꢀ6 anion in 1 is well ordered,
whereas those in 2–4 are disordered. The disorder in 2 and 4 is
modeled by assigning two sites for all of the F atoms and refining
their site occupancies with a free variable. The two sets of F atoms
in 2 and 4 settle on occupancies of 64 and 36%, and 57 and 43%,
respectively, leading to satisfactory overall refinement of the struc-
tures. The disorder in 3 is modeled by assigning two sets of sites for
four of the F atoms. A free variable refinement of their site occupan-
cies settle on 61 and 39%. The overall refinement of this structure is
poor owing to the poor diffraction of the crystals. However, the
structural parameters deduced are of sufficient quality. All hydro-
gen atoms were located and refined isotropically in the structures
of 1, 2 and 4. The hydrogen atoms were placed in calculated posi-
tions for 3 and refined isotropically with fixed thermal parameters
adapting a riding model.
4.6. Synthesis of P,P,P-tri-n-octylphosphoraniminyl-N-1,3-
dimethylimidazolinium hexafluorophosphate, (2)
A 50 mL two-necked, round bottom flask, water condenser
equipped with a gas inlet, magnetic stir bar, and a rubber septum
was purged with nitrogen. 10 mL of anhydrous toluene was intro-
duced into the flask by syringe, and then ADMImPF (2.0 g,
0.007 mol) was added directly into the flask. The azide has limited
solubility in toluene which makes a slurry. TOP (3.12 mL, 2.6 g,
0.007 mol) was added dropwise by syringe into the flask. As the
phosphine addition progressed, an intense yellow color was
observed during the reaction resulting is a yellow slurry. 20–
30 min after the phosphine was added, the yellow color subsided
and the reaction was heated to near reflux. No apparent nitrogen
off-gassing was observed as the reaction was heated to reflux.
The reaction then was refluxed overnight under nitrogen. Next
day, the reaction was allowed to cool to room temperature. Stirring
was stopped and no phase separation was observed. The reaction
solution was filtered, collected and toluene was removed by a
flowing nitrogen gas stream. The product was washed thrice with
30–40 mL of hexanes, and the remaining hexanes were removed
from bottom phase by a flowing nitrogen stream. The concentrate
was dissolved in toluene where the remaining solids were allowed
to settle and the remaining solution was removed and concen-
trated by a flowing nitrogen stream. The resulting product is a
clear, colorless liquid. The product was stable in air and remains
a liquid at room temperature. The purified product slowly crystal-
lizes in a freezer at ꢀ30 °C to yield colorless, clear prismatic crys-
tals. [73% yield, m.p. = ꢀ5 °C, viscosity = 740 cP at 25.5 °C, 31P
NMR d (d6-Acetone) = (s) 32.5, (septet, JPF = 707.0 Hz) ꢀ144.3; 19F
4.5. Synthesis of P,P,P-triphenylphosphoraniminyl-N-1,3-
dimethylimidazolinium hexafluorophosphate, (1)
A 50 mL two-necked, round bottom flask was equipped with a
water condenser, gas inlet, magnetic stir bar, and a rubber septum
was purged with nitrogen. 10 mL of anhydrous toluene was intro-
duced into the flask by syringe, and then ADMImPF (2.0 g,
0.007 mol) was added directly into the flask. The azide has limited
solubility in toluene which makes a slurry. TPP (1.9 g, 0.007 mol) is
dissolve in 10 mL of anhydrous toluene and the solution was added
dropwise by syringe into the flask. As the phosphine addition pro-
gressed, an intense yellow color was observed during the reaction.
Off gassing of nitrogen is observed as the reaction was heated to
reflux. The reaction was refluxed with monitoring for one hour
and then allowed to reflux overnight under nitrogen. The yellow
color subsided as the reaction was refluxed overnight. The reaction
then was allowed to cool to room temperature, where the desired
compound precipitated as a white solid. Excess toluene was
removed and the product was washed twice with 30–40 mL of
hexanes. The white solid was collected and re-crystallized in
tetrahydrofuran. The product showed no degradation in ambient
atmosphere and upon exposure to water at room temperature.
[87% yield, m.p. = 185 °C, 31P NMR d (d6-Acetone) = (s) 13.0, (septet,
NMR
d
(d6-Acetone) = (d, JPF = 706.0 Hz) ꢀ72.6; 1H NMR
d
(d6-Acetone) = (s) 3.66, (s) 2.98, (doublet of triplets, JPH = 6.0 Hz,
J
HH = 8.4 Hz) 2.24, (doublet of pentets, JPH = 7.8 Hz, JHH = 7.8 Hz)
1.70, (pentet, JHH = 7.8 Hz) 1.70, (m) 1.28–1.42, (m) 3.45–3.51, (t,
HH = 7.8 Hz) 0.90; 13C NMR d (d6-Acetone) = (s) 46.7, (s) 32.9, (s)
J
J
PF = 707.0 Hz) ꢀ144.2; 19F NMR d (d6-Acetone) = (d, JPF = 706.0 Hz)
31.6, (d, JPC = 15.1 Hz) 30.5, (m) 28.8, (d, JPC = 63.4 Hz) 27.0, (s)
22.4, (s) 21.6, (s) 13.4].
ꢀ72.6; 1H NMR d (d6-Acetone) = (m) 7.81–7.90, (m) 7.73, (s) 3.70,

J.R. Klaehn et al. / Inorganica Chimica Acta 466 (2017) 254–265
265
[9] E.J. Dufek, J.R. Klaehn, J.S. McNally, H.W. Rollins, D.K. Jamison, Electrochim.
Acta 209 (2016) 36–43.
[10] P. Wisian-Neilson, in: R.B. King (Ed.), Encycl. Inorg. Chem, 1st ed., John Wiley &
Sons, Chichester, England, 1994, pp. 3371–3384.
4.7. Synthesis of P,P,P-tris(2,2,2-trifluoroethoxy)phosphoraniminyl-N-
1,3-dimethylimidazolinium hexafluorophosphate, (3)
[11] Y.G. Gololobov, L.F. Kasukhin, Tetrahedron 48 (1992) 1353–1406.
[12] P. Wisian-Neilson, R.H. Neilson, J. Am. Chem. Soc. 102 (1980) 2848–2849.
[13] C.H. Honeyman, I. Manners, C.T. Morrissey, H.R. Allcock, J. Am. Chem. Soc. 117
(1995) 7035–7036.
[14] H.R. Allcock, C.A. Crane, C.T. Morrissey, J.M. Nelson, S.D. Reeves, C.H.
Honeyman, I. Manners, Macromolecules 29 (1996) 7740–7747.
[15] V. Blackstone, A.J. Lough, M. Murray, I. Manners, J. Am. Chem. Soc. 131 (2009)
3658–3667.
[16] V. Blackstone, S. Pfirrmann, H. Helten, A. Staubitz, A. Presa Soto, G.R. Whittell, I.
Manners, J. Am. Chem. Soc. 134 (2012) 15293–15296.
[17] V. Vreeken, M.A. Siegler, B. de Bruin, J.N.H. Reek, M. Lutz, J.I. van der Vlugt,
Angew. Chem. Int. Ed. 54 (2015) 7055–7059.
[18] A. Steiner, S. Zacchini, P.I. Richards, Coord. Chem. Rev. 227 (2002) 193–216.
[19] D.W. Stephan, J.C. Stewart, F. Guerin, S. Courtenay, J. Kickham, E. Hollink, C.
Beddie, A. Hoskin, T. Graham, P. Wei, R.E.V. H. Spence, W. Xu, L. Koch, X. Gao, D.
G. Harrison, Organometallics 22 (2003) 1937–1947.
[20] M.A. Benson, J. Ledger, A. Steiner, Chem. Commun. (2007) 3823–3825.
[21] F. Dielmann, C.E. Moore, A.L. Rheingold, G. Bertrand, J. Am. Chem. Soc. 135
(2013) 14071–14073.
Following the procedure as for PA 2, 10 mL of anhydrous
toluene was introduced into the flask by syringe, and then
ADMImPF (2.0 g, 0.007 mol) was added directly into the flask. TTFP
(1.54 mL, 2.3 g, 0.007 mol) was added dropwise by syringe into the
flask. As the phosphine addition progressed, no color was observed
during the reaction but remains as a slurry. 20–30 min after the
phosphine was added, no color was observed during the reaction.
No apparent off gassing (nitrogen) is observed and the reaction
was left at room temperature overnight under nitrogen. Next
day, stirring was stopped and the white, powdery product was
allowed to settle. The product was collected by decanting the
toluene and the product was washed twice with 30–40 mL of hex-
anes. The powdery solids were dried by a nitrogen gas stream over-
night. The isolated product is moderately stable in air from days to
weeks depending on humidity. The product was recrystallized in
boiling THF, and it is allowed to slowly cool to room temperature.
The final product forms colorless, clear prismatic crystals. How-
ever, long-term stability of 3 was an issue after exposure to mois-
ture, and 4 is the decomposition product of 3. [85% yield, m.p.
[22] N. Burford, T.S. Cameron, D.J. LeBlanc, A.D. Phillips, T.E. Concolino, K.-C. Lam, A.
L. Rheingold, J. Am. Chem. Soc. 122 (2000) 5413–5414.
[23] R. Boomishankar, J. Ledger, J.-B. Guilbaud, N.L. Campbell, J. Bacsa, R. Bonar-
Law, Y.Z. Khimyak, A. Steiner, Chem. Commun. (2007) 5152–5154.
[24] O. Alhomaidan, G. Bai, D.W. Stephan, Organometallics 27 (2008) 6343–6352.
[25] N.J. Van Zee, M.J. Sanford, G.W. Coates, J. Am. Chem. Soc. 138 (2016) 2755–
2761.
[26] J.K. Ruff, W.J. Schlientz, R.E. Dessy, J.M. Malm, G.R. Dobson, M.N. Memering, in:
G.W. Parshall (Ed.), Inorg. Synth, Wiley & Sons Inc, John, 1974, pp. 84–90.
[27] R.H. Neilson, J.R. Klaehn, J. Inorg. Organomet. Polym. Mater. 16 (2006) 319–
326.
[28] J.R. Klaehn, R.H. Neilson, Inorg. Chem. 41 (2002) 5859–5865.
[29] H. Staudinger, J. Meyer, Helv. Chim. Acta 2 (1919) 619–635.
[30] K. Matyjaszewski, M.S. Lindenberg, M.K. Moore, M.L. White, J. Polym. Sci. Part
Polym. Chem. 32 (1994) 465–473.
= 97–98 °C, 31P NMR
d
(d6-Acetone) = (s) ꢀ14.8, (septet,
J
PF = 707.0 Hz) ꢀ144.3; 19F NMR d (d6-Acetone) = (d, JPF = 711.4 Hz)
ꢀ72.6, (s) ꢀ75.7; 1H NMR d (d6-Acetone) = (quartets, JHF = 7.8 Hz)
5.11, (s) 3.83, (s) 3.09; 13C NMR d (d6-Acetone) = (d, JPC = 22.7 Hz)
156.0, (q, JCF = 274.8 Hz) 122.4, (d,, JPC = 7.6 Hz, JCF = 31.7 Hz) 30.5,
(s) 47.2, (s) 32.2].
[31] J. Haase, in: S. Bräse, K. Banert (Eds.), Org Azides, John Wiley & Sons Ltd, 2009,
pp. 29–51.
Acknowledgements
[32] H.C. Kolb, M.G. Finn, K.B. Sharpless, Angew. Chem. Int. Ed. 40 (2001) 2004–
2021.
[33] V. Hong, N.F. Steinmetz, M. Manchester, M.G. Finn, Bioconjug. Chem. 21 (2010)
1912–1916.
[34] H. Yoon, J.M. Lim, H.-C. Gee, C.-H. Lee, Y.-H. Jeong, D. Kim, W.-D. Jang, J. Am.
Chem. Soc. 136 (2014) 1672–1679.
[35] F. Shabanpoor, M.J. Gait, Chem. Commun. 49 (2013) 10260–10262.
[36] M. Kitamura, T. Koga, M. Yano, T. Okauchi, Synlett 23 (2012) 1335–1338.
[37] M. Kitamura, S. Kato, M. Yano, N. Tashiro, Y. Shiratake, M. Sando, T. Okauchi,
Org. Biomol. Chem. 12 (2014) 4397–4406.
[38] M. Kitamura, M. Yano, N. Tashiro, S. Miyagawa, M. Sando, T. Okauchi, Eur. J.
Org. Chem. 2011 (2011) 458–462.
[39] K. Huynh, C.P. Chun, A.J. Lough, I. Manners, Dalton Trans. 40 (2011) 10576–
10584.
[40] K. Huynh, E. Rivard, A.J. Lough, I. Manners, Chem. Eur. J. 13 (2007) 3431–3440.
[41] K. Huynh, A.J. Lough, M.A.M. Forgeron, M. Bendle, A.P. Soto, R.E. Wasylishen, I.
Manners, J. Am. Chem. Soc. 131 (2009) 7905–7916.
Work supported through the Idaho National Laboratory - Labo-
ratory Directed Research and Development (LDRD) Program, Pro-
ject 15-125, under United States - Department of Energy (US-
DOE) Idaho Operations Office (Contract No. DE-AC07-05ID14517).
This manuscript has been authored by Battelle Energy Alliance,
LLC under Contract No. DE-AC07-05ID14517 with the U.S. Depart-
ment of Energy. The United States Government retains and the
publisher, by accepting the article for publication, acknowledges
that the United States Government retains a nonexclusive, paid-
up, irrevocable, world-wide license to publish or reproduce the
published form of this manuscript, or allow others to do so, for Uni-
ted States Government purposes.
[42] C. Imrie, T.A. Modro, P.H. Van Rooyen, C.C.P. Wagener, K. Wallace, H.R. Hudson,
M. McPartlin, J.B. Nasirun, L. Powroznyk, J. Phys. Org. Chem. 8 (1995) 41–46.
[43] H. Zimmer, G. Singh, J. Org. Chem. 28 (1963) 483–486.
[44] S.P. Belov, I.V. Komlev, S.V. Kuznetsov, E.E. Nifant’ev, Russ. J. Gen. Chem. 79
(2009) 2698–2699.
References
[1] C.W. Allen, Coord. Chem. Rev. 130 (1994) 137–173.
[2] P. Potin, R. De Jaeger, Eur. Polym. J. 27 (1991) 341–348.
[3] R.H. Neilson, P. Wisian-Neilson, Chem. Rev. Wash. DC U. S. 88 (1988) 541–562.
[4] H.R. Allcock, F.W. Lampe, Contemporary Polymer Chemistry, Prentice-Hall,
1981.
[5] A.K. Andrianov, Polyphosphazenes for Biomedical Applications, John Wiley and
Sons, 2009.
[6] M. Carenza, S. Lora, L. Fambri, in: N. Hasirci, V. Hasirci (Eds.), Biomaterials,
Springer, US, 2004, pp. 113–122.
[7] M. Gleria, R.D. Jaeger, Phosphazenes: A Worldwide Insight, Nova Publishers,
2004.
[8] R.D. Jaeger, M. Gleria, Applicative Aspects of Poly(organophosphazenes), Nova
Publishers, 2004.
[45] T.W. Rave, J. Org. Chem. 32 (1967) 3461–3466.
[46] L. Maier, P.J. Diel, Phosphorus Sulfur Silicon Relat. Elem. 115 (1996) 273.
[47] I. Alkorta, G. Sánchez-Sanz, J. Elguero, J.E. Del Bene, J. Phys. Chem. A 118 (2014)
1527–1537.
[48] M.H. Holthausen, I. Mallov, D.W. Stephan, Dalton Trans. 43 (2014) 15201–
15211.
[49] F. Weller, H.-C. Kang, W. Massa, T. Rübenstahl, F. Kunkel, Kurt Dehnicke, Z.
Naturforsch. 50b (1995) 1050–1054.
[50] C. Knapp, R. Uzun, Acta Crystallogr. Sect. E Struct. Rep. Online 66 (2010).
o3185–o3185.
[51] MNova NMR, V. 10.0.2, Mestrelab Research S.L., Feliciano Barrera, Spain, 2016.
[52] APEX2 Software Suite V. 2014.11-0, Bruker AXS Inc., Madison, WI, 2014.