Alkali metals plus silica gel: Powerful reducing agents and convenient hydrogen sources

Article abstract of DOI:10.1021/ja051786+

Alkali metals absorbed into silica gel yield three stages of unique loose black powders (M-SG) that are strong reducing agents. All react nearly quantitatively with water to form hydrogen. Liquid Na-K alloys form air-sensitive powders at room temperature that can be converted at 150 °C to a form that is sensitive to moisture but not to dry air. Slowly heating sodium and silica gel to 400 °C yields a third type that can be handled in ambient air with only slow degradation by atmospheric moisture. These materials eliminate many hazards associated with pure alkali metals and provide easily handled reducing agents and hydrogen sources. They could be used in continuous-flow reactors to reduce and protonate aromatics, dechlorinate alkyl and aryl halides, and desulfurize various compounds. Copyright

Full text of DOI:10.1021/ja051786+

Published on Web 06/14/2005
Alkali Metals Plus Silica Gel: Powerful Reducing Agents and Convenient
Hydrogen Sources
James L. Dye,*,†,‡ Kevin D. Cram,† Stephanie A. Urbin,† Mikhail Y. Redko,† James E. Jackson,† and
Michael Lefenfeld*,‡
Department of Chemistry, Michigan State UniVersity, East Lansing, Michigan 48824, and SiGNa Chemistry, LLC,
530 East 76th Street, Suite 9E, New York, New York 10021
Received March 21, 2005; E-mail: dye@signachem.com; michael@signachem.com
Alkali metals absorbed into silica gel yield three types of loose
black powders (M-SG) that are strong reducing agents. All react
nearly quantitatively with water to form hydrogen. Liquid NaK
alloys form air-sensitive powders at room temperature that can be
converted at 150 °C to a form that is sensitive to moisture but not
to dry air. Slowly heating sodium and silica gel to 400 °C yields a
third type that can be handled in ambient air with only slow
degradation by atmospheric moisture. These materials eliminate
many hazards associated with pure alkali metals by providing easily
handled reducing agents and hydrogen sources. In continuous-flow
reactors, they could reductively protonate aromatics, dechlorinate
Figure 1. Room-temperature absorption of Na2K alloy into Davisil Grade
646 silica gel (Aldrich, 15-nm nominal pore size) in a 50-mL modified
Erlenmeyer flask. (A) The SG coated with alkali metal alloy just after
alkyl and aryl halides, and desulfurize various compounds.
Synthetic and industrial chemistry has long tried to take full
advantage of the reducing power of alkali metals. They have been
used as dispersions, or on inert supports, or as solutions in liquid
ammonia1 to carry out Birch reductions, Wurtz reactions, and many
others.2 Although industrially useful, these reactions can be difficult
to scale up because of low reaction efficiencies or rates and
complications in handling such pyrophoric reagents.
Alkali metals, at concentrations up to 12 mol %, can be absorbed
from the vapor phase into aluminosilicate3 or pure silica zeolites.4
This report describes the incorporation of up to 60 mol % alkali
metals from the liquid metal phase into amorphous silica gel (porous
SiO2). Three different preparations of alkali metal-silica gel adducts
(designated M-SG, stages 0, I, II) provide a range of reducing
abilities and air-sensitivities.
Liquid sodium-potassium alloys are readily absorbed by silica
gel (SG) to form shiny black stage 0 powders. Figure 1 shows the
conversion of SG particles, initially coated with 35 wt % of a liquid
Na2K alloy, to a powder that has absorbed all the alloy. Some
nanoscale alkali metal particles are present, as shown by the
characteristic reduction of the melting point commonly seen for
small particles.5 The differential scanning calorimetry (DSC) trace
for stage 0 Na2K (Figure 2A) shows that melting begins at -25 °C
rather than the normal value of -12.6 °C. Stage 0 powders must
be handled in an inert atmosphere (e.g., within a nitrogen-filled
glovebag) to avoid hydrogen ignition in high humidity air. However,
they are loose powders that can easily be packed in chromatographic
columns to provide a medium for continuous-flow chemical
reductions. Of course, the solvent used must not be reducible by
M-SG.
mixing. (B) The final loose black powder (stage 0) formed by shaking the
sample shown in A for a few minutes.
been stored for up to eight months in screw-cap vials without any
change in the reducing capacity.
Uniform black powders of stage I Na-SG can be made by
heating Na with SG to ∼165 °C with continual agitation. DSC traces
show no melting endotherm of Na metal but only a broad exotherm
(∆H ≈ -25 kJ/mol of Na) from 100 to 450 °C. When heated slowly
to 400 °C with intermittent agitation, sodium metal and SG react
to form uniform powders of stage II Na-SG. The bulk density at
this concentration is about 0.7 g/cm3. DSC traces of Na with SG
(Figure 2C) show that the process is exothermic, with ∆H in the
range of -75 kJ/mol of metal. The melting endotherm of Na metal,
present in the initial trace at 98 °C, is absent in the subsequent
run, showing that the metal has reacted with the silica gel. The
highest concentration at which all of the Na reacts is ∼40 wt %
Na. Stage II Na-SG shows no peaks in the DSC up to 450 °C.
Stage II powders of Na-SG are easily handled in an open
ambient environment and do not change with time when kept in a
closed container. They are slightly less reducing than stages 0 and
I but react rapidly with water to yield the same amounts of pure
hydrogen gas, as measured by a method described elsewhere.4b
Because of its significantly lower reactivity toward air, even in the
presence of mild humidity, stage II Na-SG reduced silica would
be the material of choice for hydrogen production, solvent drying,
and those reduction reactions that do not require the stronger
reducing properties of stages 0 or I.
Preliminary atomic pair distribution function measurements
similar to those applied to “inorganic electrides”4c suggest that stage
I samples retain mostly SiO2 character and that stage II Na-SG
contains nanoparticles of Na4Si4.6 The fate of the released electrons
in stage I M-SG remains a mystery. Are they delocalized in a
high energy band of silica, or have they partially reduced the silica,
or are they present in the void spaces along with partially ionized
cations, as with “inorganic electrides”?
Stage 0 samples, when heated to about 150 °C, form stage I
powders that show little or no melting endotherm by DSC (Figure
2B). Exposure to dry oxygen for up to a day does not change the
reducing capacity (the amount of hydrogen produced by reaction
with water). Stage I samples are stable in closed columns and have
† Michigan State University.
‡ SiGNa Chemistry, LLC.
9
9338
J. AM. CHEM. SOC. 2005, 127, 9338-9339
10.1021/ja051786+ CCC: $30.25 © 2005 American Chemical Society

C O M M U N I C A T I O N S
Figure 2. (A) DSC traces for 9.2 mg of stage 0 Na2K in SG. The inset shows the region of melting. The overall value of ∆H is ∼-100 kJ/mol of metal.
(B) DSC traces for 12.5 mg of stage I NaK2 in SG. Note the absence of a melting endotherm. The overall value of ∆H is ∼-35 kJ/mol of metal. (C) DSC
traces for a mixture of 4.6 mg of Na and 6.6 mg of SG, which react exothermically to form stage II material. The endothermic heat of melting of Na (113
J/g Na), which appears at 98 °C in the initial scan, is absent in the repeat scan.
Organic reductions in tetrahydrofuran (THF) were carried out
at room temperature in a nitrogen-filled glovebag, either by stirring
batch reactions with M-SG or by passing the solution through a
2-mL Pasteur pipet that had been plugged with Pyrex glass wool
and filled with M-SG. Analyses before and after reaction were
stage 0 material. Surprisingly, the historically difficult9 reduction
of diphenyl sulfide to biphenyl was also accomplished in 2 h via
bulk reaction with stage I Na2K-SG.
The powerful reducing properties of alkali metal-silica gel
adducts should permit the use of continuous-flow columns for the
reduction of organic and inorganic compounds that are now reduced
by alkali metals. In addition, they provide the ability to provide
hydrogen gas upon demand from air-stable, easily handled sources
that can be stored indefinitely.
1
made by H and 13C NMR and GC-MS. The effective reduction
potentials of the various stages of M-SG can be determined by
comparing their ability to form highly colored aromatic radical
anions with the reduction potentials listed by Szwarc7 (biphenyl )
0). Stage 0 M-SG can reduce biphenyl to its radical anion. Stage
I powders reduce naphthalene (0.043 V) and phenanthrene (0.142
V). Even stage II materials, which can be handled in air, are able
to reduce pyrene (0.529 V) and anthracene (0.642 V) to their radical
anions. M-SG materials can serve as effective drying agents and
anti-oxidants for aprotic solvents such as THF. Care must be taken
not to expose flammable solvents that contain particles of M-SG
to air, since spontaneous ignition could occur.
Acknowledgment. We are grateful to Dr. Rui H. Huang for
assistance with GC-MS, Prof. Simon Billinge for PDF studies,
and the Camille and Henry Dreyfus Foundation for partial support
of undergraduates S.A.U. and K.D.C.
Supporting Information Available: Materials and methods, reac-
tion schemes, and product identification. This material is available free
of charge via the Internet at http://pubs.acs.org.
Anthracene in THF underwent Birch reduction1a (5 min elution
time) to 9,10-dihydroanthracene in >99% purity upon elution
through a chromatography column that contained either stage 0 or
stage I M-SG mixed with an equal volume of silica gel (to provide
a proton source). In the absence of a proton source, all three stages
yielded stable THF solutions that contained anthracene radical
anions, as verified by the optical spectrum (Supporting Information).
Reductions such as the Wurtz reduction of halogenated organic
compounds1b,c can be violent with alkali metals but occur smoothly
with M-SG. Benzyl chloride reacted by Wurtz coupling in both
chromatographic (stages 0 and I) and batch (stage II) processes to
form bibenzyl as the only product. Other dehalogenations include
the dechlorination of chlorobenzene, 1,2-dichlorobenzene, and
neopentyl chloride. Thus, both aromatic and aliphatic halocarbons
can be dehalogenated by M-SG.
References
(1) (a) Savoia, D.; Trombini, C.; Umani-Ronchi, A. Pure Appl. Chem. 1985,
57, 1887-1896. (b) Levy, J.; Tamarkin, D.; Selig, H.; Rabinovitz, M.
Angew. Chem., Int. Ed. Engl. 1981, 20, 1033. (c) Rabideau, P. W.
Tetrahedron 1989, 45, 1579-1603.
(2) (a) Birch, A. J. Pure Appl. Chem. 1999, 68, 553-556. (b) Billington, D.
C. In ComprehensiVe Organic Synthesis; Trost, B. M., Ed.; Pergamon
Press: New York, 1991; Vol. 3, pp 413-423.
(3) (a) Edwards, P. P.; Anderson, P. A.; Thomas, J. M. Acc. Chem. Res. 1996,
29, 23-29. (b) Rabo, J. A.; Angell, P. H.; Kasai, P. H.; Schomaker, V.
Discuss. Faraday Soc. 1966, 41, 328-349. (c) Srdanov, V. I.; Stucky, G.
D.; Lippma, E.; Engelhardt, G. Phys. ReV. Lett. 1998, 80, 2449-2452.
(4) (a) Ichimura, A. S.; Dye, J. L.; Camblor, M. A.; Villaescusa, L. A. J. Am.
Chem. Soc. 2002, 124, 1170-1171. (b) Wernette, D. P.; Ichimura, A. S.;
Urbin, S. A.; Dye, J. L. Chem. Mater. 2003, 15, 1441-1448. (c) Petkov,
V.; Billinge, S. J. L.; Vogt, T.; Ichimura, A. S.; Dye, J. L. Phys. ReV.
Lett. 2002, 89, 075502.
(5) Unruh, K. M.; Huber, T. E.; Huber, C. A. Phys. ReV. B 1993, 48, 9021-
9027.
(6) (a) Witte, J.; Schnering, H. G. Z. Anorg. Allg. Chem. 1964, 327, 260-
273. (b) He, J.; Klug, D. D.; Uehara, K.; Preston, K. F.; Ratcliffe, C. I.;
Tse, J. S. J. Phys. Chem. B 2001, 105, 3475-3485.
Slow alkali metal reduction of aromatic sulfur compounds in
hydrocarbon solvents can occur at high temperatures.8 With M-SG
the reduction of dibenzothiophene in THF to biphenyl by continuous
(stage 0) and batch reactions (0 and I) is readily accomplished at
room temperature. Even the normally difficult desulfurization of
4,6-dimethyl dibenzothiophene occurred in a batch process with
(7) Szwarc, M. Carbanions, LiVing Polymers and Electron-Transfer Processes;
Interscience: New York, 1968; p 314.
(8) Yu, Z.; Verkade, J. G. Energy Fuels 1999, 13, 23-28.
(9) Hauptman, H.; Walter, W. F.; Marino, C. J. Am. Chem. Soc. 1958, 80,
5832-5835.
JA051786+
9
J. AM. CHEM. SOC. VOL. 127, NO. 26, 2005 9339