Formation of Pd2X2(μ-S)(μ-dpm)2 by reaction of PdX2(dpm) with H2S in the presence of alumina (X = halogen, dpm - diphenylphosphinomethane)

Article abstract of DOI:10.1016/S1381-1169(98)00201-5

Alumina promotes the reaction of PdX₂(dpm) (3) (X = Cl, Br, I) with H₂S to form Pd₂X₂(μ-S)(dpm)₂ (2) and HX, with chemisorption of the HX (quantified by titration) providing the driving force; althoug

Full text of DOI:10.1016/S1381-1169(98)00201-5

Journal of Molecular Catalysis A: Chemical 139 Ž1999. 159–169
ž
/ž
/
ž
/
Formation of Pd X m-S m-dpm by reaction of PdX dpm with
2
2
2
2
ž
H S in the presence of alumina Xshalogen,
2
/
dpmsdiphenylphosphinomethane
)
Terrance Y.H. Wong, Brian R. James , Philip C. Wong, Keith A.R. Mitchell
Department of Chemistry, UniÕersity of British Columbia, VancouÕer, BC, Canada V6T 1Z1
Received 20 February 1998; accepted 27 May 1998
Abstract
1
Alumina promotes the reaction of PdX Ždpm. Ž3. ŽXsCl, Br, I. with H S to form Pd X Žm-S.Ždpm. Ž2. and HX,
2
2
2
2
2
with chemisorption of the HX Žquantified by titration. providing the driving force; although no intermediate species are
detected, the reaction likely proceeds via mercapto species formed via ‘alumina activated H S’. XPS studies show
2
interactions of species 2 and 3 with alumina, these interactions being weak as demonstrated by recovery of the species on
dissolution using an appropriate solvent. For the iodide system, adsorbed HI on alumina undergoes partial photodecomposi-
tion to give I2 and H atoms. q 1999 Elsevier Science B.V. All rights reserved.
Keywords: Pd X Žm-S.Žm-dpm. ; PdX Ždpm.; H S; Alumina
2
2
2
2
2
1
. Introduction
We reported recently the first homogeneously catalyzed conversion of H S to H using dinuclear
2
2
Pd –dpm complexes
Ž
dpmsbis
Ž
diphenylphosphino
.
methane. ŽScheme 1
.
w1x. Kinetic and mechanis-
2
tic studies have been carried out for both the forward
the forward reaction, hydrido mercapto intermediates were detected in low temperature NMR studies
Ž
1™2
.
and reverse
Ž
2™1
.
reactions and, for
w1–4x.
y
Besides dpm, several reagents
Ž
including O-atom donors, and phosphines, CO, CN , biphenyl, or
butadiene
w3–5x. Reactions with halogens remove the sulfur as elemental S8
occur under ambient conditions in two kinetically observable stages via a rapid transannular oxidative
addition to give with generation of S8 a tetrahalo-dipalladium II
The by-product PdX2
.
can remove completely or partially in solution the bridged S in ‘coordinated forms’ from 2
Ž
Eq.
Ž
1
.. w4,6,7x, and these reactions
Ž
.
Ž
.
intermediate w8x that subsequently
Ph PCH P Ph
undergoes slower unimolecular decomposition w4,6,7x.
Ž
Ž
Ž
S
.
.
,
2
2
2
)
Corresponding author. Tel.: q1-604-822-32-66; Fax: q1-604-822-28-47; E-mail: brj@chem.ubc.ca
For convenience, m-dpm in 2 is written simply as dpm.
1
1
381-1169r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved.
PII: S1381-1169Ž98.00201-5

1
60
T.Y.H. Wong et al.rJournal of Molecular Catalysis A: Chemical 139 (1999) 159–169
Scheme 1.
crystallographically characterized
ring, forms under certain conditions w6,7x.
Ž
with XsCl or I
.
and containing a 5-membered
q1r8S8
P–S chelating
Ž .
.
Pd X2ŽmyS.Ždpm. Ž qX ™2PdX2
2
.
Ž
dpm.Ž
3
.
1
Ž .
2
2
2
The catalyzed conversion of H S to H and elemental sulfur is attractive commercially w1,4x, and
2
2
thus efforts to convert 3 to 2 by reaction with H S have been undertaken. This present paper reports
2
on reaction
g-alumina. Implications for catalysis can be seen with the iodide system, where combination of
reactions and , coupled with the known photo-decomposition of HI to generate H and I , leads
to the net reaction: H S™H q1r8 S .
2 which occurs rapidly and completely under ambient conditions in the presence of
Ž .
Ž
1
.
2
Ž .
2
2
2
2
8
2
PdX2
Ž
dpm.Ž
3
.
qH S™Pd X2ŽmyS.Ždpm. Ž
2
.
q2HX
2
Ž .
2
2
2
2
. Experimental
The materials used, synthetic procedures for the complexes, and instrumentation used for H and
1
3
1
1
P H NMR and UV–Vis spectroscopies, and gas chromatography, have been largely described
Ä 4
previously w1–3x. Complexes 2a–c w3x and 3a–c w8–10x were synthesized using established methods.
All experiments were performed under Ar unless otherwise specified. Alumina powder and plates
were purchased from Fisher
Ž
neutral g-Al O3., Alfa Products
Ž
a-Al O , corundum
.
, and Merck
2
2
3
Ž
neutral g-Al O plates, mesh 60, layer thickness 0.2 mm
.
and were pre-dried at 75 or 1508C for 24 h
2
3
before use. Finely ground alumina was prepared via a manual grinding process using a mortar and
pestle. Aliquots of gases were added to solutions from gas-tight syringes, the volumes referring to 1
atm pressure at ;208C.
X-ray photoelectron spectra
Ž
XPS
.
were measured with a Leybold MAX200 spectrometer using a
1235.6 eV which was operated at 10 kV and 20
mA; the emitted photoelectrons were collected from a 2=4 mm area. Survey spectra were measured
with the pass energy set at 192 eV, while the higher resolution narrow scan spectra were measured
non-monochromatized Mg K excitation source
Ž
.
a
2
Ž
.
with a 48 eV pass energy. The binding energies for assignments in the XPS spectra are referenced to
the C 1s peak for adventitious carbon at 285.0 eV. It took approximately 4 h to take samples from
atmospheric pressure to the system operating pressure at 4=10y9 mbar at room temperature
which these spectra were recorded.
r.t. at
.
Ž

T.Y.H. Wong et al.rJournal of Molecular Catalysis A: Chemical 139 (1999) 159–169
161
2
.1. Heterogeneous alumina systems
1.0=1.0 cm2., with or without added reagents, was placed
15 mg or an alumina plate
. Ž
Alumina
either in an NMR tube fitted with a PTFE J. Young valve or septum, or in a Schlenk tube fitted with a
septum. CDCl3 0.7 or 1.0 ml was added, and the systems were either shaken or stirred and left at r.t.
for up to 72 h before analysis; the added reagents were 3a 5 mg, 0.0089 mmol , gaseous H S 110
ml or 0.5 ml, 0.0045 or 0.021 mmol, respectively, gaseous HCl 0.5 ml, 0.021 mmol 12 mg,
as required. In experiments where both 3a and H S were used, 3a was introduced as a
either before or after the addition of H S, with 0.4 or 0.7 ml CDCl initially
Ž
Ž
.
Ž
.
Ž
2
Ž
, or S8
. Ž
0
0
.047 mmol
.3 ml CDCl solution
.
Ž
2
.
3
2
3
being placed in the NMR or Schlenk tube, respectively. Analyses were carried out using NMR
spectroscopy on the liquid phases, and XPS on the alumina plates after drying them for a few minutes
at r.t. under a flow of Ar.
In some cases, an H S solution-pretreated alumina plate was dried under Ar, and placed in another
2
Schlenk tube for reaction with a 1.0 ml CDCl solution of 3a prior to analysis.
3
2
.2. Homogeneous systems
PdCl2
of NaSH
Ž
dpm. Ž3a, 10 mg, 0.018 mmol
was dissolved in CHCl3 10 ml , and an MeOH solution 10
. Ž . Ž
ml
.
Ž
1.0 mg, 0.018 mmol was added dropwise over 30 min. The solvents of the resulting
.
yellow solution were removed, and the remaining yellow residue was dried in vacuo and dissolved in
CDCl3 0.5 ml for analysis by NMR spectroscopy.
Ž
.
2
.3. Low temperature studies
A single, low temperature NMR study was carried out in which H S
Ž
0.25 ml, 0.010 mmol
injected into a septum-sealed NMR tube containing a 0.5 ml CDCl solution of 3a 5 mg, 0.0089
. The sample
.
was
2
Ž
3
mmol and finely ground alumina 15 mg in a MeCNrliquid N slush bath y428C
. Ž . Ž .
was analyzed at y508C 1 h later using NMR spectroscopy, and then placed at r.t. for 24 h and
2
re-analyzed.
2
.4. Synthetic-scale studies
In a Schlenk tube fitted with a septum were placed 25.0 mg PdX2
Ždpm. Ž3a, 0.045 mmol; 3b,
0
.038 mmol; 3c, 0.034 mmol
5 ml, 0.20 mmol
solution turned brown within seconds. The mixture was stirred continuously for 4 h before being
reduced in volume to ;0.5 ml with evacuation of the H S. CH Cl 20 ml was then added to
dissolve all Pd species, and the alumina was filtered out, washed with CH Cl2 2=10 ml , MeOH
, and then dried in vacuo. The
solvents from the filtrates and washings were removed by rotary evaporation, and the resulting residue
was dissolved in CDCl3 ;0.5 ml
and analyzed by NMR spectroscopy. Yield of Pd X2Žm-S.Ždpm.2
a, 24.0 mg )99% ; 2b, 22.0 mg
flask, and 10 ml H O and a few drops of phenolphthalein indicator were added for titration with a
.
, 75.0 mg finely ground alumina, and 2.5 ml CHCl . The mixture was
3
rapidly stirred and H S
Ž
was injected. The initial yellow 3a, 3b or orange 3c
. Ž . Ž .
2
Ž
.
Ž
2
.
Ž
2
2
.
2
Ž
, and CH Cl 10 ml
10 ml. Žthereby eluting all absorbed Pd species. Ž .
2 2
Ž
.
:
2
2
Ž
.
98% ; 2c, 21.0 mg 99% . The isolated alumina was placed in a
Ž . Ž .
2
NaOH solution
placed with ;0.7 ml CDCl in an NMR tube fitted with a J. Young valve. The resulting mixture was
0.010 M . For the iodide system, the isolated alumina from a repeat experiment was
Ž .
3
then placed in the presence of light laboratory, sunlight, or that from an 18 W TLC Hg vapour lamp
Ž .

1
62
T.Y.H. Wong et al.rJournal of Molecular Catalysis A: Chemical 139 (1999) 159–169
for up to 4 days, and the system then analyzed for H in the head-space
Ž
using NMR and GC
.
and I2
2
in solution using UV–Vis ; the alumina was subsequently filtered off, washed with MeOH, and the
Ž .
washings analyzed using UV–Vis and titrated with NaOH solution. Blank titrations were also
performed using unused alumina, and alumina isolated from the above procedure for the chloride
system when H S was not used.
2
Possible reactivity of 3a with S in the presence of alumina was investigated. A mixture of 3a
8
Ž
25.0 mg, 0.045 mmol
ml was rapidly stirred for 4 h in a Schlenk tube, when the initially white alumina gradually turned
orange although the colour of the supernatant remained unchanged yellow . CH Cl 20 ml was
then added to dissolve all Pd species, and the mixture was filtered to separate out the alumina, which
was subsequently washed with CH Cl2 2=10 ml , MeOH 10 ml , and CH Cl 10 ml . The
solvents from the filtrate and washing were removed by evaporation, and the resulting residue was
dissolved in CDCl3 ;0.5 ml and analyzed using NMR spectroscopy.
Possible decomposition of H S on alumina was also investigated. An alumina plate
, finely ground alumina 75.0 mg , S8 14 mg, 0.45 mmol , and CHCl3 2.5
. Ž . Ž . Ž
.
Ž
.
Ž
2
.
2
Ž
.
Ž
.
Ž
2
.
2
2
Ž
.
Ž
1.0=10.0
2
2
cm
taken out and dried in air. The alumina was then removed from the plate, washed with CHCl3
ml , and the washings collected and analyzed by UV–Vis. A control experiment was also carried out
using alumina in the absence of H S.
. Ž .
, CHCl3 20 ml , and 1 atm H S in a Schlenk tube were left at r.t. for 4 h when the plate was
2
Ž
;5
.
2
2
.5. Photodecomposition of HI (see also Section 2.4)
Photodecomposition of HI
Alumina 15 mg was placed in an NMR tube fitted with a J. Young valve. Then
was added followed by HI and the mixture left in laboratory light for 24 h
aq.. Ž2 ml, 0.016 mmol
prior to analysis for H and I see Section 2.4 , and HI
aq.. Ž2 ml of a 7.58 M solution, 0.016
was added first, with the head space analyzed ‘immediately’ or after 4 days, followed by
addition of CDCl3 0.5 ml with the mixture again left in laboratory light for 24 h prior to analysis. A
control sample of I2 0.01 mg, 0.052 mmol , alumina 15 mg , and CDCl3 0.5 ml was placed at r.t.
for 24 h in the presence of laboratory light before being analyzed using UV–Vis spectroscopy.
Ž .
aq. in the presence of alumina in CDCl was studied qualitatively.
3
Ž
.
a CDCl3 0.5 ml
Ž . Ž .
Ž
.
Ž
2
.
Ž
b
.
Ž
2
mmol
.
Ž
.
Ž
.
Ž
.
Ž
.
2
.6. Hydrogenation of cis-cyclooctene
A synthetic-scale reaction was repeated for the iodide system see Section 2.4 , and the isolated
Ž .
alumina
Ž
washed and dried in vacuo
.
was placed under an atmosphere of Ar in a septum-sealed NMR
2 ml,
was then injected, and the NMR tube was shaken. The sample was then left at r.t. for
4 h before being analyzed. Control studies using irradiated and non-irradiated, ‘unreacted’ alumina
.. were also performed.
tube. The sample was then irradiated using the TLC Hg vapour lamp for 24 h. Cis-cyclooctene
Ž
y
5
1
2
Ž
.5=10 mol
.
i.e., samples not subjected to reaction
Ž
2
3
. Results
Table 1 summarizes NMR data for the well characterized complexes Pd X2Žm-S.Ždpm.2 Ž
2
.
and
2
3
1
1
PdX2
Ž
dpm. Ž
3
.
. The
P H NMR spectra of 2a–c reveal a singlet at d 5.5–6.1 for the four
Ä 4
equivalent P atoms, and singlets are seen for the two equivalent P atoms of 3a–c in the d y55 to

T.Y.H. Wong et al.rJournal of Molecular Catalysis A: Chemical 139 (1999) 159–169
163
Table 1
NMR data for the palladium complexes
Compounda
1
b
31
1
c
d Ž H.
d Ž PÄ H4.
d,e
5.52d
Pd Cl Žm-S.Ždpm. Ž2a.
2.79 Ž12.6, 3.5.
4
2
2
2
e
.73 Ž12.6, 6.1.
d,e
5.96d
6.14
Pd Br Žm-S.Ždpm. Ž2b.
2.88 Ž12.8, 3.2. Ž2.90.
2
2
2
e
4
.83 Ž12.8, 7.6.
d,e
d
Pd I Žm-S.Ždpm. Ž2c.
3.06 Ž14.0, 3.0.
6.08
2
2
2
e
4
.95 Ž14.0, 6.0.
f
PdCl Ždpm. Ž3a.
4.21 Ž10.8.
y54.7
y56.2
y63.2
2
f
PdBr Ždpm. Ž3b.
4.37 Ž10.5.
2
f
PdI Ždpm. Ž3c.
4.42 Ž10.0.
2
a
For convenience, m-dpm in 2a–c is written as dpm.
b
In CDCl , unless stated otherwise, at 208C with respect to TMS; J
andror JPH values in Hz are given in parentheses; signals for CH2
3
HH
protons.
c
Unless otherwise stated, singlets in CDCl3 at 208C with respect to 85% H PO , downfield being positive.
3
4
d
e
f
In CD Cl .
2
2
Doublets of quintets for each of two sets of CH2 protons.
Triplet.
1
y63 region w3,8–10x. In the H NMR spectra, the CH resonances of 2a–c appear as AB doublets
2
with additional coupling to the four P atoms, while those of 3a–c appear as 1:2:1 triplets w3,8–10x.
1
H NMR analysis of a suspension of g-alumina
Ž
mesh size ;200
presumably due to surface OH groups . With H S added, the expected H
singlet at d 0.82 for free H S w2x is not observed because this species is adsorbed onto the alumina
.
in CDCl reveals a broad,
3
1
unresolved signal at d 1.2
Ž
.
2
2
Ž
see Section 4
in the alumina from white to orange; NMR analysis of the filtrate shows only the presence of 3a.
Subsequent introduction of a stoichiometric amount of H S see Eq. , H S:3as0.5 resulted in an
immediate colour change of the alumina to orange–brown and the solution to a brown–yellow colour.
H and
.Ždpm.2
already treated with a stoichiometric amount of H S, also yielded an immediate colour change, both
.
. Addition of 3a to a suspension of alumina causes over ;1 h a gradual colour change
Ž
Ž
2
.
.
2
2
1
31
1
P H
Ä 4 NMR spectra here show, in addition to unreacted 3a, trace amounts of Pd Cl2Žm-
Ž .
2a . Reversing the order of addition, i.e., adding 3a to a CDCl suspension of alumina
3
2
S
2
the alumina and solvent turning orange–brown, but 2a was now formed in ;10% yield
from the relative integrated areas of the respective CH proton signals , or ;60% if the alumina was
Ž
determined
.
2
pre-dried at 150 instead of 758C. No 2a was formed, however, if the alumina was isolated after the
addition of H S and, prior to the addition of 3a, placed in fresh, H S-free CDCl , because H S
2
2
3
2
desorbs from the alumina during the isolation process see Section 4 . A quantitative yield of 2a can
Ž .
be affected using alumina of lower mesh size. For example, NMR analyses show, that after 4 h at r.t.,
a was completely converted to 2a when alumina of mesh size 60 i.e., from a TLC plate or a finely
ground ;200 mesh size was used. Use of corundum a-alumina with an extremely low mesh size of
y100 gave only a 10% yield after 24 h and reached a maximum 50% yield after 72 h Fig. 1
3
Ž
.
Ž
.
Ž
.
.
XPS analysis of g-alumina reveals signals arising from Al 2p, Al 2s, and O 1s photoelectrons with
corresponding binding energies of 74.5, 119.8, and 531.8 eV. Additional signals arising from
adventitious carbon and nitrogen are also observed, the C 1s and N 1s photoelectrons, for instance,
having binding energies of 285.0 and 399.3 eV, respectively; the data are similar to those found in the
literature w11x. With 3a adsorbed on the alumina, the XPS spectrum also shows the Cl 2p, Pd 3d5r2
,
and Pd 3d photoelectron signals with binding energies of 198.2, 337.0, and 342.5 eV, respectively
3
r2
Ž
Figs. 2 and 3 . For alumina isolated from reaction 2 , two additional signals are also seen, namely,
. Ž .

1
64
T.Y.H. Wong et al.rJournal of Molecular Catalysis A: Chemical 139 (1999) 159–169
1
y3
Fig. 1. H NMR spectra Ž300 MHz. showing the conversion of PdCl Ždpm. Ž3a, 8.9=10
M. with 2.5 mole equivalent H S to
2
2
Pd Cl Žm-S.Ždpm. Ž2a. in CDCl3 in the presence of corundum Ža-alumina. Ž15 mg. at room temperature.
2
2
2
the S 2p signals at 161.9 and 169.1 eV
or S show a Cl 2p signal at 199.5 eV for the HCl sample, and one broad weak S 2p signal at 169.1
Ž . Ž . Ž .
Fig. 3 . XPS analyses of alumina exposed to HCl g , H S g ,
2
8
eV for the H S or S samples Fig. 4 . Of note, some decomposition of H S to elemental sulfur over
Ž .
2
8
2
g-alumina was observed; for example, alumina previously exposed to H S for 4 h was washed with
2
CHCl , and the eluate was analyzed using UV–Vis spectroscopy which revealed the presence of S
3
8
y
1
y1
Ž
lmax 244, 266 nm; ´s830 and 910 M cm , respectively w6x
The Pd X products from reaction were recovered in quantitative yields
m-S.Ždpm
from synthetic-scale experiments, while the amounts of HX adsorbed on the alumina
.
.
Ž
.
2
Ž
2
.
2
Ž .
2
2
Ž
)98%
.
Ž
.
y
5
y
5
y
y
determined by titration were 5.2, 3.4, and 3.4=10 mol "0.2=10 mol for the Cl , Br , and
Ž .
Fig. 2. X-ray photoelectron spectrum ŽMg K , r.t.. showing the Pd 3d photoelectron signals resulting from adsorption of PdCl Ždpm. Ž3a.
a
2
andror Pd Cl Žm-S.Ždpm. Ž2a. on alumina.
2
2
2

T.Y.H. Wong et al.rJournal of Molecular Catalysis A: Chemical 139 (1999) 159–169
165
Fig. 3. X-ray photoelectron spectrum ŽMg K , r.t.. showing the Cl 2p and S 2p photoelectron signals resulting from adsorption of
a
PdCl Ždpm. Ž3a. andror Pd Cl Žm-S.Ždpm. Ž2a. on alumina; the signal at 169.1 eV is due to S8 Žsee text..
2
2
2
2
Iy systems, respectively, allowing for blank titrations giving 1.1 =10y mol acid. These
"0.1
Ž .
5
y
5
titration results are in good agreement with the theoretical values of 4.5, 3.8, and 3.4=10 mol for
the respective systems. In the 4-day laboratoryrsunlight photochemical experiment with the iodide
system, the alumina acquired a yellow colour but the CDCl solvent remained colourless. Subsequent
3
exposure to 254 nm light for 4 h led to the alumina becoming orange and the solvent purple. UV–Vis
analysis of the solvent revealed an absorption band at 512 nm indicating the presence of I
2
y
7
Ž
;1=10 mol
from the orange alumina contained 1.8=10
.
w6x, but GC and NMR analyses revealed no detectable H . The MeOH washing
2
y
6
y
y
6
y
mol I
Ž
i.e., 5.4=10
mol I as determined by
.
3
y
4
UV–Vis on comparison with an authentic sample of I
1
contain 2.8=10
w2.8q0.54q2
available see above
alumina without prior exposure to light.
Reaction was studied at low temperatures, but no intermediate species were detected. For
Ž
lmax 292, 360 nm; ´s2.89=10 and
3
4
y1
y1
.66=10 M cm , respectively w6x. ŽFig. 5
.
. The filtered alumina was then found by titration to
mol acid, and the total number of moles of I
x=10 , i.e., ;3.4=10 mol, in excellent agreement with the total iodide
. Of note, some I also forms from addition of a CHCl solution of I to
y
5
y
or HI generated equals
Ž .
y
5
y5
Ž
0.01
.
.
y
Ž
3
3
2
2
Ž .
example, for the chloro system at y428C, subsequent NMR analysis at y508C revealed only
unreacted 3a and 2a in 50% yield. Some mechanistic insight, however, came from reaction of 3a with
Fig. 4. X-ray photoelectron spectra ŽMg K , r.t.. showing the S 2p and Cl 2p photoelectron signals resulting from adsorption of H S, S , or
a
2
8
HClŽg. on alumina.

1
66
T.Y.H. Wong et al.rJournal of Molecular Catalysis A: Chemical 139 (1999) 159–169
Fig. 5. Electronic spectrum Ž1 cm cell. of Iy3 Žin 25.0 ml MeOH. isolated from synthetic-scale studies.
1
equiv. NaSH in solution which effected complete conversion to 2a, with liberation of H S. Possible
2
reactivity of 3a with S in the presence of alumina to form 2a was examined, but NMR analysis
8
revealed only unreacted 3a.
Qualitative studies showed that an HI
produce I2 within minutes at r.t. in laboratory light
orange colour generated in the alumina is due to I3
Ž
aq.
.
rCDCl solution over alumina photo-decomposes to
3
Ž
.
Ž
as evidenced by UV–Vis spectroscopy. An
see above ; no H was detected. Of interest,
y
.
2
1
however, H NMR analysis revealed that when corundum was used, a broad singlet at d 5.2 was seen
Fig. 6 and this is attributable to a CDHX2 XsCl andror I species see Section 4 . Control
samples of alumina in CDCl with or without I in CDCl , even using 254 nm light, showed no
Ž
.
Ž
.
Ž
.
3
2
3
changes, the alumina remaining white and the I 512 nm absorption band remaining constant in
2
intensity.
1
Fig. 6. H NMR spectrum Ž300 MHz. showing the results of the photodecomposition of HIŽaq.. Ž;2 ml of a 7.58 M aq. solution. at r.t. in
CDCl3 in the presence of corundum Ža-alumina. Ž15 mg.; XsCl andror I.

T.Y.H. Wong et al.rJournal of Molecular Catalysis A: Chemical 139 (1999) 159–169
167
8
There was small but significant conversion
cyclooctene using alumina recovered from reaction
conditions described in Section 2.6; no conversion was seen with ‘blank alumina’.
Ž
5=10y mol
.
to cyclooctane from the added excess
Ž
Ž
2
.
for the iodide system
.
on irradiation under the
4
. Discussion
The PdX2
heterogeneous system with g-alumina
and HX Eq.
.Ždpm.2 Ž ..; no intermediate species were observed. Of note, the reverse reaction
using solutions of 2 and HX in the ‘free state’ proceeds rapidly and completely, and thus, in the
Ž . Ž .
dpm complexes 3 react with H S cleanly and quantitatively within minutes in a
2
Ž
low mesh, or finely ground higher mesh
. to form Pd X2Žm-
2
S
2
.
Ž
Ž
2
Ž
.
absence of alumina, the forward reaction does not occur w3x. As discussed below, the HX species
chemisorbs on the alumina surface during reaction
forward reaction.
2 and this provides the driving force for the
Ž .
The NMR and XPS data show that adsorption of 2, 3 and H S occurs on alumina, but the
2
interactions are relatively weak. For H S, desorption readily takes place when the alumina is removed
2
from the H S environment, and this is evident from non-reactivity, i.e., within reaction
Ž
2
.
, when
such alumina was subsequently treated with a solution of 3a. Species 2 and 3 readily elute with
solvent, implying physisorption rather than chemisorption as seen for HX see below . XPS
2
Ž
.
measurements and recovery of the Pd species on elution indicate that the complexes remain
unchanged when adsorbed onto the alumina. Thus the observed Pd 3d5r2 binding energy of 337.0 eV
corresponds to those found for PdX2
energy of 198.2 eV is in the range of those reported for chloro complexes
presence of 2 on alumina is further defined by the S 2p signal at 161.9 eV, in the range of 160–162
Ž
phosphine.2 Ž337–338 eV
.
w12x and the observed Cl 2p binding
Ž
197–199 eV
.
w13x. The
2
y
eV seen for S in metal sulfides w14x.
Although facile desorption of H S is evident, XPS analyses of alumina isolated from reaction
Ž
2
.
2
reveal a weak, second S 2p photoelectron signal at 169.1 eV that is also observed in control studies of
alumina treated only with H S or S ; UV–Vis data furthermore show that S which can be eluted
Ž
8
2
8
from the alumina with corresponding loss of the 169.1 eV signal
.
can be formed from decomposition
the
. The 169.1 eV binding
168.8 eV
Ž . w14x, suggesting that the environment of surface
of H S. These studies show that the 169.1 eV signal results from the presence of elemental sulfur
Ž
2
S 2p binding energy associated with non-adsorbed solid S is 164 eV w14x
.
8
2
4
y
energy is close to that of S in ‘free’ SO
O-atoms shifts the S signal to the higher 169.1 eV value. Decomposition of H S to S over alumina
8
2
8
is reported, but the mechanism is poorly understood w15,16x.
Chemisorption of hydrohalic acids by transition aluminas is known w17x. One component of the
XPS Cl 2p signal centered at 198.2 eV noted for the chloride system is the Cl 2p signal of HCl; a
control sample of alumina treated with HCl
Ž
g
.
shows a signal at 199.5 eV.
probably via an S–H moiety ,
.
Reaction proceeds largely via initial H S activation on alumina
Ž
2
.
Ž
2
as seen from only trace conversion to 2 when adsorption of 3 occurred prior to addition of H S, and
2
more obviously when 3 was introduced following desorption of H S during isolation of the alumina.
2
Surprisingly, use of corundum with an extremely low mesh size of y100 yielded only 50%
conversion
hinders the progress of reaction
with use of lower mesh or more finely ground g-alumina, where presumably more active sites are
available; the detailed nature of the sites involved remains
acid–base or H-bonding w17–19x
unknown. Involvement of such active sites is supported by: increased conversion to 2 on drying the
after 72 h and, as noted below, the absence of ‘active’ sites here for S–H bond activation
Ž .
2 . Improved conversions up to quantitative of 3 to 2 were realized
Ž . Ž .
Ž
.
i
Ž .

1
68
T.Y.H. Wong et al.rJournal of Molecular Catalysis A: Chemical 139 (1999) 159–169
alumina at higher temperatures, and ii adsorption of 3 prior to the introduction of H S giving only
Ž .
2
trace amounts of 2. The drying process presumably exposes more active sites as more H O molecules
2
are driven from the surface w20x, while occupation of these sites by 3 would explain zero conversion
with subsequent addition of H S as ‘S–H activation’ could not take place. The relatively low activity
2
of corundum
cally inactive w17x and chemically inert w21x, and thus H S activation by interaction with polar Al–O
of low mesh corresponds to its general properties; the material is both chromatographi-
Ž .
2
bonds in this material is not effective. Not all active sites on g-alumina activate H S for reaction 2 ,
Ž .
2
because this reaction is not quite stoichiometric with respect to H S
Ž
a stoichiometric amount gives
2
only 90% conversion to 2a from 3a, and XPS studies reveal slight decomposition of H S to S8
.
.
2
After activation of H S, anionic ligand exchange within 3 could form the mercapto intermediate
2
X
PdX
Although 3 was not detected by low temperature NMR studies, the observed interaction between 3a
and 1 equiv. of NaSH in solution Eq.
.. provides indirect evidence for such an intermediate
XsCl
3 that could subsequently couple with elimination of H S to give 2 Eq. 3 .
ŽSH.Ždpm. Ž . Ž . Ž Ž ..
2
X
Ž
Ž
4
Ž
.
.
2
H 2S
yH 2S
2
PdX2
Ž
dpm
.
™ 2PdX
Ž
SH.Ždpm
.
™ Pd X2ŽmyS.Ždpm.2
Ž
3
.
2
y2HX
X
3
3
2
2
PdCl2
Ž
. .
dpm.Ž3a q2NaSH™Pd Cl2ŽmyS.Ždpm. Ž2a q2NaClqH2
Ž
4
.
2
2
Previous studies from this laboratory w3x have shown that 3a with excess NaSH gives exclusively
y
Pd2
ligands, or via coupling of Pd
ŽSH. Žm-S.Ždpm.
, which could be formed from 2a by nucleophilic substitution of both Cl
2 2
Ž
SH. Ždpm
.
with concomitant elimination of H S
Ž
cf. Eq.
4
Ž ... Coupling
2
2
X
of 3 could go via Pd X2ŽSH. Ždpm.2, and such species have been detected in low temperature NMR
2
2
studies of the solution reaction between Pd2
tionrprotonation process w22,23x then gives 2 and H S. The net reaction
via the steps shown in Eq.
Ž
SH. Žm-S.Ždpm.2 and I w6,7x
Ž
Eq.
would then be realized
could couple with PdX2 dpm , and the
5
Ž ..; a deprotona-
2
2
2
Ž .
2
Ž
3
.
. Alternatively, PdX
Ž
SH.Ždpm
.
Ž
.
steps in Eq.
Ž
6
.
could constitute reaction
Ž
2
.
.
Ž
5
.
H 2S
3
PdX2
Ž
dpm
.
™ PdX
Ž
SHX.Ždpm
.
™ Pd X2ŽmyS.Ždpm.2
Ž
6
.
2
yHX
yHX
3
3
2
Adsorbed HI on alumina undergoes partial photodecomposition to form I , this then reacting with
2
y
y
I on the surface to form I , which is readily quantified
Ž .
Typically, up to 15% photodecomposition was observed with alumina isolated from reaction 2
Ž
following elution
.
by UV–Vis spectroscopy.
for
the iodide system; the strongly light-absorbing Iy3 appears to limit the extent of photodecomposition
of HI. The co-photoproduct the adsorbed hydrogen was not detected as H , but there is some
evidence for the formation of H-atoms. Based on comparison of H NMR data for CH Cl
3
Ž
.
2
1
Ž
d 5.2 in
2
2
CDCl3 , the d 5.2 signal shown in Fig. 6 for the generally less active a-alumina system is attributed
. Ž .

T.Y.H. Wong et al.rJournal of Molecular Catalysis A: Chemical 139 (1999) 159–169
169
to CDHX2
XsCl andror I , which could be formed by reaction of H-atoms with CDCl to give
Ž .
3
initially HCl and CDCl P radicals, with these then reacting with H- or I-atoms to give CDHCl or
2
2
CDCl I, respectively
Ž
note that I-atoms do not react with CHCl w24x
.
. Use of g-alumina gave no
2
3
formation of this type of species, but in hydrogenation studies using cis-cyclooctene, cyclooctane was
formed in up to ;5% yield based on 15% photodecomposition of HI.
5
. Conclusions
Concerning potential catalytic cycles for the conversion of H S to H and elemental sulfur, we
2
2
have shown that the dinuclear bridged-sulfide species Pd X2
Ž
m-S.Ždpm.2 can be formed from the
2
reaction of H S with PdX2Ždpm
liberated HX
.
on alumina, the reaction being induced by chemisorption of the
2
Ž .
Xshalide . The reaction involves activation of the H S on the alumina, and a
plausible mechanism via mercapto intermediates is presented. For the iodide system, photodecomposi-
tion of the absorbed HI can generate I , but not very efficiently, and H is not produced; there is
2
2
2
some evidence for formation of H-atoms, but not to the quantitative amount based on the I formed.
2
References
w1x T.Y.H. Wong, A.F. Barnabas, D. Sallin, B.R. James, Inorg. Chem. 34 Ž1995. 2278.
w2x A.F. Barnabas, D. Sallin, B.R. James, Can. J. Chem. 67 Ž1989. 2009.
w3x G. Besenyei, C.-L. Lee, J. Gulinski, B.R. James, D.A. Nelson, M.A. Lilga, Inorg. Chem. 26 Ž1987. 3622.
w4x B.R. James, Pure Appl. Chem. 69 Ž1997. 2213.
w5x W.Z. Weng, D. Sallin, B.R. James, to be published.
w6x T.Y.H. Wong, PhD Dissertation, University of British Columbia, Vancouver, 1996.
w7x T.Y.H. Wong, S.J. Rettig, B.R. James, to be published.
w8x C.T. Hunt, A.L. Balch, Inorg. Chem. 20 Ž1981. 2267.
w9x J.M. Jenkins, J.G. Verkade, Inorg. Synth. 11 Ž1868. 108.
w10x W.L. Steffen, G. Palenik, Inorg. Chem. 15 Ž1976. 2432.
w11x C.D. Wagner, W.M. Riggs, L.E. Davis, J.F. Moulder, in: G.E. Muilenberg ŽEd.., Handbook of X-ray Photoelectron Spectroscopy,
Perkin-Elmer, Physical Electronics Div., Eden Prairie, MN, 1979, p. 43.
w12x C.D. Wagner, W.M. Riggs, L.E. Davis, J.F. Moulder, in: G.E. Muilenberg ŽEd.., Handbook of X-ray Photoelectron Spectroscopy,
Perkin-Elmer, Physical Electronics Div., Eden Prairie, MN, 1979, p. 100.
w13x C.D. Wagner, W.M. Riggs, L.E. Davis, J.F. Moulder, in: G.E. Muilenberg ŽEd.., Handbook of X-ray Photoelectron Spectroscopy,
Perkin-Elmer, Physical Electronics Div., Eden Prairie, MN, 1979, p. 58.
w14x C.D. Wagner, W.M. Riggs, L.E. Davis, J.F. Moulder, in: G.E. Muilenberg ŽEd.., Handbook of X-ray Photoelectron Spectroscopy,
Perkin-Elmer, Physical Electronics Div., Eden Prairie, MN, 1979, p. 56.
w15x J.R. West, in: M. Grayson ŽEd.., Sulfur Recovery, Kirk-Othmer Encyclopedia of Chemical Technology, 3rd edn., Vol. 22, Wiley,
Toronto, 1983, p. 267.
w16x F. Bandermann, K.B. Harder, Int. J. Hydrogen Energy 10 Ž1985. 21.
w17x H. Knozinger, P. Ratnasamy, Catal. Rev.-Sci. Eng. 17 Ž1978. 31.
w18x L.R. Snyder, Principles of Adsorption Chromatography, Chap. 7, Marcel Dekker, New York, 1968.
w19x J.B. Peri, J. Phys. Chem. 69 Ž1965. 220.
w20x J.B. Peri, J. Phys. Chem. 69 Ž1965. 211.
w21x R.K. Oberlander, in: B.E. Leach ŽEd.., Applied Industrial Catalysis, Chap. 4, Vol. 3, Academic Press, Toronto, 1984.
w22x P.G. Jessop, C.-L. Lee, G. Rastar, B.R. James, C.J.L. Lock, R. Faggiani, Inorg. Chem. 31 Ž1992. 4601.
w23x M.C. Jennings, N.C. Payne, R.J. Puddephatt, J. Chem. Soc., Chem. Commun., Ž1986. 1809.
w24x A. Streitwieser Jr., C.H. Heathcock, Introduction to Organic Chemistry, 3rd edn., Chap. 6, Macmillan, New York, 1985.