Unexpected Reactivity of Red Phosphorus in Ionic Liquids

Article abstract of DOI:10.1002/ejic.201500502

Red phosphorus is far less reactive than the white allotrope. On the other hand, it is easier to handle and not as toxic as white phosphorus. In the Lewis-acidic ionic liquid (IL) [BMIm]Cl·2AlCl₃ ([BMIm] = 1-butyl-3-methylimidazolium), red phosphorus and elemental iodine form several iodides at moderate temperature. ³¹P liquid- and solid-state NMR spectroscopy was used to rationalize the reaction at various temperatures and ratios of the starting materials. Monitoring of the reaction revealed nanoscale red-phosphorus particles. In addition to this top-down formation, phosphorus nanoparticles were also obtained in a bottom-up synthesis by dissociation of P₂I₄ in the IL. Depending on the ratio of red phosphorus and iodine, as well as the reaction temperature, P₂I₄, PI₃, or P₂I₅⁺ dominate. Red phosphorus readily reacts with iodine in a Lewis-acidic ionic liquid at moderate temperature. Besides iodides, phosphorus nanoparticles are formed. The latter are also obtained in a bottom-up process starting from P₂I₄. In situ liquid- and solid-state NMR spectroscopy reveals details of the reactions.

Full text of DOI:10.1002/ejic.201500502

FULL PAPER
DOI:10.1002/ejic.201500502
Unexpected Reactivity of Red Phosphorus in Ionic Liquids
[
a]
[a]
[a]
Matthias F. Groh, Silvia Paasch, Alexander Weiz,
Michael Ruck,*[
a,b]
[a]
COVER PICTURE
and Eike Brunner*
Dedicated to Manfred Scheer on the occasion of his 60th birthday
Keywords: In situ spectroscopy / Ionic liquids / Nanoparticles / NMR spectroscopy / Phosphorus
Red phosphorus is far less reactive than the white allotrope.
On the other hand, it is easier to handle and not as toxic as
white phosphorus. In the Lewis-acidic ionic liquid (IL)
peratures and ratios of the starting materials. Monitoring of
the reaction revealed nanoscale red-phosphorus particles. In
addition to this top-down formation, phosphorus nanopar-
ticles were also obtained in a bottom-up synthesis by dissoci-
[
BMIm]Cl·2AlCl
3
([BMIm] = 1-butyl-3-methylimidazolium),
red phosphorus and elemental iodine form several iodides at
moderate temperature. P liquid- and solid-state NMR spec-
troscopy was used to rationalize the reaction at various tem- PI
ation of P
2 4
I in the IL. Depending on the ratio of red phos-
3
1
phorus and iodine, as well as the reaction temperature, P I ,
2 4
+
3
2 5
, or P I dominate.
Introduction
gations in ILs with NMR techniques. As we have shown
previously, red phosphorus is reactive in [BMIm]Cl·nAlCl3
Various inorganic materials can be synthesized at or near
ambient conditions with the help of ionic liquids (ILs).
This approach enables reduction of energy usage and tech-
nical efforts as well as better control of the reactions com-
pared to commonly applied high-temperature processes.
[2a,2o]
if combined with a suitable reaction partner.
The cho-
[
1,2]
sen amount of AlCl is the highest one for a homogeneous
3
system at ambient conditions.
[
2a]
Moreover, syntheses in ILs provide great opportunities to
discover new compounds with potentially outstanding and
useful chemical and physical properties.[ Still, the mecha-
nisms of product formation are barely examined although
this understanding would be crucial for the directed synthe-
sis of a desired material. A (semi-)continuous reaction mon-
itoring with high-resolution (HR) liquid-state NMR as-
sisted by solid-state NMR is especially suitable for mechan-
Results and Discussion
3]
To facilitate NMR signal assignment, the red phos-
phorus used as starting material as well as the potential
31
reaction products PI and P I were analyzed by using
P
3
2 4
solid- and liquid-state NMR spectroscopy (see Supporting
Information). The measurements were carried out between
room temperature (RT) and 80 °C; the latter being the high-
est accessible temperature. Red phosphorus as an amorph-
3
1
istic investigations. Already in 1967, early P NMR mea-
surements had been applied successfully to follow the reac-
[5]
ous solid has a poorly defined structure. Consequently, its
[4]
tion of white phosphorus with iodine in CS2. We decided
to substitute white phosphorus by its more convenient
physical and chemical properties are variable to some de-
gree, and the linewidths and exact chemical shift of the
NMR signals can vary for different batches and sources.
Nevertheless, red phosphorus can easily be distinguished
from the other common allotropes by its chemical shift. We
observed a broad line centered at about 50 ppm (Figures S1
and S2, Supporting Information) in agreement with the lit-
erature, which reports values of 60 to 65 ppm. Black
phosphorus appears at δ = 22.2 ppm, Hittorf’s phos-
phorus exhibits a broader and more complex spectrum
ranging from –84.5 to 171.3 ppm, and white phosphorus
red allotrope and the neurotoxic and highly reactive CS2
with the comparatively inert and redox-innocent IL
[
BMIm]Cl·2AlCl3
([BMIm]
=
1-butyl-3-methylimid-
azolium) in order to assess the viability of in situ investi-
[
a] Fachrichtung Chemie und Lebensmittelchemie, Technische
Universität Dresden
[6]
[
7]
0
1062 Dresden, Germany
E-mail: michael.ruck@tu-dresden.de
eike.brunner@tu-dresden.de
[8]
http://www.chm.tu-dresden.de/ac2/
[
9]
http://www.chm.tu-dresden.de/anc1/
resonates at δ = –527 ppm. Crystalline fibrous phos-
[
b] Max-Planck-Institut für Chemische Physik fester Stoffe
Nöthnitzer Str. 40, 01187 Dresden, Germany
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejic.201500502.
[10]
phorus
has not yet been characterized by NMR spec-
troscopy, but owing to its structural similarity to Hittorf’s
phosphorus, one can expect a similarly complex spectrum.
Eur. J. Inorg. Chem. 2015, 3991–3994
3991
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
One of the two non-crystalline, fibrous forms mimics the are cut off the solid. A rough estimation of the particle size
NMR signal of red phosphorus at δ = 65 ppm. However, based on the linewidth yielded diameters of 20Ϯ10 nm (see
this modification of phosphorus has only been obtained Supporting Information).
[
11]
from a CuI matrix.
For the monitoring of the first reaction, we chose a mo-
lar ratio P:I = 1:2. PI forms upon mixing phosphorus and
2
3
3
1
iodine in the IL. This is visible in the P NMR spectrum
at 25 °C (Figure 1, top), which is dominated by the CSA-
broadened signal characteristic of PI (Figure S3, Support-
3
ing Information). After heating to 80 °C inside the spec-
trometer, all PI is in liquid phase. Ex situ conditioning at
3
100 °C results in the formation of a phosphorus iodide dif-
ferent from PI . It gives rise to two rather broad signals at
3
room temperature (provided the solution is quickly cooled).
These signals at about 130 ppm and –150 ppm (Figure 1)
Figure 2. 31P NMR spectra of the reaction mixture P:I
2
= 1:1
+
are characteristic of P2I5 (solid P I AlI : 114 ppm for R– (phosphorus batch 1). Top: liquid-state at room temperature
2 5
4
PI2, –142 ppm for R–PI ;[
12]
P I [Al(OC(CF ) ) ] in (black) and at 80 °C (red) after mixing. Middle: liquid-state at RT
3
2 5 3 3 4
[13]
+
(black) and at 80 °C (red) after ex situ annealing at 100 °C for 24 h.
Bottom: 14 kHz MAS spectra at RT of the used red phosphorus
CD Cl : 127.4 and –156.3 ppm ). The P2I5 cation domi-
2
2
nates after conditioning at 120 °C (and subsequent fast
cooling) and vanishes completely upon annealing at 55 °C,
(
blue) and of the precipitate after conditioning at 120 °C (black).
The asterisk marks the signal at δ = 166 ppm.
while PI remains the only visible species. Therefore, we can
propose a reaction scheme for the P:I2 = 1:2 case
3
In addition, an unknown species with a chemical shift of
166 ppm is visible (Figure 2, marked with an asterisk). The
latter must be either a monophosphorus or a symmetric
diphosphorus species similar to PI , P I , or the –PI part
(Scheme 1).
3
2 4
2
+
31
of P I . Owing to strong spin-orbit coupling, the
P
2
5
signals of four-coordinate phosphorus atoms bonded to
+
iodine atoms are strongly shifted upfield (e.g. PI , –PI
4
3
+
[12–15]
part of P I ).
By quantum chemical calculations, the
hypothetical ions PI2 and P2I4 could also be ruled out.
2
5
+
2+
The chemical shift of 166 ppm resembles PI dissolved in IL
3
(δ =187 ppm). Since this signal is rather broad, we assume it
is related to PI forming on the surface of the nanoparticles.
3
After conditioning the reaction solution at 100 °C for
2
4 h, the red phosphorus is completely converted into P2I4
and PI , and the signals at δ = 166 ppm and 50 ppm have
1:2 (phosphorus batch 1). Top: at room temperature (black) and vanished (Figure 2, middle and bottom, Figure S7, Sup-
Figure 1. Liquid-state 31P NMR spectra of a reaction mixture P:I
2
3
=
at 80 °C (red) after mixing. Middle: at RT (black) and at 80 °C
red) after ex situ annealing at 100 °C for 15 h and fast cooling.
porting Information). The annealing seems to reduce the
solubility of the iodides by aggregation. Correspondingly,
their NMR signals are broad. At the chosen element ratio,
(
Bottom: at RT (black) and at 55 °C (red) after annealing at 120 °C
for 24 h and fast cooling.
the reaction is limited by the equilibrium between PI and
3
P I . We propose a two-step reaction scheme (Scheme 2).
2
4
Scheme 1. Proposed reaction of red phosphorus and iodine at the
2 3
molar ratio P:I = 1:2 in [BMIm]Cl·2AlCl .
Scheme 2. Proposed reaction of red phosphorus and iodine at the
molar ratio P:I = 1:1 in [BMIm]Cl·2AlCl
2
3
.
A molar ratio of the starting materials of P:I = 1:1 also
2
leads to the formation of PI upon mixing of the elements
Exactly this two-step process was observed by Carroll et
3
[
4]
in the IL at 25 °C, as indicated by the characteristic broad al. for white phosphorus in CS2. However, the reaction in
3
1
signal in the P NMR spectrum (Figure 2, top). After heat- CS proceeds to P I at room temperature. By increasing
2
2 4
ing the sample to 55 °C inside the spectrometer, an ad- the phosphorus to iodine ratio in the reaction mixture to
ditional signal at about 50 ppm appears which intensifies at 2:1, the formation of PI is disfavored, and P I becomes
3
2 4
higher temperature (Figures 2 and S6, Supporting Infor- the only detectable reaction product at 70 °C (Figure S5,
mation). The chemical shift is characteristic of red phos- Supporting Information; keeping in mind that solid red
phorus (Figure S1, Supporting Information) but the signal phosphorus is virtually invisible).
is rather narrow. We assume that, in the course of etching
We have used two batches of red phosphorus from dif-
of the red phosphorus with dissolved iodine, nanoparticles ferent sources. The 31P MAS NMR spectrum revealed a
Eur. J. Inorg. Chem. 2015, 3991–3994
3992
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
minor difference between the two batches visible in an up- (see EDX line scans and mapping, Figures S10 to S12, Sup-
field “shoulder” of the signal of the first sample (Figure S2, porting Information). Upon washing with dichloromethane
Supporting Information). The intensity of the nanoparticle (DCM) and ethanol (96%), the surfactants are either re-
signal at about 50 ppm depends on the chosen batch moved or converted into undefined compounds containing
(source) of red phosphorus as starting material. A second aluminum, carbon, and oxygen.
batch of red phosphorus yielded fewer nanoparticles under
otherwise same conditions. Moreover, the second batch of
red phosphorus enabled the synthesis of P I at lower tem-
2
4
perature (Ͻ70 °C) and thus monitoring of the reaction in
the spectrometer (Figure S8, Supporting Information; note
the isosbestic point at δ = 150 ppm). The lower reaction
temperature gives rise to much lower signal intensity for the
nanoparticles because phosphorus is mainly transformed
into iodides before appearing as a signal at δ = 50 ppm in
the spectrum. Moreover, differences in the (micro-)structure Figure 4. SEM image (SE detector, left) and STEM image (TE de-
of the amorphous solid seem to correlate with different tector, right) of hollow phosphorus nanoparticles after washing
with DCM and EtOH (96%). The insets show the EDX mapping
etching behavior, perceived as the intensity of the nanopart-
for phosphorus (red, left) and aluminum (blue, right). Note the
icle signal. Together with the signal of the phosphorus
second wall formed by aluminum-containing surfactants within the
particles.
nanoparticles at δ = 50 ppm, the additional signal at δ =
66 ppm – attributed to phosphorus iodides at the particle
1
surface (Figure 2, top) – vanishes.
P I , taken as starting material, partly dissociates to PI
2
4
3
Conclusions
and phosphorus in the IL at 80 °C (Figure S5, Supporting
Information). Thereby, a small amount of yellowish powder
is formed (Figures 3 and 4 and Figure S9, Supporting Infor-
mation) and a narrow line at δ = 50 ppm, which indicates
phosphorus nanoparticles (vide supra), appears in the
NMR spectrum. This resembles the observations by Carroll
et al. as well as Baudler et al. who described an iodine-
containing colloidal phosphorus polymer as one of the
The following can be concluded: (i) The reactivity of red
phosphorus is strongly increased by iodine in [BMIm]Cl·
2
AlCl . (ii) The reaction pathways can be followed in detail
3
31
by a combination of liquid- and solid-state P NMR
spectroscopy. (iii) Phosphorus nanoparticles, terminated by
iodine on the surface, are formed in the course of the reac-
tion (top-down). Similar nanoparticles are formed as a
dissociation product of P I (bottom-up). The discovery of
[4,16]
dissociation products of P I in CS .
On the basis of
2
4
2
2
4
the linewidth, we estimate particle sizes of 22Ϯ10 nm (cf.
Supporting Information). We were unable to isolate these,
owing to hindered sedimentation of very small particles, but
detected massive (Figure 3) as well as hollow, spherical
particles (Figure 4) of 60 to 500 nm in the scanning electron
phosphorus nanoparticles may stimulate the development
of a reactive form of phosphorus without the drawbacks of
the white allotrope.
microscope (SEM). Chloridoaluminates originating from Experimental Section
the IL cover the outer and inner surfaces of the particles
Sample Preparation: All compounds were handled in an argon-
filled glove box (M. Braun; p(O )/p Ͻ 1 ppm, p(H O)/p Ͻ 1 ppm).
2 2
0
0
Starting materials: phosphorus “batch 1” (Sigma Aldrich, 99.95%),
phosphorus “batch 2” (ABCR, 99.999%), iodine (Fluka, 98%),
AlCl (Fluka 98%, sublimated three times), and [BMIm]Cl (Sigma
3
Aldrich, Ն 98% HPLC, dried in dynamic vacuum). Both phos-
phorus samples were washed with sodium hydroxide, refluxed in
water, and dried in vacuo.[ PI
17]
3 2 4
(ABCR, 98%) and P I (Aldrich,
9
5%) were used without further purification. To utilize the full
range of the NMR probe, a minimum of 982 mg (5.6 mmol) of
BMIm]Cl and the corresponding amount of AlCl (n(AlCl ) =
n([BMIm]Cl)) was filled in a commercial 10 mm NMR glass tube.
Owing to the high enthalpy of mixing of [BMIm]Cl and AlCl as
[
2
3
3
3
well as the absence of a stirring bar, the two compounds were alter-
natingly mixed before phosphorus (n(P) = 0.27–0.47n([BMIm]Cl))
and iodine or P I (n(P) = 0.27n([BMIm]Cl)) were added. The glass
2 4
tubes were sealed under dynamic vacuum. Subsequently, the reac-
tion mixture was cautiously homogenized with a vortex mixer.
NMR Spectroscopy: All liquid-state NMR experiments were car-
ried out on a Bruker Avance 300 spectrometer at a phosphorus
resonance frequency of 121.5 MHz by using a commercial 10 mm
HR probe. MAS NMR experiments were performed on a Bruker
Figure 3. SEM images (SE detector) of phosphorus nanoparticles
with diameters between 80 and 430 nm on a silicon wafer after
washing with DCM and EtOH (96%).
Eur. J. Inorg. Chem. 2015, 3991–3994
3993
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
FULL PAPER
Avance 300 spectrometer with commercial double resonance 2.5
and 4 mm MAS NMR probes operating at a resonance frequency
of 121.5 MHz and a MAS frequency of 14 kHz. The phosphorus
Isaeva, M. Ruck, Z. Anorg. Allg. Chem. 2015, 641, 279–284;
m) M. F. Groh, J. Breternitz, E. Ahmed, A. Isaeva, A. Efimova,
P. Schmidt, M. Ruck, Z. Anorg. Allg. Chem. 2015, 641, 388–
3
93; n) T. Alammar, Y.-K. Chow, A.-V. Mudring, New J. Chem.
3 4
chemical shifts were referenced relative to H PO .
2015, 39, 1339–1347; o) K.-O. Feldmann, T. Wiegand, J. Ren,
H. Eckert, J. Breternitz, M. F. Groh, U. Müller, M. Ruck, B.
Maryasin, C. Ochsenfeld, O. Schön, K. Karaghiosoff, J. J. Wei-
gand, Chem. Eur. J. 2015, DOI: 10.1002/chem.201406476.
3] a) A. M. Guloy, R. Ramlau, Z. Tang, W. Schnelle, M. Bait-
inger, Y. Grin, Nature 2006, 443, 320–323; b) E. Ahmed, J.
Beck, J. Daniels, Th. Doert, S. J. Eck, A. Heerwig, A. Isaeva,
S. Lidin, M. Ruck, W. Schnelle, A. Stankowski, Angew. Chem.
Int. Ed. 2012, 51, 8106–8109; Angew. Chem. 2012, 124, 8230–
8233.
SEM/STEM: Scanning (transmission) electron microscopy was
performed by using a SU8020 (Hitachi) with a triple detector sys-
tem for secondary and low-energy backscattered electrons (U
0–30 kV). Energy-dispersive X-ray (EDX) spectra were collected
= 8–30 kV) using a Silicon Drift Detector (SDD) X-MaxN (Ox-
a
=
[
1
(U
a
ford). Prior to the measurements, the samples were washed with
dry DCM and subsequently with EtOH (96%), and in each wash-
ing step centrifuged and redispersed using a supersonic bath. Fi-
nally, the dispersed particles were transferred either on a TEM grid
or a polished silicon wafer.
[4] R. L. Carroll, R. P. Carter, Inorg. Chem. 1967, 6, 401–403.
[
5] a) J. S. Lannin, B. V. Shanabrook, F. Gompf, J. Non-Cryst. Sol-
ids 1982, 49, 209–219; b) G. Fasol, M. Cardona, W. Hönle,
H. G. von Schnering, Solid State Commun. 1984, 52, 307–310;
c) D. J. Olego, J. A. Baumann, M. A. Kuck, R. Schachter, C. G.
Michel, Solid State Commun. 1984, 52, 311–314; d) H. Farman,
D. Blakey, J. C. Dore, M.-C. Bellissent-Funel, S. R. Elliott,
Phys. Scr. 1994, 49, 634–636; e) P. Jovari, L. Pusztai, Appl.
Phys. A 2002, 74, S1092–S1094.
XRD: Data collection was performed at 296(1) K on an XЈPert Pro
MPD diffractometer (PANalytical) equipped with
Ge(111) monochromator by using Cu-Kα1 radiation (λ
54.056 pm). The air-sensitive samples were protected by a
a curved
=
1
polyimide (Kapton®) film (Chemplex).
[
6] a) A. Bytchkov, F. Fayon, D. Massiot, L. Hennet, D. L. Price,
Phys. Chem. Chem. Phys. 2010, 12, 1535–1542; b) T. Rami-
reddy, T. Xing, M. M. Rahman, Y. Chen, Q. Dutercq, D. Gun-
zelmann, A. M. Glushenkov, J. Mater. Chem. A 2015, 3, 5572–
5584.
Acknowledgments
We gratefully acknowledge J. Pallmann (TU Dresden) for her help
with the MAS NMR measurements. This work was supported by
the Deutsche Forschungsgemeinschaft (DFG).
[7] S. Lange, P. Schmidt, T. Nilges, Inorg. Chem. 2007, 46, 4028–
4035.
[
8] T. Wiegand, H. Eckert, S. Grimme, D. Hoppe, M. Ruck, Chem.
Eur. J. 2011, 17, 8739–8748.
[9] M. Di Vaira, M. Peruzzini, S. S. Constantini, P. Stoppioni, J.
Organomet. Chem. 2006, 691, 3931–3937.
[10] M. Ruck, D. Hoppe, B. Wahl, P. Simon, Y. Wang, G. Seifert,
Angew. Chem. Int. Ed. 2005, 44, 7616–7619; Angew. Chem.
2005, 117, 7788–7792.
[11] A. Pfitzner, M. F. Bräu, J. Zweck, G. Brunklaus, H. Eckert,
Angew. Chem. Int. Ed. 2004, 43, 4228–4231; Angew. Chem.
2004, 116, 4324–4327.
[
[
1] Ionic Liquids in Synthesis (Eds.: P. Wasserscheid, T. Welton),
nd ed., Wiley-VCH, Weinheim 2007.
2
2] a) M. Groh, U. Müller, E. Ahmed, A. Rothenberger, M. Ruck,
Z. Naturforsch., B: J. Chem. Sci. 2013, 68, 1108–1122; b) A.
Taubert, Z. Li, Dalton Trans. 2007, 723–727; c) K. Biswas, Q.
Zhang, I. Chung, J.-H. Song, J. Androulakis, A. J. Freeman,
M. G. Kanatzidis, J. Am. Chem. Soc. 2010, 132, 14760–14762;
d) D. Freudenmann, S. Wolf, M. Wolff, C. Feldmann, Angew.
Chem. Int. Ed. 2011, 50, 11050–11060; Angew. Chem. 2011, [12] C. Aubauer, G. Engelhardt, T. M. Klapötke, A. Schulz, J.
1
23, 11244–11255; e) C. Vollmer, C. Janiak, Coord. Chem. Rev.
Chem. Soc., Dalton Trans. 1999, 1729–1734.
[13] I. Krossing, J. Chem. Soc., Dalton Trans. 2002, 500–512.
[14] K. B. Dillon, M. G. Craveirinha Dillon, T. C. Waddington, J.
Inorg. Nucl. Chem. 1976, 38, 1149–1150.
2
011, 255, 2039–2057; f) M. Wolff, J. Meyer, C. Feldmann, An-
gew. Chem. Int. Ed. 2011, 50, 4970–4973; Angew. Chem. 2011,
1
23, 5073–5077; g) E. Ahmed, J. Breternitz, M. F. Groh, M.
Ruck, CrystEngComm 2012, 14, 4874–4885; h) M. F. Groh, A.
Isaeva, M. Ruck, Chem. Eur. J. 2012, 18, 10886–10891; i) K.
Schütte, H. Meyer, C. Gemel, J. Barthel, R. A. Fischer, C. Jan-
iak, Nanoscale 2014, 6, 3116–3126; j) S. Tyrrell, M. Swadzba-
Kwasny, P. Nockemann, J. Mater. Chem. A 2014, 2, 2616–2622;
k) G. Gebresilassie Eshetu, M. Armand, B. Scrosati, S. Pass-
erini, Angew. Chem. Int. Ed. 2014, 53, 13342–13359; Angew.
Chem. 2014, 126, 13558–13576; l) M. F. Groh, M. Knies, A.
[15] M. Kaupp, C. Aubauer, G. Engelhardt, T. M. Klapötke, O. L.
Malkina, J. Chem. Phys. 1999, 110, 3897–3902.
[16] M. Baudler, P. Junkes, G. Sadri, Z. Naturforsch., B: J. Chem.
Sci. 1971, 26, 759–762.
[17] G. Brauer, Handbuch der Präparativen Anorganischen Chemie,
3rd ed., vol. 1, F. Enke, Stuttgart 1962, p. 506.
Received: May 8, 2015
Published Online: June 3, 2015
Eur. J. Inorg. Chem. 2015, 3991–3994
3994
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim