Note: Descriptions are shown in the official language in which they were submitted.
CA 02202446 1997-04-11
METHOD OF DEPOSITION OF A METAL ON A METAL SURFACE AND THE
PRODUCT THEREOF
FIELD OF INVENTION
s
The present invention relates to a method, and the product thereof, of
reducing one or more metal centres of one or more organometallic precursors in the
presence of a metal surface to deposit sub-monolayer, monolayer or multilayer
quantities of the metal centres thereon. Further, the invention relates to a method,
0 and the product thereof, of hydrogenating an organometallic precursor over a metal
surface to deposit sub-monolayer, monolayer or multilayer quantities of the metal
centre of the precursor thereon. Preferably, the method permits real time control over
surface stoichiometry of the evolving surface resulting from the reaction. Finally, the
present invention relates to a method of manufacturing electrodes having bimetallic
or polymetallic clusters or surfaces for use in methanol fuel cells.
BACKGROUND OF INVENTION
Bimetallic clusters of ruthenium (Ru) and platinum (Pt) have been used
20 to catalyze the hydrogenation of arenes, the hydrogenolysis of propane, the
hydrogenation of carbon monoxide and the electrochemical oxidation of methanol in
fuel cells. Ru-Pt metallic clusters are typically prepared by chemical or electrochemical
reduction of Ru and Pt compounds (e. g. chlorides, oxides, hydroxides and amines) in
the presence of a support. However, the surface compositions and oxidation states of
2s the resulting bimetallic clusters tend to be ambiguous. Sources of ambiguity include
variations in the rates of mass transport to the sites of cluster growth, surface
segregation during high-temperature stages of the synthesis (e. g. during reduction of
precursors and during electrode pressing), interactions with the support, and
dissolution of Ru during pretreatment by potential cycling.
A further method now used to synthesize bimetallic clusters that may
provide some control over surface composition involves a two step process in which a
hydrocarbonyl compound of one of a first metal is deposited on a support containing
metal clusters of a second metal. The hydrocarbonyl compound is then reduced by
3s dihydrogen gas at elevated temperatures. The hydrocarbonyl compound of the first
metal is either physisorbed on the support and migrates to the metal clusters of the
second metal during the reduction by the dihydrogen, or the hydrocarbonyl is grafted
--1-
CA 02202446 1997-04-11
onto the surface of the second metal before reduction by dihydrogen. Control over the
surface composition of these bimetallic clusters is achieved by adjusting the ratio of the
amount of hydrocarbonyl originally introduced to the system to the number of active
sites on the clusters of the second metal (i. e. limiting-reagent control).
As well, it has been reported that (diolefin)dialkylPt(II) complexes have
been reduced by dihydrogen gas over Pt black to generate the corresponding alkanes
and Pt(0) that was then incorporated onto the surface of Pt black.
0 Finally, it has been found that the use of a single metal catalyst electrode,
such as a platinum electrode, in methanol fuel cells is relatively ineffective.
Specifically, the platinum appears to be poisoned by carbon monoxide or a related
species during the reaction and the current produced by the fuel cell quickly declines.
There is therefore a need in many industries for a method of depositing
sub-monolayer, monolayer or multilayer quantities of one metal upon another in amanner which preferably permits real time control over the surface stoichiometry.
Further, there is a need in industry for a method of manufacturing a catalytic electrode
for use in fuel cells, and in particular methanol fuel cells, which electrode is relatively
effective as compared to known fuel cell electrodes.
SUMMARY OF INVENTION
The present invention relates to a method of depositing sub-monolayer,
2s monolayer or multilayer quantities of one or more metals upon another metal, and
also relates to the product thereof. The present invention provides a relatively clean
deposition reaction that allows for the monitoring of the extent of deposition of one
metal upon another in situ, and that permits interruption when the desired surface
stoichiometry is achieved by either reaction-rate control or mass transport control. In
this manner, the method may permit real time control over the surface stoichiometry
of the resulting reaction surface.
Further, the invention relates to a method, and the product thereof, of
reducing one or more metal centres of one or more organometallic precursors in the
presence of a metal surface to deposit sub-monolayer, monolayer or multilayer
quantities of the metal centres thereon. Again, the method may permit real time
control over the surface stoichiometry of the evolving surface resulting from the
--2-
CA 02202446 1997-04-11
reaction. The reducing step may be comprised of reacting the metal centres of the
organometallic precursor with a reducing agent in the presence of the metal surface. In
addition, the method may be comprised of dissolving the organometallic precursor in a
solvent and reacting the solution with the reducing agent in the presence of the metal
s surface.
As well, the reducing step may be comprised of hydrogenating the
organometallic precursors in the presence of the metal surface to deposit sub-
monolayer, monolayer or multilayer quantities of the metal centres of the precursor
0 thereon. The method may also be comprised of dissolving the organometallic
precursor in a solvent and hydrogenating the solution in the presence of the metal
surface. In these instances, the reducing agent may be comprised of dihydrogen gas or
some other source of hydrogen atoms that hydrogenates the organometallic precursor.
In the preferred embodiment, the invention relates to the controlled
deposit of submonolayer, monolayer or multilayer quantities of ruthenium (Ru)
adatoms on platinum (Pt) metal using a chemical reaction between a dissolved Ru
hydrocarbonyl compound and the Pt surface. Preferably, the method is comprised of
the hydrogenation of Ru(COD)( 113-C3Hs)2 (preferably COD is 1,5-cyclooctadiene) over
20 black platinum by dihydrogen gas in hexanes.
The hydrogenation may be carried out at any temperature. However,
preferably the hydrogenation is conducted at a temperature of between about 20 degrees
Celsius and -40 degrees Celsius. In the preferred embodiment, the hydrogenation is
2s carried out at a temperature of between about -10 and -15 degrees Celsius.
The hydrogenation preferably results in adsorption of Ru adatoms by the
surface of Pt with concomitant formation of reaction products. Typically, the total
amount of the reaction products in solution is approximately equal to the amount of
30 the metal precursor, being Compound 1 in the preferred embodiment, consumed at all
stages of the hydrogenation. Accordingly, the method permits the observation in real
time of both the stoichiometry and the activity of the evolving Ru-Pt surface bymonitoring the concentrations of either the metal precursor or of the reaction products
in solution.
The present invention also relates to a method of manufacturing a
catalytic electrode for use in fuel cells, and in particular methanol fuel cells, which
-3-
CA 02202446 1997-04-11
electrode is relatively effective as compared to known fuel cell electrodes. Preferably,
the present invention relates to a method of manufacturing electrodes, having
bimetallic clusters or surfaces, for use in methanol fuel cells.
During experiments conducted using the within method, a catalyst surface
resulting from hydrogenation of 0.11 equiv Ru has been found to be up to ~ 14 times
more active than bare Pt for the potentiodynamic oxidation of methanol ([MeOH] = 1.0
M, [H2SO4] = 0.5 M, 40 degrees Celsius, sweep rate = 5 mV/sec), a catalyst surface
resulting from hydrogenation of 0.33 equiv Ru has been found to oxidize methanol0 potentiostatically at 0.158 V ([MeOH] = 0.5 M, [H2SO4] = 0.5 M, 25 degrees Celsius) for 45
min. with ~ 13 times the activity of Pt under the same conditions, and a catalyst surface
resulting from deposition of 0.8 equiv Ru has been found to oxidize methanol
potentiostatically at 0.256 V under the above conditions for a total of 1.5 h with
negligible dissolution of Ru into the electrolyte.
SUMMARY OF DRAWINGS
Embodiments of the invention will now be described with reference to
the accompanying drawings, in which:
Figure 1 shows plots of equiv Ru adatoms deposited on Pt by
hydrogenation of Compound 1 versus time for reactions carried out at 20 ~C and at -10
~C. The equiv Ru were determined from the total amount of C8 product hydrocarbons
in solution;
Figure 2 shows plots of equiv Ru consumed and total equiv C8
hydrocarbon products produced in solution versus time for a hydrogenation of
Compound 1 carried out at -10 ~C.;
Figure 3 (a) shows cyclic voltammograms (sweep rate = 5 mV-(sec)~1 in 1.0
M H2SO4) of a catalyst surface resulting from hydrogenation of 2.7 equiv Ru recorded
before and after adsorption of a monolayer of carbon monoxide at -0.19 V. in which the
voltammograms are not normalized for surface area;
3s Figure 3 (b) shows cyclic voltammograms of a black Ru electrode recorded
under identical conditions as those for Figure 3 (a) in which the voltammograms are
not normalized for surface area;
-4-
CA 02202446 1997-04-11
Figure 4 shows cyclic voltammograms (sweep rate = 5 mV-(sec)~1, [H2SO4]
= 0.5 M, [MeOH] = 0.5 M, 25 ~C) of a catalyst surface generated by hydrogenation of 3.5
equiv Ru;
Figure 5 shows a potentiodynamic oxidation of methanol ([MeOH] = 1.0
M, [H2SO4] = 0.5 M, 40 ~C, sweep rate = 5 mV-(sec~1)) by black Pt and by a Ru-Pt catalyst
surface resulting from hydrogenation of 0.11 equiv Ru, in which the currents arenormalized to 4.1 ,umol surface atoms; and
Figure 6 shows a potentiostatic oxidation of methanol ([MeOH] = 0.5 M,
[H2SO4] = 0.5 M, 25 ~C) by black Pt and by a Ru-Pt catalyst surface resulting from
hydrogenation of 0.33 equiv Ru, in which the currents are normalized to 4.1 ,umol
surface atoms.
DETAILED DESCRIPTION
The present invention relates to a method, and the product thereof, of
reducing one or more metal centres of one or more organometallic precursors in the
20 presence of a metal surface to deposit sub-monolayer, monolayer or multilayerquantities of the metal centres thereon. In the preferred embodiment, the invention
relates to a method, and the product thereof, of hydrogenating the organometallic
precursor in the presence of the metal surface to deposit sub-monolayer or multilayer
quantities of the metal centre of the precursor thereon. The method permits real time
2s control over surface stoichiometry and the evolving bimetallic or polymetallic surface.
Thus, thin films, sub-monolayers, monolayers or multilayers of the metal centre of the
precursor may be applied to the metal surface.
The metal surface may be comprised of any metal or metals that effects or
30 catalyzes the reduction of the metal centre or centres of the selected organometallic
precursor. More particularly, in the preferred embodiment, the metal surface may be
comprised of any metal or metals that effect or catalyze the hydrogenation of the
selected organometallic precursor. Most preferably, the metal surface is capable of
effecting the hydrogenation of olefins.
The metal surface may be comprised of one or more metals from groups 8,
9 or 10 of the periodic table of elements. Group 8 metals include iron, ruthenium and
-5-
CA 02202446 1997-04-11
osmium. Group 9 metals include cobalt, rhodium and iridium. Group 10 metals
include nickel, palladium and platinum. As well, the metal surface may be comprised
of one or more metals from group 11 or 12 of the periodic table. Group 11 metalsinclude copper, silver and gold. Group 12 metals include zinc, cadmium and mercury.
5 Further, the metal surface may be comprised of one or more metals containing
interstitial hydrides or one or more metals that activate hydrogen, co-ordinate olefins
and undergo migratory insertion reactions.
In addition, the metal surface may be comprised of small highly dispersed
powders, smooth surfaces or solid surfaces with varying degrees of roughness.
However, as the reaction tends to occur on the surface of the metal only, the metal
surface is preferably comprised of metal clusters in order to maximize the available
surface area for the reaction to occur. As well, if desired, the metal surface may be
supported on an inert substrate. In this case, substrate supports free of oxygen or oxide
films, such as carbon, are most appropriate.
As indicated, in the preferred embodiment, the metal surface acts to
catalyze the hydrogenation of the organometallic precursor and the metal centre or
centres of the precursor are reduced and adsorbed by the metal surface. The adsorption
20 preferably results in the metal centre and the metal surface physically being on the
same surface or being no greater than one atom apart. The composition of the metal
surface is constantly changing during the reaction until the desired coverage isobtained. By monitoring the extent of the deposition on the metal surface during the
reaction, the reaction may be interrupted when the desired surface stoichiometry is
25 achieved. Thus reaction-rate control may be achieved.
To perform the within method, the initial surface area of the metal
surface in atomic terms is preferably known. However, under some conditions, theinitial surface area may be inferred from the variation of the reaction rate with surface
30 composition.
In the preferred embodiment, the metal surface is comprised of platinum
(Pt). More particularly, a shiny 52 mesh blacked Pt gauze, threaded with Pt wire leads,
may be used. Its surface area may be determined from the coulombic charge in the35 cathodic hydride region of cyclic voltammograms recorded in a 1.0 M aqueous solution
H2SO4 under argon pursuant to conventional procedures.
CA 02202446 1997-04-11
Once the surface area is determined, the metal surface is transferred to the
reactor or reaction apparatus in which the reduction or hydrogenation is to take place
by any suitable means or method. However, in the preferred embodiment, when the
surface area of the black Pt is determined in an aqueous acid electrolyte solution, the
5 following procedure is used to transfer the gauze from the aqueous acid electrolyte to
the reactor. First, the surface of the gauze is protected as oxides by holding the potential
at 1.2 V (all potentials in this application are reported relative to the saturated calomel
electrode (SCE)) for 2 min.). Second, the gauze is raised above the electrolyte and rinsed
with purified water under argon. Third, while protected by drops of purified water, the
0 gauze is quickly transferred through air to the reactor. Fourth, the gauze is dried under
a stream of argon and then placed under an atmosphere of dihydrogen gas to reduce
the surface oxides to hydrides.
The reactor may be comprised of apparatus suitable for, and compatible
with, its intended purpose as described herein. Preferably, the reactor includes any
suitable device, mechanism or structure for stirring the contents of the reactor and any
suitable device, mechanism or structure for supporting the metal surface in the reactor.
In the preferred embodiment, the reactor is comprised of a Pyrex (trade-mark) tube
containing a Teflon (trade-mark) coated magnetic stir bar and fitted with a rubber septa
20 pierced with a pipette to act as a reducing agent inlet. Further, the reactor may be
comprised of a glass tube for supporting the metal surface in the reactor.
A measured or quantified amount of a solution of the organometallic
metal precursor dissolved in a solvent is either transferred to the reactor or prepared
2s within the reactor. Once the solution is present in the reactor, the solution is
preferably rapidly stirred. In addition, a relatively slow continuous stream of a
reducing agent is introduced to the solution, which is maintained throughout thereaction. The metal surface may be introduced to the solvent, the organometallicprecursor or the solution at any time during the method, however, the metal surface is
30 preferably maintained out of contact with the solution until the reducing agent is
introduced to the solution.
As stated, the method is comprised of reducing the metal centre or centres
of the organometallic precursor in the presence of a metal surface. In particular, the
3s metal centres are exposed to a reducing agent in the presence of the metal surface. In
the preferred embodiment, the method is comprised of the hydrogenation of the
organometallic precursor in the presence of the metal surface and the reducing agent is
-7-
CA 02202446 1997-04-11
comprised of dihydrogen gas. The dihydrogen gas may be at any pressure, however
preferably the dihydrogen gas is at a pressure of approximately one atmosphere. The
solvent is saturated with the dihydrogen gas. The reducing agent is preferably present
in excess to ensure reaction rate limited kinetics (RRL) and thus ensure that mass
s transfer of the reducing agent to the metal surface is not rate limiting. However, the
reaction will still occur and may also be desirable under conditions where mass
transport of dihydrogen to the surface is rate limiting.
The reducing agent may also be comprised of compounds other than
o dihydrogen gas, although the reducing agent is preferably comprised of other sources of
hydrogen atoms. Preferably, the reducing agent does not fully react with the
organometallic compound in the absence of the metal surface. Sources of hydrogenatoms are preferred as the reducing agent since the hydrogen saturates the ligands of
the organometallic precursor, rendering the resulting reaction products less reactive.
15 The reaction may also be carried out with metal surfaces initially saturated without
having excess dihydrogen present, which may allow for control of the surface
composition by variation of the number of hydrides required to reduce the metal
precursor.
As indicated, the solution of the organometallic precursor may be
comprised of the organometallic precursor dissolved in a solvent. The solution may
also include an internal standard such as decane. Preferably, the metal precursor is
comprised of an organometallic compound that does not fully react with the reducing
agent in the absence of the selected metal surface. The metal precursor may partially
2s react with the reducing agent, being dihydrogen, to produce hydride, dihydrogen, or
hydrogen insertion reactions prior to reaction with the metal surface. The metalprecursor may contain any ligands that can be reduced to saturated hydrocarbons by
dihydrogen.
The metal precursor is preferably an organometallic precursor, which is
preferably comprised of metal centres connected to hydrocarbons. The hydrocarbons
are preferably comprised of alkanes. As well, the hydrocarbons preferably include no
impurities, although a level of impurities not interfering with the function of the
metal precursor may be present.
3s
CA 02202446 1997-04-11
The metal atoms or centres of the precursor must be capable of being
adsorbed onto the metal surface by reduction with the reducing agent, which in the
preferred embodiment is dihydrogen gas.
s The metal centre of the organometallic precursor may be comprised of
early transition metals (groups 4, 5, 6, and 7 of the periodic table), late transition metals
(groups 8, 9,10,11 or 12 of the periodic table of elements as set out above) or the main
group metals. Group 4 metals include titanium, zirconium and hafnium. Group 5
metals include vanadium, niobium and tantalum. Group 6 metals include chromium,
0 molybdenum and tungsten. Group 7 metals include manganese, technetium and
rhenium. Main group metals include aluminum, gallium, indium, thallium,
germanium, tin, lead, antimony, bismuth and polonium. Not all of these metals are
capable of deposition on the metal surface to the same extent or at the same reaction
rate. The metal centre of the organometallic precursor should therefore be selected so
as to achieve the desired deposition and reaction rate. The metal centres of theorganometallic precursor may be in any oxidation state.
Further, combinations of metal precursors containing different metal
centres may be used. The composition of the resulting surface will depend on therelative reactivities and concentrations of the different precursors in solution.
Homopolymetallic and heteropolymetallic precursors may be used. Oligomeric and
polymeric metal precursors may also be used.
In the preferred embodiment, the metal centre or centres are ruthenium
2s and the metal precursor is comprised of a ruthenium (Ru) hydrocarbonyl compound.
In particular, the metal precursor is preferably comprised of Ru(COD)( 11 3-c3H5)2 (in
Compound 1, COD is 1,5-cyclooctadiene). Compound 1 may be prepared
conventionally and is preferably sublimed immediately before use. This Compound 1
has attractive features in that it can be prepared in pure form and it contains no
components (i.e. halide ions, phosphines, carbonyls, heteroatoms) that might act as
catalyst poisons.
The solvent is comprised of a compound which is capable of dissolving
the organometallic compound and which does not completely inhibit the deposition of
the metal centre on the metal surface. Further, the solvent must interact with the
metal surface to provide an interface that is suitable to allow the reduction of the metal
centre to occur and the reducing agent must be miscible in the solvent. The solvent is
g
CA 02202446 1997-04-11
preferably volatile, which may facilitate its removal from the reactor and the metal
surface. In the preferred embodiment, the solvent is comprised of an alkane, preferably
being dihydrogen-purged hexanes. However, other solvents such as ethers, alcohols
and per-fluorinated solvents may be used.
As well, the solvent is preferably a non-polar, saturated hydrocarbon.
However, polar solvents may be used provided they do not completely inhibit the
metal precursor, the metal surface or the reducing agent.
o In the preferred embodiment, the reaction is carried out as a mixture of
the solvent, containing reducing agent, and the organometallic precursor in
combination with the solid metal surface. However, a gas/solid reaction system where
the organometallic precursor and the reducing agent are gases reacting with the solid
metal surface may also be used.
Thus, in the preferred embodiment, the dihydrogen-purged hexanes are
transferred to the reactor containing the Pt metal surface. A slow stream of dihydrogen
gas is then passed through the solution at approximately one atmosphere throughout
the reaction. Freshly sublimed Compound 1 is weighed and dissolved in the hexanes,
preferably along with a weighed amount of decane, to produce the reaction mixture.
The reaction mixture is then mixed or stirred and the Pt metal surface is immersed
into the reaction mixture.
In the preferred embodiment, prior to the immersion of the metal surface
2s in the reaction mixture, the reactor is immersed in a cooling bath. Although the
reaction will occur at virtually any temperature, preferably the reaction mixture is
cooled to a sufficient temperature to slow the reaction, when the metal surface is
immersed in the reaction mixture, so that the evolving organometallic precursor-metal (Ru-Pt) surface may be easily and effectively monitored to observe the
stoichiometry and activity of the evolving surface. However, too low a temperature
will render the method no longer feasible or practical given the slowness of thereaction. It has been found that at a temperature of about 20 degrees Celsius, the
reaction occurs very quickly and interruption of the reaction at the desired surface
composition is difficult to achieve. Conversely, cooling to a temperature of about -40
3s degrees Celsius slows the reaction to a level such that the use of the method may no
longer be feasible. Thus, a temperature of between about 20 degrees Celsius and -40
degrees Celsius is preferred. More preferably, it has been found that the evolving
-10-
CA 02202446 1997-04-11
surface may be easily and effectively monitored when the reaction mixture is cooled to
between about -10 and -15 degrees Celsius.
The hydrogenation of the metal precursor results in the adsorption of
s adatoms of the reduced metal centres of the metal precursor by the metal surface, with
concomitant formation of reaction products. Typically, the total amount of the
reaction products is equal to the amount of the metal precursor consumed at all stages
of the hydrogenation, unless there has been some unaccounted for decomposition of
Compound 1. In this manner, the stoichiometry and the activity of the evolving
surface may be observed in situ in real time by monitoring either the concentration of
the metal precursor or the concentration of the reaction products in the reaction
mixture.
In the preferred embodiment, the hydrogenation results in adsorption of
Ru adatoms by the Pt surface with concomitant formation of propane, cyclooctane, and
small amounts of cis-bicyclo[3.3.0]octane and n-octane. The total amount of
cyclooctane, bicyclo[3.3.0]octane, and n-octane in solution is approximately equal to the
amount of Compound 1 consumed at all stages of the hydrogenation. Therefore, theconcentrations of either Compound 1 or of the C8 product hydrocarbons in solution
20 may be monitored.
To monitor the concentrations, aliquots are removed at timed intervals
and analyzed by conventional means, such as either by gas chromatography, by mass
spectrometry or by the UV-VIS absorbance of the solution. Gas chromatography
25 permits the concentration of the reaction products to be determined, while UV-VIS
absorbance permits the decrease in the concentration of the metal precursor to be
determined.
It has been found that the nature of the reaction products (as determined
30 by gas chromatography-mass spectrometry) from the hydrogenation of the metal
precursor may depend upon the reaction temperature. For instance, in the preferred
embodiment, propane, cyclooctane (89 % of C8 products), cis-bicyclo[3.3.0]octane (8 % of
C8 products), and n-octane (3 % of C8 products) tend to be produced at room
temperature. Propane, cyclooctane (95 % of C8 products), and cis-bicyclo[3.3.0]octane (5
35 % of C8 products) tend to be produced at -10 degrees Celsius. The ratio of C8 products
does not tend to change within experimental error over the course of the
hydrogenation.
-11-
CA 02202446 1997-04-11
Once the desired coverage of the metal surface is achieved for the
particular purpose or use of the metal surface, the hydrogenation may be interrupted.
The hydrogenation may be interrupted in any suitable manner or by any suitable
5 means. However, in the preferred embodiment, the hydrogenation is interrupted by
quickly raising the metal surface above the reaction mixture and removing the reaction
mixture from the reactor. The metal surface may then be rinsed, preferably with cooled
hexanes, and dried under a stream of dihydrogen gas.
0 As stated, the desired deposited amount of equivalent ruthenium on the
metal surface will vary depending upon the particular purpose of the method and
intended use or application of the metal surface. In the preferred embodiment, where
the metal surface is intended for use as a catalytic electrode in a methanol fuel cell, a
ratio of the metal centre of the organometallic precursor to the metal surface of
S between about 9:1 to 7:3 is preferred (i.e. about 0.11 to 0.43 equiv Ru). It has been found
that for this particular application, however, a deposition of about 0.11 equivalent
ruthenium is most preferred. For other applications, any ratio which produces a metal
surface compatible with its intended may be used.
In the preferred embodiment, once the hydrogenation is interrupted, as
described above, the resulting metal surface may be transferred to the electrochemical
cell by any suitable means or method, such as through an atmosphere of argon in a
glovebox.
2s It has been found that the within method is a relatively low energy
process, as compared to known conventional processes, for the deposition of adatoms
of a foreign metal on the surface of another metal. It allows for the generation of a
prototypic kinetic bimetallic or polymetallic surface with real time control over the
stoichiometry and activity of the evolving surface. This system may offer certain
advantages over conventional methods of deposition of metal adatoms (e. g. metalatom evaporation and chemical vapor deposition). For instance, it may: provide
uniform coverage over all exposed sides of a rough, blacked metal surface after
adsorption of less than two equiv adatoms; allow for deposition of sub-monolayerequivalents of adatoms under reaction-rate control at low temperatures; and allow for
3s use of simple, bench-top equipment and techniques. Further, it preferably uses a Ru
precursor (Compound 1) containing no components that may poison the resulting
catalyst surface.
-12-
CA 02202446 1997-04-11
As well, in the preferred embodiment, the resulting Ru-Pt surfaces have
shown appreciable activities and durability towards both the potentiodynamic and the
potentiostatic oxidation of methanol under conditions that are typical for an operating
5 fuel cell. Since this deposition proceeded via a reaction with a metal surface, it will in
principle allow for self-directed depositions on Pt clusters dispersed on an inert
support.
Finally, this methodology may be applied to a number of metal systems
provided the appropriate precursors are employed. Sub-monolayer coverages of onemetal upon another may be readily prepared. The resulting bimetallic surfaces may be
useful heterogeneous catalysts. For instance, platinum/ruthenium bimetallic surfaces
may be useful for the oxidation of small organic molecules containing carbon,
hydrogen and oxygen. Increased activity for the anodic oxidation of methanol, ethanol,
l5 ethylene glycol and glucose relative to pure Pt electrodes has been found. The
composition of these surfaces may be systematically optimized for the fuel in question,
or for the desired reaction.
More elaborate surfaces, composed of several different metals may be
20 prepared using appropriate combinations of support(s), surface(s), and precursor(s). As
well, single monolayer coverage's and multi-layer coverage's may be prepared. Thin
films of different metals may be deposited to give layered structures of uniform atomic
thickness. Layered structures with novel magnetic (magnetization) and/or conductive
properties (2D conductors) may be prepared. Such structures may be useful for the
2s preparation of magnetic storage devices or electronic applications (integrated circuits,
or circuit boards).
Since this method is reaction rate limiting, the metal atom of the
precursor may be deposited in thin uniform layers onto surfaces of varying
30 morphology. This method may therefore be used to cover a less noble metal with a
protective layer of a noble metal. Thus the reaction could be used to apply a protective
layer to machined metal, and also to biomedical applications (a thin protective coat to
avoid allergic reactions). Also heterogeneous catalysts may be prepared using a cheap
metal as the base thus minimizing the amount of expensive noble metal required. For
35 example a noble metal (Pt) may be deposited onto a cheap base metal (Ni) and the thin
noble metal layer may be further reacted with the appropriate metal precursor to form
a more elaborate surface composition. The above procedure would also work with
-13-
CA 02202446 1997-04-11
supported metals. In addition polymetallic surfaces may be prepared by reversing the
combination of metal precursor and metal surface.
As the reaction conditions are mild (non-corrosive solvent, low
s temperature), fragile metal surfaces that would be adversely affected by high
temperature, aqueous acid or alkaline environments may be used in the method.
By varying the reaction conditions from reaction rate limited (RRL) to
mass transport limited (MTL) both the uniformity of the deposit and the morphology
0 (roughness) of the deposit may be controlled. RRL conditions will favor uniform
coverage whereas MTL conditions will favor non-uniform coverage and an increase in
the surface roughness (porosity).
EXAMPLES
The following examples serve more fully to illustrate the invention. The
major objective of these examples was to deposit controlled submonolayer or
multilayer quantities of Ru adatoms on Pt metal using a chemical reaction between a
dissolved Ru hydrocarbonyl compound and the Pt surface. In particular, this example
20 describes the hydrogenation of (COD)Ru( 113-C3Hs)2 (Compound 1, COD is 1,5-
cyclooctadiene) by dihydrogen gas in hexane solution over black Pt.
It has been found that the hydrogenation, as described above, resulted in
adsorption of Ru adatoms by the surface of Pt in accordance with the following
25 equation:
~ Ru + 5H2 blaCknP~l Ru(O) + ~+ O + t~
+
3s
-14-
CA 02202446 1997-04-11
Results and Discussion
General Methods
Compound 1 was prepared by known conventional processes and
sublimed immediately before use. This compound is preferable given that it can be
prepared in pure form and that it contains no components (halide ions, phosphines,
carbonyls, heteroatoms) that might act as catalyst poisons. A blacked Pt gauze was used
as the substrate. The surface area of the black Pt was determined from the coulombic
o charge in the cathodic hydride region of cyclic voltammograms recorded in a 1.0 M
aqueous solution H2SO4 under argon.
The following procedure was then used to transfer the gauze from the
aqueous acid electrolyte to the vessel in which the hydrogenation was carried out.
First, the surface of the gauze was protected as oxides by holding the potential at 1.2 V
(all potentials herein are reported relative to the saturated calomel electrode (SCE)) for
2 min.). Second, the gauze was raised above the electrolyte and rinsed with purified
water under argon. Third, while protected by drops of purified water, the gauze was
quickly transferred through air to the hydrogenation vessel. Fourth, the gauze was
dried under a stream of argon and then placed under an atmosphere of dihydrogen gas
to reduce the surface oxides to hydrides.
The gauze was then immersed in dihydrogen-saturated hexanes at the
desired reaction temperature. The reaction mixture was rapidly stirred (800 rpm) and a
continuous stream of dihydrogen gas (~ 10 mL/min.) was bubbled through the solution
during the hydrogenation to ensure that mass transfer of dihydrogen to the surface of
the gauze was not rate limiting (reaction-rate limiting conditions rather than mass-
transport limiting conditions).
To determine if significant changes in surface area occurred during these
manipulations, a control experiment was performed in which the above procedure was
repeated in the absence of Compound 1. After being immersed in a blank hexanes
solution under dihydrogen gas for a typical time of reaction, the Pt gauze was lifted
above the hexanes under dihydrogen gas in the hydrogenation vessel, the hexanes
were removed by cannula from the vessel, and the gauze was dried under a stream of
dihydrogen gas. Protected under several drops of purified water, the gauze was quickly
transferred through air to an electrochemical cell containing a 1.0 M solution of H2SO4
-15-
CA 02202446 1997-04-11
in water under argon, and the surface area was determined again using cyclic
voltammetry. Little change (a decrease by ~ 6 %) in the surface area of the black Pt had
occurred during these manipulations.
s Hydrogenation of Compound 1
To begin a hydrogenation, a solution of Compound 1 and a decane
internal standard dissolved in hexanes were quickly and quantitatively transferred to
the reactor under dihydrogen. Aliquots were removed at timed intervals. The aliquots
0 were analyzed either by gas chromatography (GC) and the concentrations of product
hydrocarbons were determined, or by the UV-VIS absorbance of the solution and the
decrease in concentration of Compound 1 was determined. Since the products of the
hydrogenation were alkanes and Ru(0) adsorbed on the surface of Pt, the only
component of the hydrogenation mixture with a significant UV-VIS absorbance was
lS Compound 1. Control experiments showed that reaction between Compound 1 and
dihydrogen gas did not occur under these conditions in the absence of Pt.
It was found that the nature of the hydrocarbon products (as determined
by gas chromatography-mass spectrometry) from the hydrogenation of Compound 1
20 was dependent upon the reaction temperature. Propane, cyclooctane (89 % of C8products), cis-bicyclo[3.3.0]octane (8 % of C8 products), and n-octane (3 % of C8 products)
were produced at room temperature. Free COD in solution was not detected. Propane,
cyclooctane (95 % of C8 products), and cis-bicyclo[3.3.0]octane (5 % of C8 products) were
produced at -10 degrees Celsius. The ratio of C8 products did not change within
2s experimental error over the course of the hydrogenation. The absence of n-octane at -
10 degrees Celsius may indicate that its formation proceeded via a higher energypathway than hydrogenation of Compound 1.
Figure 1 shows plots of equiv Ru adatoms (based on hydrocarbon products
30 in solution adsorbed by Pt) versus time for hydrogenations carried out at 20 degrees
Celsius and at -10 degrees Celsius. The number of equiv Ru are relative to the initial
number of sites on black Pt, and they were assumed to equal the total of cyclooctane,
cis-bicyclo[3.3.0]octane, and n-octane in each aliquot as determined by GC.
3s Referring to Figure 1, there was a kinetic burst during the initial stages of
the hydrogenations carried out at -10 degrees Celsius that ended after deposition of 0.2
to 0.5 equivalents ("equiv") of Ru. The size of the burst varied among Pt samples. The
-16-
CA 02202446 1997-04-11
rate of hydrogenation slowed after the burst, and then increased as more Ru was
deposited on Pt to reach a maximum, constant rate after deposition of 1.5 - 1.8 equiv
Ru. This rate remained constant until Compound 1 was depleted from solution.
Figure 2 shows a plot of consumption of Compound 1 and total
production of all C8 product hydrocarbons (cyclooctane and cis-bicyclo[3.3.0]octane) in
solution between hydrogenation of 0 and ~ 1.8 equiv Compound 1 at -10 degrees
Celsius, when submonolayer quantities of Ru adatoms were adsorbed by the surface of
Pt. The rates of consumption of Compound 1 and of formation of C8 hydrocarbons
were found to be the same within experimental certainty. The decrease in Compound
1 and the increase in C8 hydrocarbons in solution were also found to be the samewithin experimental certainty at all stages of the hydrogenation. Since there was no
observable time lag between consumption of Compound 1 and the formation of C8
hydrocarbons in solution, it may be concluded that the lifetimes of the surface
hydrocarbonyls were short on the time scale of the hydrogenation. It was therefore
possible to observe in real time the number of Ru adatoms adsorbed by the Pt surface by
monitoring either the concentration of Compound 1 or the total concentration of C8
hydrocarbon products in solution.
As the lifetimes of the surface hydrocarbonyls appear short on the time
scale of the hydrogenation, the curves in Figures 1 and 2 show the actual evolution of
the surface's activity towards hydrogenation of Compound 1 as the coverage of Pt by
Ru increased. Kinetic bursts similar to those observed during low temperature
hydrogenations of Compound 1 (Figures 1 and 2) have also been previously observed
2s by others during low temperature hydrogenations of (COD)Pt(Me)2 over Pt black
carried out under reaction-rate limiting conditions. It has been theorized that the
bursts during hydrogenation of Compound 1 result from a proportion of the activesites on Pt being substantially more active than the others, and that these highly active
sites react quickly during the initial stages of the hydrogenation. Alternatively, it may
be that the initial burst is caused by the surface being saturated with dihydrogen at the
beginning of the hydrogenation, and that the decrease in rate after the burst results
from limitations in mass transport of dihydrogen to the surface. It is not believed that
mass transport of dihydrogen limited the rate of hydrogenation after the initial burst
because the rate of reaction increased as more Ru was deposited on Pt.
3s
The increase in rate after the burst may indicate that Ru was more active
than Pt towards hydrogenation of Compound 1. It is believed that the rate increased
-17-
CA 02202446 1997-04-11
until all the active sites on Pt were covered by Ru. To investigate this possibility,
hydrogenations were interrupted after the maximum, constant rate was achieved and
the resulting surfaces were analyzed using cyclic voltammetry.
5 Characterization of the Ru Surface
Hydrogenations were interrupted after the desired number of equiv Ru was
deposited on Pt (as determined by GC or by UV-VIS absorbance) by quickly lifting the
gauze above the reaction mixture, rinsing the gauze in the reactor with cooled (-10
0 degrees Celsius) dihydrogen-saturated hexanes, removing the hexanes from the
reactor, and drying the gauze under a stream of dihydrogen gas. The reactor was
transferred to a glovebox and the gauze, protected by surface hydrides, was transferred
through the atmosphere of argon in the glovebox to an electrochemical cell. An argon-
saturated 1.0 M aqueous solution of H2SO4 was transferred to the cell while holding
the gauze above the level of the solution, and according to known procedures, the
potential of the gauze was set to -0.19 V concurrent with immersion of the gauze into
the electrolyte.
A surface resulting from hydrogenation of 2.7 equiv Ru was analyzed by
20 recording voltammograms before and after adsorption of a monolayer of carbon
monoxide. The voltammograms were compared to those of a control black Ru surfaceprepared by electrochemical deposition of excess Ru on a black Pt gauze. As described
by other workers, the carbon monoxide was adsorbed at -0.19 V before initiating the
potential sweeps. These other workers showed, using bulk Ru-Pt alloys, that the
25 potential of the anodic wave for oxidation of adsorbed carbon monoxide depended on
the ratio of Ru and Pt atoms at the surface.
Figure 3 shows the cyclic voltammograms. The voltammograms of the
control black Ru surface and the surface resulting from hydrogenation of 2.7 equiv Ru
30 were quite similar both in shape and in peak positions before and after adsorption of
the carbon monoxide. To further characterize the surface after the maximum, constant
rate was achieved during hydrogenation of Compound 1, voltammograms were
recorded in an aqueous solution of H2SO4 and MeOH of a surface resulting form
hydrogenation of 3.5 equiv Compound 1. Figure 4 shows the resulting
35 voltammograms. There was neither adsorption of methanol nor were there
appreciable oxidation currents below 0.46 V. This behavior is identical to that observed
by others using a pure Ru surface.
-18-
CA 02202446 1997-04-11
It has been theorized that the nearly identical electrochemical behaviors of
the control black Ru surface and of the surfaces resulting from hydrogenation of 2.7
and of 3.5 equiv Compound 1, and that the maximum, constant activity of the surface
5 towards hydrogenation of Compound 1, indicate that the coverage of Pt by Ru isessentially complete after adsorption of 1.5-1.8 equiv Ru by the surface of Pt. That more
than 1 equiv of Ru was required to cover the active sites on Pt is theorized to be due to
a combination of factors. These factors may include uncertainties in the measured
surface area of black Pt, hydrogenation occurring on the adsorbed Ru as well as on Pt,
0 the mobility of the Ru atoms on the surface, the relative affinities of Ru adatoms for
Ru and for Pt, and that more than one equiv of Ru adatoms may likely be required to
cover the rough atomic morphology of black Pt. It is also noted that the radius of Ruis
~ 96 % that of Pt. The number of equiv Ru in Figures 1 and 2 is therefore used as an
approximate measure of surface composition.
The surface area of the catalyst surface was estimated after hydrogenation
of 2.7 equiv Compound 1 using the charge associated with oxidation of a monolayer of
adsorbed carbon monoxide. The measured surface area was 67 % that of the black Pt
before hydrogenation of Compound 1. Oxidation of carbon monoxide can only be used
20 to approximate the surface area of Ru. It was therefore proposed that major changes in
surface area did not occur during hydrogenation of Compound 1.
Composition of the Ru Deposit
2s The amount of Ru deposited on Pt was determined after several
hydrogenations by anodic stripping of Ru from the resulting surfaces. Anodic stripping
was carried out in 1.0 M aqueous solution of NaOH at room temperature using a 9 V
battery as power source. UV-VIS spectroscopy showed that Ru in the resulting
electrolyte was mainly in the form of sodium ruthenate. The amount of Ru in
solution was determined by inductively coupled plasma spectrophotometry (ICP) and
was found to equal the total of cyclooctane, cis-bicyclo[3.3.0]octane, and n-octane in the
hydrogenation mixture as determined by GC. All the Ru atoms generated by
hydrogenation of Compound 1 were therefore adsorbed by the surface of Pt. These
results, taken together with the results of the GC-UV-VIS studies shown in Figure 2
3s indicate that little, if any, carbon from COD was trapped by the Ru deposit.
-19-
CA 02202446 1997-04-11
The chemical composition of the Ru deposit was further investigated by
hydrogenating at room temperature ~ 30 equiv Compound 1 over Pt black powder with
an approximate dispersion of 8 %. The molar ratio of Ru to Pt in the resulting particles
was 2.9: 1 according to neutron activation analysis.
Oxidation of Methanol by Ru-Pt Surfaces
Figure 5 shows the potentiodynamic activity for oxidation of methanol by
black Pt and by a black Ru-Pt catalyst resulting from interrupting a hydrogenation of
0 Compound 1 after deposition of 0.11 equiv Ru. The onset of oxidation was ~100 mV
lower for Ru-Pt than for Pt, and the current remained higher (by up to a factor of ~ 14)
until the potential reached ~ 0.65 V. This behavior is similar to those of the bulk alloys
of Ru-Pt studied by others for which the surface compositions were precisely known.
The activity of the Ru-Pt surface noticeably decreased after four sweeps up to 1.0 V. Ru
dissolves in sulfuric acid solutions at potentials above ~ 0.65 v,5(a-c) and it is believed
that this decrease in activity was caused by dissolution of Ru above this potential.
Analysis (ICP) of the electrolyte for Ru ions after 6 sweeps showed that ~ 54 % of the Ru
originally on the surface had dissolved.
To study the stability of these surfaces near the upper operating potentials
of an anode in a methanol fuel cell, a potentiostatic oxidation of methanol was carried
out at 0.158 V for 45 min. using a Ru-Pt catalyst surface resulting from interrupting a
hydrogenation of Compound 1 after deposition of 0.33 equiv Ru. Figure 6 shows the
variation of current with time for the potentiostatic oxidation of methanol using both
2s the Ru-Pt catalyst and black Pt in aqueous solutions of H2SO4 and MeOH. A high
initial current followed by a rapid decrease to a lower, more steady value is typical
behavior for the electrochemical oxidation of methanol using catalysts that contain Pt.
It is believed that the high initial current results from the rapid dehydrogenation of
methanol and oxidation of the resulting surface hydrides. The loss in initial current is
proposed to result from poisoning of the catalyst surface by carbon monoxide or COH.
For both the Ru-Pt and the Pt catalysts, the current generated by the poisoned surfaces
decreased slightly over the course of the oxidations. The current generated by the
poisoned Ru-Pt surface remained higher than that of the Pt surface by a factor of ~ 13,
indicating that surfaces resulting from deposition of submonolayer quantities of Ru on
3s Pt by hydrogenation of Compound 1 may be appreciably durable under these
conditions.
-20-
CA 02202446 1997-04-11
To determine if dissolution of Ru into the electrolyte occurred under
these conditions, we carried out a potentiostatic oxidation of methanol at 0.256 V (a
higher potential than the previous potentiostatic oxidation) first for 0.5 h, and then for
an additional 1 h using a Ru-Pt catalyst surface generated by hydrogenation of 0.8 equiv
5 Compound 1. Analysis using ICP showed only traces of Ru ions in the electrolyte after
the first 0.5 h. The signal was too small to accurately quantify, but roughly
corresponded to 0.06 % of the Ru originally on the surface. ICP analysis of the same
electrolyte after oxidation of methanol for a further 1 h showed no change in the
amount of Ru ions in solution.
Experimental Section
General
lS Argon (pre-purified) was passed through molecular sieves (activated type 4 A)
prior to use. Dihydrogen (pre-purified) and carbon monoxide (ultra high purity) were
used as received. Water was deionized, distilled from alkaline permanganate under
nitrogen, and purged with argon for 30 min. prior to use. Hexanes (HPLC grade) were
passed through aluminum oxide (grade 1), hydrogenated (pressure dihydrogen = 1 atm)
20 over platinum black for 24 hr, and distilled from potassium under argon. Heptane and
decane were purified similarly. Methanol (HPLC grade ) was distilled from Mg(OCH3)2
under argon. Diethyl ether was distilled from potassium/benzophenone under argon.
Cyclohexane-d12 was flash distilled from potassium/benzophenone under argon and
degassed by 3 freeze-pump-thaw cycles. H2SO4 ( ACS grade) was used as received.
2s Rubber septa were extracted for 24 h with HPLC grade hexane in a soxlet extractor, and
dried under vacuum. All glassware was rinsed with a 1:5 mixture of 30% aqueous
hydrogen peroxide: concentrated H2SO4, water, a 5% mixture of ammonium
hydroxide in absolute ethanol, ethanol, and dried in an oven. ( 11 4-1, 5-
cyclooctadiene)Ru( 11 3-C3Hs)2 (Compound 1) was prepared by a known method and
30 twice sublimed under vacuum.
The reactor used for the hydrogenations was a 2.3 x 10.3 cm Pyrex (trade-
mark) tube containing a 4 x 14 mm Teflon-coated magnetic stir bar and fitted with a
rubber septa pierced with a disposable pipette (used as a dihydrogen gas inlet) and a
3s glass tube supporting the blacked platinum gauze.
CA 02202446 1997-04-11
Electrochemical experiments were performed using a Pine Bipotentiostat
Model AFCBP1 controlled with Pinechem 2.00 software or using a homemade
potentiostat equipped with a Hewlett Packard 7004 chart recorder. Inductively Coupled
Plasma Spectroscopy (ICP) was performed using a Perkin Elmer Optima equipped with
s an atomic emission detector. Gas Chromatography - Mass Spectrometry was performed
using a VG-7070E with a Varian 6000 GC fitted with a 30 m J&B DB5 column using aMSS data system. Electrolytes were purged with argon for at least 10 min. prior to use
and electrochemical experiments were performed under argon unless stated otherwise.
The reference electrode was an anodized silver wire behind a D-porosity glass frit, but
0 potentials are referred to a standard calomel electrode in the same electrolyte. The
counter electrode was a blacked platinum wire behind a D-porosity glass frit. Gas
chromatography (GC) was performed on a Hewlett Packard series 530 ,u 10 m methylsilicone column # 19057-121 fitted to a Hewlett Packard 5980A gas chromatograph with
a Hewlett Packard 3392A integrator. 1H NMR Spectra were measured on a Bruker AM-
400 NMR spectrometer operating at 400.13 MHz.
Blacked Platinum
Platinum gauze (52 mesh, 99.9 %, 25 x 25 mm, Aldrich) was threaded with
20 platinum wire (~ 200 mm in length, 0.127 mm in diameter, 99.9 %, Aldrich) andsupported by flame sealing the wire leads through 3 mm uranium glass tubing. Thegauze was blacked and its surface area determined from the coulometric charge of the
hydrogen adsorption region in cyclic voltammograms recorded in 1.0 M H2SO4
according to known procedures. The blacked platinum was then held at 1.2 V for 22s min., rinsed with four 2 mL portions of purified water under argon, and quickly
transferred wet through air to the hydrogenation reactor. The blacked platinum was
dried with a stream of argon for 3 hr, and exposed to a stream of dihydrogen gas for a
further 1 hr to reduce the surface oxides and to remove the resulting water.
30 Hydrogenation of Compound 1 Over Blacked Platinum
Dihydrogen-purged hexanes (15 mL) were transferred via cannula to the
reactor containing the hydrogen-saturated blacked platinum and the reactor was
immersed into the cooling bath. The blacked Pt was kept above the hexanes until the
3s hydrogenation was begun. A slow stream (~ 10 mL/min.) of dihydrogen was passed
through the solution throughout the reaction. Freshly sublimed Compound 1 (10 to 12
mg) was weighed into a small vial in a dry box and the vial was capped with a rubber
-22-
CA 02202446 1997-04-11
septum. A weighed amount of decane was added to the vial containing Compound 1
using a 5 ~L syringe to act as an internal standard for GC analysis. The contents of the
vial were dissolved in hexanes then quantitatively transferred to the reactor with a
cannula and with several rinses with hexanes, and the volume of solution in the
5 reactor was brought up to 20 mL. The stir rate was set to 800 rpm, the platinum was
immersed into the reaction mixture, and timed aliquots (100 to 300 ,uL) were
cannulated from the reaction mixture. The aliquots were analyzed using GC and UVspectrophotometry. To interrupt the hydrogenation at a desired coverage, the
platinum was raised above the solution, and the solution was cannulated out of the
0 reactor. The catalyst surface and the reactor walls were rinsed with four 5 mL portions
of hexanes and the catalyst was dried under a stream of dihydrogen gas for 30 min.
Control Experiments
The entire hydrogenation procedure was repeated in the absence of
Compound 1. The surface area of the platinum as determined using cyclic
voltammetry had decreased by 6 % during the treatment by dihydrogen.
Dihydrogen (pressure ~ 1 atm) was bubbled through a hexane (40 mL)
20 solution of Compound 1 (25 mg) for 30 min. at room temperature. Analysis of the
solution by GC showed no reaction occurred. This experiment was repeated in
cyclohexane-d12 at a higher concentration (39.6 mg of Compound 1 in 2.5 mL solvent).
Analysis by 1H NMR spectroscopy also showed that no reaction occurred after 30 min.
25 Anodic Stripping of Ruthenium
The ruthenium-platinum surface obtained by hydrogenation of
Compound 1 was transferred in air to an electrochemical cell that contained a Teflon-
coated stirbar, a 1.0 M solution of NaOH in argon-purged water (80 mL). The a black
30 platinum counter electrode was fitted behind a D-porosity glass frit. The cathode ("+"
terminal) of a 9 volt battery was connected to the ruthenium-platinum electrode, and
the anode ("-" terminal) was connected to the counter electrode. The solution was
stirred under argon for 10-30 min. as the colour of the solution turned orange. The
solution was quantitatively transferred to a 100 mL volumetric flask and diluted to
35 volume with 1.0 M NaOH. The amount of ruthenium in solution was determined
using ICP analysis. UV-VIS spectra of the solutions indicated that the product of the
-23-
CA 02202446 1997-04-11
anodic stripping was sodium ruthenate. Further anodic stripping in fresh electrolyte
showed that all the ruthenium was stripped from the electrode by this treatment.Hydrogenation of Compound 1 Over Platinum Black
s
In air, platinum black (10 to 11 mg; Johnson Matthey, fuel cell grade) was
weighed into a 3 dram, 21 x 50 mm vial containing a 4 x 14 mm Teflon-coated magnetic
stir bar and capped with a rubber septum pierced with two steel needles used as a gas
inlet and outlet. The reactor was flushed with argon for 10 min. and then placed under
0 an atmosphere of dihydrogen gas. Hexanes (~ 3 mL) were added via cannula to a 3-
dram vial that was capped with a rubber septum and that contained freshly sublimed
Compound 1 (42 to 48 mg; weighed in a dry box). The solution of Compound 1 in
hexanes was flushed with a stream of dihydrogen gas (~ 20 mL/min.) for 2 min.
Hexanes (~ 1 mL) were added via cannula to the reactor containing the hydrogen-
saturated platinum black and a stream of dihydrogen gas (~ 20 mL/min.) was passed
through the hexanes for 1 min. while stirring the mixture at ~ 400 rpm. The solution
of Compound 1 in hexanes was transferred to the reactor with a cannula using ~ 2 mL
more hexanes for rinses. The total volume of the solution in the reactor was brought
up 6 mL. A stream of dihydrogen (~ 10 mL/min.) was passed through the reaction for
the duration of the reaction. After 24 h at room temperature, the solution was drained
using a cannula and the contents of the reactor were rinsed with hexanes (three ~5 mL
portions). The contents of the reactor were then dried at room temperature undervacuum (~ 0.01 torr) for 2 h.
2s Neutron Activation Analysis
Samples and standards, packed in HNO3 washed polyethylene 100 IlL
tubes, were individually irradiated for 300 s at a neutron flux of 1 x 1011 n cm~2 s-1 in
an inner site of the University of Alberta, Canada SLOWPOKE II Nuclear Reactor.
Following a decay period of 18 minutes, each sample was counted for 300 s at a sample-
to-detector distance of 3 cm using a 34 % hyperpure Ge detector attached to an 8k
channel PC-based multichannel analyzer. Analysis was performed by the comparatormethod of INAA using RuCl3-3H2O and K2PtCl6 (98 %) as standards. Platinum was
quantified using the 542.96 and 185.76 keV y-ray emissions of 199Pt (T1/2 = 30.8 min.)
produced via the reaction 198Pt(n, y ) 199pt while Ru was determined using the 724.27
keV y-emission of 105Ru ((T1/2 = 4.44 h) produced via the thermal neutron reaction
104RU(n, y ) 105Ru.
-24-
CA 02202446 1997-04-11
Adsorption and Oxidation of Carbon Monoxide
A ruthenium-platinum surface was prepared by hydrogenation of 2.7
s equiv Compound 1. The reaction vessel containing the rinsed and dried electrode was
transferred to a glovebox, and the electrode was transferred to the electrochemical cell
under argon. The cell was removed from the glovebox, flushed with argon, and fitted
with two disposable pipettes (to be used as gas inlets), with a reference electrode, and
with a counter electrode. An argon-purged solution of H2SO4 (1.0 M in water) was
0 transferred to the cell via a polyethylene cannula. The potential of the ruthenium-
platinum working electrode was set to - 0.19 V concurrent with immersion into the
electrolyte. With stirring (700 rpm) at room temperature, carbon monoxide was
bubbled through the solution for 30 min. followed by bubbling argon through the
solution for 2 min. The stirring was stopped and the potential of the working electrode
was swept at 5 mV/s up to 0.723 V then down to -0.262 V.
A control black ruthenium electrode was prepared by potentiostatic
deposition of ruthenium on a black platinum electrode at - 0.267 V for 15 min. from a
stirred solution of RuC13-3H2O (.005 M) and H2SO4 (1.0 M) in water. Using the
20 coloumbic charge passed during the deposition, we calculate that 15 equiv of Ru were
deposited on the electrode surface. The electrode was rinsed with H2SO4, transferred
quickly in air to the electrochemical cell, and the oxidation of adsorbed carbonmonoxide was repeated as described above.
2s Oxidations of Methanol
A hydrogenation of Compound 1 was interrupted as described above after
deposition of 0.11 equiv ruthenium. The rinsed and dried catalyst surface was
transferred in a glovebox to an electrochemical cell. An argon-purged aqueous
30 solution (90 mL) of H2SO4 (0.5 M) and methanol (1.0 M) at 40 degrees Celsius was
transferred via cannula to the cell. The potential of the ruthenium-platinum surface
was set to - 0.200 V concurrent with immersion into the electrolyte and then swept to
1.0 V at 5 mV/s while maintaining the temperature of the electrolyte at 40 degrees
Celsius using a heated oil bath.
3s
The potentiostatic oxidations of methanol were carried out similarly at 25
degrees Celsius, [H2SO4] = 0.5 M, [MeOH] = 0.5 M. The potential of the catalyst surfaces
-25-
. CA 02202446 1997-04-11
were set to - 0.17 V concurrent with immersion into the electrolyte, and then set to the
desired potential to begin the oxidation.
-26-
