New phosphacyclic diphosphines for rhodium-catalyzed hydroformylation

Article abstract of DOI:10.1021/om990523j

The use of phosphacyclic diphosphines based on the xanthene backbone as ligands in rhodium-catalyzed hydroformylation was studied. New phosphacyclic xantphos ligands with wide natural bite angles were synthesized, and a short, efficient route toward the synthesis of 10-chlorophenoxaphosphine and 10-chlorophenothiaphosphine was developed. The effect of the phosphacyclic moieties on the coordination chemistry in the (diphosphine)Rh(CO)₂H complexes was investigated using NMR and IR spectroscopy. Both NMR and IR spectroscopy showed that the phosphacyclic xantphos ligands exhibit an enhanced preference for diequatorial (ee) chelation compared to the diphenylphosphino-substituted parent compound. In the hydroformylation of 1-octene the introduction of the phosphacyclic moieties leads to higher reaction rates. More importantly, the dibenzophospholyl- and phenoxaphosphino-substituted xantphos ligands exhibit an unprecedented high activity and selectivity in the hydroformylation of trans 2- and 4-octene to linear nonanal. The high activities of the phosphacyclic xantphos ligands are explained by the lower phosphine basicity and the wider natural bite angles of the phosphacyclic ligands. The extraordinary high activity of the phenoxaphosphino-substituted xantphos ligand can be attributed to the 4- to 6-fold higher rate of CO dissociation compared to the other xantphos ligands. CO dissociation rates from the (diphosphine)Rh(CO)₂H complexes were determined using ¹³CO labeling in rapid-scan IR experiments.

Full text of DOI:10.1021/om990523j

Organometallics 1999, 18, 4765-4777
4765
New P h osp h a cyclic Dip h osp h in es for
Rh od iu m -Ca ta lyzed Hyd r ofor m yla tion
Lars A. van der Veen, Paul C. J . Kamer, and Piet W. N. M. van Leeuwen*
Institute of Molecular Chemistry, University of Amsterdam, Nieuwe Achtergracht 166,
1
018 WV Amsterdam, The Netherlands
Received J uly 7, 1999
The use of phosphacyclic diphosphines based on the xanthene backbone as ligands in
rhodium-catalyzed hydroformylation was studied. New phosphacyclic xantphos ligands with
wide natural bite angles were synthesized, and a short, efficient route toward the synthesis
of 10-chlorophenoxaphosphine and 10-chlorophenothiaphosphine was developed. The effect
2
of the phosphacyclic moieties on the coordination chemistry in the (diphosphine)Rh(CO) H
complexes was investigated using NMR and IR spectroscopy. Both NMR and IR spectroscopy
showed that the phosphacyclic xantphos ligands exhibit an enhanced preference for
diequatorial (ee) chelation compared to the diphenylphosphino-substituted parent compound.
In the hydroformylation of 1-octene the introduction of the phosphacyclic moieties leads to
higher reaction rates. More importantly, the dibenzophospholyl- and phenoxaphosphino-
substituted xantphos ligands exhibit an unprecedented high activity and selectivity in the
hydroformylation of trans 2- and 4-octene to linear nonanal. The high activities of the
phosphacyclic xantphos ligands are explained by the lower phosphine basicity and the wider
natural bite angles of the phosphacyclic ligands. The extraordinary high activity of the
phenoxaphosphino-substituted xantphos ligand can be attributed to the 4- to 6-fold higher
rate of CO dissociation compared to the other xantphos ligands. CO dissociation rates from
1
3
the (diphosphine)Rh(CO)
IR experiments.
2
H complexes were determined using CO labeling in rapid-scan
In tr od u ction
complex. The linear L2Rh(CO)(alkyl) species will result
in formation of the linear aldehyde, while the branched
L2Rh(CO)(alkyl) species can either give the branched
aldehyde or, via â-hydride elimination, give rise to
alkene isomerization.
Rhodium-catalyzed hydroformylation is one of the
world’s largest homogeneously catalyzed processes in
1
-3
industry.
The key issue in this process, where linear
aldehydes are the most desired products, is the control
of regioselectivity. Since the stereoelectronic properties
of ligands have a pronounced influence on the catalytic
performance, a wide variety of ligands have been
synthesized and tested in the hydroformylation reaction.
Among the different classes of compounds, phosphines
are the most widely used ligands.
The trigonal bipyramidal rhodium complexes contain-
ing two coordinated triphenylphosphine ligands can
exist in two isomeric forms, in which the phosphine
ligands coordinate in a diequatorial (ee) or a equatorial-
5
apical (ea) fashion. At room temperature these isomers
are in rapid equilibrium with an ee:ea ratio of 85:15.
Diphosphine ligands can in principle stabilize either one
of the two isomers, depending on the bridge between
the two phosphine moieties. This so-called ligand back-
bone restricts the P-M-P bite angle, and therefore the
conformation of the rhodium-diphosphine complexes,
to specific geometries.
The generally accepted, dissociative mechanism for
rhodium triphenylphosphine-catalyzed hydroformyla-
4
tion as proposed by Wilkinson is shown in Scheme 1.
The active catalyst is a trigonal bipyramidal hydrido-
rhodium complex, which usually contains two phospho-
rus donor ligands. Under moderate reaction conditions,
the rate-limiting step in the catalytic cycle is often the
displacement of a carbonyl ligand by the incoming
alkene. The regiochemistry of the reaction is determined
by the mode of alkene insertion in the hydridorhodium
(1) Parshall, G. W. Homogeneous Catalysis: The Applications and
Chemistry of Catalysis by Soluble Transition Metal Complexes; Wiley:
New York, 1980.
(
2) Beller, M.; Cornils, B.; Frohning, C. D.; Kohlpaintner, C. W. J .
Mol. Catal. A: Chem. 1995, 104, 17-85.
3) Frohning, C. D.; Kohlpaintner, C. W. In Applied Homogeneous
Casey and Whiteker have developed the concept of
the natural bite angle to calculate the preferred chela-
tion angle of diphosphine ligands using molecular
(
Catalysis with Organometallic Compounds: A Comprehensive Hand-
book in Two Volumes; Cornils, B., Herrmann, W. A., Eds.; VCH:
Weinheim, 1996; Vol. 1, Chapter 2, pp 27-104.
(4) Evans, D.; Osborn, J . A.; Wilkinson, G. J . Chem. Soc. A 1968,
(5) Brown, J . M.; Kent, A. G. J . Chem. Soc., Perkin Trans. 2 1987,
1597-1607.
3
133-3142.
1
0.1021/om990523j CCC: $18.00 © 1999 American Chemical Society
Publication on Web 10/13/1999

4
766 Organometallics, Vol. 18, No. 23, 1999
van der Veen et al.
Sch em e 1
6
mechanics. They were the first to show that the bite
angle has a great influence on the regioselectivity in the
discovered that not only the variation of the ligand
backbone but also substitution of the standard diphen-
ylphosphine moieties have a direct influence on the
ligand bite angle. Molecular mechanics calculations
revealed that the introduction of phosphacyclic moieties
on the xanthene backbone yields diphosphine ligands
with bite angles of 120-126° (Figure 2, Table 1). Herein
we describe the synthesis of new phosphacyclic ligands,
their coordination chemistry in the (diphosphine)Rh-
(CO)2H complexes, and their catalytic performance in
the hydroformylation of 1-octene. The phosphacyclic
xantphos derivatives show an increased preference for
ee coordination and display a higher activity in the
hydroformylation of 1-octene than the noncyclic parent
ligand. Moreover, DBP-xantphos (2) and POP-xant-
phos (5) give very active and selective catalysts for the
hydroformylation of internal octenes to linear nonal.10
The very high hydroformylation and isomerization
activity of ligand 5 is the direct result of an enhanced
rate of CO dissociation from the (diphosphine)Rh(CO)2H
7
rhodium-catalyzed hydroformylation. Using BISBI, an
ee coordinating diphosphine with a wide natural bite
angle of approximately 113°, a linear-to-branched (l:b)
aldehyde ratio of 66:1 was obtained, while the ea
coordinating diphosphine dppe with a narrow natural
bite angle of 85° gave a l:b ratio of only 2:1. To
investigate the exact nature of the bite angle effect, we
have developed a series of diphosphine ligands based
on the xanthene-type backbones (Figure 1).8 In this
series of ligands, a clear trend of increasing selectivity
for linear aldehyde formation with increasing natural
bite angle was observed.
,9
1
3
complex, which was measured using CO labeling in
rapid-scan IR experiments.
F igu r e 1. Xantphos family.
Resu lts
On the basis of these results it was concluded that
for inducing high regioselectivity in the rhodium-
catalyzed hydroformylation diphosphine ligands with
wide natural bite angles are required. In our effort to
develop new ligands with wide natural bite angles, we
Syn th esis. The ligands 1-3, 5, and 6 were prepared
by dilithiation of 4,5-dibromo-2,7-di-tert-butyl-9,9-di-
methylxanthene (8) with n-butyllithium at -60 °C,
followed by reaction with the corresponding chlorophos-
phines (Scheme 2). Ligand 4 was prepared by a slightly
modified procedure, in which the intermediate dilithio
species was converted into the dizinc compound before
it was made to react with 10-chlorophosphorine. This
way the occurrence of deprotonation of the methylene
group at the 9-position of the phosphorine ring is
(
6) Casey, C. P.; Whiteker, G. T. Isr. J . Chem. 1990, 30, 299-304.
(7) Casey, C. P.; Whiteker, G. T.; Melville, M. G.; Petrovich, L. M.;
Gavney, J . A., J r.; Powell, D. R. J . Am. Chem. Soc. 1992, 114, 5535-
543.
8) Kranenburg, M.; van der Burgt, Y. E. M.; Kamer, P. C. J .; van
Leeuwen, P. W. N. M. Organometallics 1995, 14, 3081-3089.
9) van der Veen, L. A.; Keeven, P. H.; Schoemaker, G. C.; Reek, J .
5
(
(
N. H.; Kamer, P. C. J .; van Leeuwen, P. W. N. M.; Lutz, M.; Spek, A.
L. Submitted.
(10) van der Veen, L. A.; Kamer, P. C. J .; van Leeuwen, P. W. N.
M. Angew. Chem., Int. Ed. 1999, 38, 336-338.

New Phosphacyclic Diphosphines
Organometallics, Vol. 18, No. 23, 1999 4767
F igu r e 2. Phosphacyclic xantphos derivatives.
Sch em e 2a
Ta ble 1. Ca lcu la ted Na tu r a l Bite An gle a n d
F lexibility Ra n ge for th e Xa n tp h os Liga n d s
flexibility
ân, deg range, deg
a
a
ligand
phosphacycle
di-t-buxantphos (1)
110.1
98-134
107-138
111-139
DBP-xantphos (2) [b,d]dibenzophosphole 121.4
TPP-xantphos (3) 1,2,3,4-tetraphenyl-
phosphole
124.3
PCP-xantphos (4) phosphorine
POP-xantphos (5) phenoxaphosphine
PSP-xantphos (6) phenothiaphosphine
POP-DPEphos (7) phenoxaphosphine
123.4
123.1
126.3
111.7
108-143
107-142
110-146
95-129
a
The natural bite angle (ân) and the flexibility range were
calculated analogously to the method used by Casey and White-
ker. ân is defined as the preferred chelation angle determined only
6
by ligand backbone constraints and not by metal valence angles.
The flexibility range is defined as the accessible range of bite
-1
angles within 3 kcal mol excess strain energy from the calculated
natural bite angle.
prevented. Ligand 7 was synthesized by dilithiation of
a
(
i) n-Buli/THF/-60 °C; (ii) (ZnCl
2 2
/Et O/)chlorophosphine/-
4
,4′-di-tert-butyldiphenyl ether (9) using n-butyllithium/
60 °C to rt; (iii) n-Buli/TMEDA/Et O/25 °C; (iv) (i-Pr) NPCl /
hexanes/0 °C; (v) PCl /110 °C.
3
2
2
2
TMEDA and successive reaction with 10-chlorophen-
1
1
oxaphosphine (10). 9-Chlorodibenzo[b,d]phosphole,
1
1
5
-chloro-1,2,3,4-tetraphenylphosphole, and 10-chloro-
10 via a modification of the procedure developed by
Teunissen and Bickelhaupt for the synthesis of 9-chlo-
rodibenzo[b,d]phosphole. This resulted in a one-step
1
2
dibenzo[b,e]phosphorine were prepared according to
literature procedures. The synthesis of compound 10 has
11
1
3
been reported before by Doak, Freedman, and Levy.
synthesis of 10 in 35% yield based on diphenyl ether.
Even higher yields were obtained using a two-step
procedure employing the intermediate 10-N,N-diisopro-
pylamino-substituted phosphacycle. Both chlorophos-
phines 10 and 11 were synthesized by dilithiation of
diphenyl ether or diphenyl sulfide using n-butyllithium/
TMEDA, followed by reaction with N,N-diisopropyl-
phosphoramidous dichloride, and refluxing the resulting
diisopropylamine substituted phosphacycle in phospho-
rus trichloride. The pure products were obtained after
They prepared compound 10 in three steps from o-
nitrophenyl phenyl ether by cyclodehydrohalogenation
in a total yield of 17%. The low total yield of this
procedure prompted us to develop a shorter and more
efficient route to chlorophosphine 10. We first prepared
(
11) Teunissen, H. T.; Bickelhaupt, F. Phosphorus, Sulfur 1996, 118,
3
09-312.
(
(
12) de Koe, P.; Bickelhaupt, F. Angew. Chem. 1967, 79, 533-534.
13) Doak, G. O.; Freedman, L. D.; Levy, J . B. J . Org. Chem. 1964,
2
9, 2382-2385.

4
768 Organometallics, Vol. 18, No. 23, 1999
van der Veen et al.
F igu r e 3. Variable-temperature 31P{ H} NMR spectra in CDCl
1
and wire models of the major and minor conformers of
3
PSP-xantphos (6).
Ta ble 2. Selected NMR Da ta for
sublimation in relatively good yields of 54% and 45%,
respectively.
(
Dip h osp h in e)Rh (CO)
2
H Com p lexesa
ligand ân, deg 1J (Rh,H), Hz 1J (Rh,P), Hz 2J av(P,H), Hz
For ligands 2, 3, 5, and 7 sharp lines were observed
in the NMR spectra at room temperature. Ligand 4
1
3
4
5
6
7
110.1
124.3
123.4
123.1
126.3
120.6
6.6
126
138
135
151
140
117
16.5
12.2
1.8
10.4
7.7
showed some line broadening in the 1H and P{ H}
31
1
<2
3.9
2.4
3.1
12.3
NMR spectra at room temperature, probably caused by
flexing of the phosphacyclic ring between two “butterfly”
conformations on the NMR time scale. Due to fluxional
behavior of the phenothiaphosphine moieties, only very
broad signals were observed in the NMR spectra of
ligand 6. At low temperature the fluxional processes
b
17.3/12.3
a
b
In C6D6 at 298 K. Only the ea complex isomer was observed.
moieties on the xanthene backbone affects the chelation
behavior of the xantphos ligands, we studied the solu-
tion structures of the (diphosphine)Rh(CO)2H com-
plexes, the catalyst resting state under hydroformyla-
1
could be frozen out completely. Both in the H and in
3
1
1
the P{ H} NMR spectrum two different conformers of
ligand 6 are observed at 223 K in a 5:2 ratio (Figure 3).
The two phosphorus atoms of the major conformer are
chemically different (∆δ ) 14.3 ppm, J (P,P) ) 6.7 Hz),
while they are equivalent in the minor isomer. The
absolute configurations of the conformers are not known,
but the NMR data and molecular mechanics suggest
that in the major conformer one of the phenothiaphos-
phine rings is pointing up, while the other ring is
pointing down (Figure 3). In the minor conformer there
must be a plane of symmetry, since both phosphorus
atoms are identical. In this structure the butterfly
conformations of the phenothiaphosphine moieties should
be mirror images and both pointing down. Above 300 K
the conformers rapidly interconvert on the NMR time
scale by flexing one of the phenothiaphosphine rings,
and only one signal is observed in the 31P{ H} NMR
spectrum at -27.9 ppm.
1
4
tion conditions, by NMR and IR spectroscopy. The
complexes were prepared in situ from Rh(CO)2(acac) and
diphosphine under an atmosphere of CO/H2 (1:1). The
characteristic heteronuclear coupling constants for the
complexes are shown in Table 2. The IR frequencies of
the absorption bands of the complexes in the carbonyl
region are summarized in Table 3.
For ligands 1 and 3-6, the formation of the (diphos-
phine)Rh(CO)2H complexes was evidenced by the ap-
3
1
1
pearance of a doublet in the P{ H} NMR spectra and
an apparent triplet of doublets in the hydride region of
the H NMR spectra. For ligand 2 the (diphosphine)-
Rh(CO)2H complex was not formed. Instead, only the
formation of the [(diphosphine)Rh(CO)2]2 complex was
1
1
(
14) van der Veen, L. A.; Boele, M. D. K.; Bregman, F. R.; Kamer,
P. C. J .; van Leeuwen, P. W. N. M.; Goubitz, K.; Fraanje, J .; Schenk,
(
Dip h osp h in e)Rh (CO)2H Com p lexes. To deter-
mine whether the introduction of the phosphacyclic
H.; Bo, C. J . Am. Chem. Soc. 1998, 120, 11616-11626.

New Phosphacyclic Diphosphines
Organometallics, Vol. 18, No. 23, 1999 4769
Ta ble 3. Selected HP -IR Da ta for
the relatively large rhodium-proton coupling constant
and the relatively small rhodium-phosphorus coupling
constants (vide infra). The trans phosphorus-proton
coupling constant of 17.3 Hz, however, is very small for
an ea complex, indicating that the complex is most likely
distorted from ideal geometry. The observation of only
the ea isomer in the NMR spectra of the complex with
ligand 7 is confirmed by the IR spectroscopy. In the IR
spectrum the intensities of the two carbonyl absorption
bands of the ee complex isomer are very low compared
a
(
Dip h osp h in e)Rh (CO)
2
H Com p lexes
1
ν2, cm-1
ν3, cm-1
ν4, cm-1
ligand
ν1, cm-
1
3
4
5
6
7
2039 (m)
2028 (m)
2037 (s)
2048 (s)
2044 (s)
2014 (vw)
1996 (w)
1974 (m)
1980 (m)
1981 (s)
1988 (s)
1985 (s)
1973 (vw)
1949 (m)
1947 (vw)
1958 (w)
2004 (vw)
2010 (w)
2000 (s)
1963 (m)
1962 (vs)
a
In cyclohexane at 353 K and 20 bar of CO/H2 (1:1) (carbonyl
region).
1
4
to those of the ea isomer.
The NMR data for the complexes of ligands 1 and 3-6
Table 2) reveal that the introduction of the phos-
(
phacycles on the xanthene backbone leads to reduced
1
rhodium-proton ( J (Rh,H)) and phosphorus-proton
2
(
J (P,H)) coupling constants and to enhanced rhodium-
1
phosphorus ( J (Rh,P)) coupling constants. From previ-
ous work it is known that (diphosphine)Rh(CO)2H
complexes consist of mixtures of ee and ea isomers that
5
,9,14
are in dynamic equilibrium.
Since the fluxional
behavior could not be frozen out at low temperature (163
K), estimations of the equilibrium compositions of these
mixtures were obtained using the averaged phosphorus-
proton coupling constants. Unfortunately, in the present
series of ligands the averaged phosphorus-proton cou-
pling constants cannot be used for quantitative estima-
tions of the ee:ea ratios, since it is not likely that these
coupling constants are exactly the same for the different
phosphacycles. The rhodium-proton coupling constants
in the (diphosphine)Rh(CO)2H complexes, however, also
prove to be indicative of the ee:ea ratio. For ea isomers
of (diphosphine)Rh(CO)2H complexes relatively large
rhodium-proton coupling constants of 11-13 Hz are
observed, while for the purely ee coordinating BISBI a
1
8
coupling constant of only 2 Hz is obtained. Moreover,
in previous studies increasing ee:ea ratios were ac-
companied by decreasing rhodium-proton coupling
F igu r e 4. Observed (top) and calculated (bottom) 31P{ H}
1
NMR spectrum of [(DBP-xantphos)Rh(CO)
2 2
] .
9
,14
constants.
observed. This dimer complex exhibits a very charac-
From the observed small rhodium-proton coupling
constants for ligands 3-6 it can be concluded that the
introduction of the phosphacycles on the xanthene
backbone shifts the equilibrium composition of the
teristic splitting pattern in the 31P{ H} NMR spectrum
1
1
consistent with an AA′A′′A′′′XX′ spin system ( J (Rh,P)
2
2
3
)
)
149 Hz, J (Rh,Rh) < 3 Hz, J (Rh,P) ) 10 Hz, J (P,P)
-16 Hz) (Figure 4).1
5,16
The IR spectrum of the
(
diphosphine)Rh(CO)2H complexes almost completely to
complex containing ligand 2 is also typical of that of a
dimer complex. It exhibits two absorption bands for
the terminal carbonyl ligands (2010 and 1980 cm ) and
two absorption bands in the region for bridging carbonyl
ligands (1797 and 1768 cm ). The dimer complex proves
to be the resting state of the catalyst, since the IR
spectra of the rhodium species present during a hydro-
formylation experiment with ligand 2 and the [(diphos-
phine)Rh(CO)2]2 complex are identical. Similar dimer
complexes were also formed for ligands 5 and 7 upon
long reaction times.
the ee isomer. Since the rhodium-proton coupling
constant of 6.6 Hz for ligand 1 corresponds to an ee:ea
isomer ratio of 7:3 (based on the phosphorus-proton
coupling constant), it can be concluded that for ligands
and 6 the ee:ea ratio is more than 8:2. The small
rhodium-proton coupling constants observed for ligands
and 5 indicate that these ligands coordinate almost
1
7
-1
-1
4
3
exclusively in the ee fashion.
The effect of the phosphacycles on the ee:ea ratio is
confirmed by the IR spectra of the (diphosphine)Rh-
CO)2H complexes. For ligand 1 two pairs of absorption
(
For ligand 7 a double doublet was observed in the 31P-
H} NMR, indicating that the two phosphorus atoms
in the (diphosphine)Rh(CO)2H complex are not identical.
The inequivalence of the phosphorus atoms indicates
that only the ea isomer of the complex is formed
employing this ligand. This is in good agreement with
bands of comparable intensities are observed, which are
assigned to the ee and the ea complex isomer.14 In the
IR spectra of the complexes of ligands 4 and 6 the
intensities of the bands belonging to the ea isomer are
considerably lower than those of the ee isomer. The IR
spectra of the complexes of ligands 3 and 5 show only
very minor contributions of the ea isomer to the complex
mixture. Furthermore, Table 3 shows that on the whole
{1
(
15) J ames, B. R.; Mahajan, D.; Rettig, S. J .; Williams, G. M.
Organometallics 1983, 2, 1452-1458.
16) Castellanos-P a´ ez, A.; Castill o´ n, S.; Claver, C.; van Leeuwen,
P. W. N. M.; de Lange, W. G. J . Organometallics 1998, 17, 2543-2552.
17) Evans, D.; Yagupsky, G.; Wilkinson, G. J . Chem. Soc. A 1968,
660-2665.
(
(
(18) Boele, M. D. K.; Bregman, F. R.; Ernsting, J . M.; Muller, F.;
van der Veen, L. A.; Elsevier: C. J . J . Organomet. Chem., in press.
2

4
770 Organometallics, Vol. 18, No. 23, 1999
van der Veen et al.
Ta ble 4. Resu lts of th e Hyd r ofor m yla tion of
Ta ble 5. Resu lts of th e Hyd r ofor m yla tion of
a
a
1
-Octen e a t 80 °C
tr a n s-2- a n d 4-Octen e
%
linear
%
l:b
% linear
ligand ân, deg l:b ratiob aldehyde % isomerb
b
tofb,c
ligand substrate t (h) conversionb ratiob aldehyde tofd
c
1
2
2
3
4
5
6
7
110.1 50.4 ( 1.7 94.3 ( 1.0
3.9 ( 1.0
244 ( 30
453 ( 86
676 ( 33
(PPh3
2
5
(PPh3
2
5
2-octene
2-octene
2-octene
4-octene 17
4-octene 17
4-octene 17
1.0
1.0
1.0
8.5
10
22
9.0
54
67
0.9
9.5
9.2
0.3
6.1
4.4
46
90
90
23
86
81
39)
65
112
2.4)
15
121.4 62.3 ( 3.4 87.8 ( 3.0 10.8 ( 3.0
121.4 47.3 ( 2.8 83.3 ( 0.1 15.0 ( 0.2
d
124.3
3.7 ( 0.2 72.6 ( 1.3
7.9 ( 0.6 1332 ( 260
4.3 ( 1.3 787 ( 190
123.4 18.1 ( 0.6 90.7 ( 1.4
123.1 66.9 ( 5.4 88.7 ( 1.9 10.0 ( 1.9 1557 ( 330
20
126.3 12.8 ( 2.7 88.0 ( 2.2
111.7 2.6 ( 0.1 70.4 ( 0.9
4.9 ( 0.7
2.3 ( 0.3
438 ( 72
100 ( 24
a
As Table 4, but at 120 °C and a constant pressure of CO/H2
b
(
1:1) of 2 bar and number of experiments ) 2. l:b ratio includes
a
c
Conditions: P(CO/H2) (1:1) ) 20 bar, ligand/Rh ) 5, substrate/
all branched aldehydes. Percentage of linear aldehyde of all
Rh ) 637, [Rh] ) 1.00 mM in toluene, number of experiments )
. In none of the experiments was hydrogenation observed. Linear
products other than octenes. dSee Table 4.
b
3
over branched ratio, percent linear aldehyde, percent isomerization
Ligand 7 is not as active as the other phosphacyclic
ligands. In this case substitution of the backbone with
the phenoxaphosphino moieties does not lead to the
large increase in activity as observed for ligand 5. The
selectivity obtained with ligand 7 is lower compared to
the parent compound DPEphos and characteristic of an
to 2-octene, and turnover frequency were determined at 20-30%
c
alkene conversion. Turnover frequency ) (mol of aldehyde)(mol
-
1
-1 d
of Rh)
h
. P(CO) ) 10 bar, P(H2) ) 50 bar.
the absorption bands of ligands 4-6 are shifted to
higher wavenumbers compared to ligand 1.
Hyd r ofor m yla tion of 1-Octen e. The activity and
selectivity of ligands 1-7 were tested in the hydrofor-
mylation of 1-octene. The reactions were carried out at
7,8
ea coordinating diphosphine.
Hyd r ofor m yla tion of In ter n a l Octen es. The high
activities of ligands 2 and 5 in the hydroformylation of
8
0 °C and 20 bar of CO/H2 (1:1) using a 1.0 mM solution
of rhodium diphosphine catalyst prepared from Rh(CO)2-
acac) and 5 equiv of ligand. The production of octene
1
-octene, combined with the very high l:b ratios and the
rates of isomerization to internal olefins, prompted us
to test these ligands in the hydroformylation of internal
octenes. The hydroformylation experiments were per-
formed at a relatively high temperature of 120 °C and
a low pressure of 2 bar CO/H2 (1:1) to enhance the rate
of isomerization and to prevent the hydroformylation
of the internal alkenes. The results of the experiments
are summarized in Table 5.
(
isomers, nonanal, and 2-methyloctanal was monitored
by gas chromatography. The results of the experiments
are summarized in Table 4. The averaged turnover
frequencies were determined at 20-30% conversion.
The introduction of the phosphacycles on the xan-
thene backbone clearly results in considerably higher
turnover frequencies compared to the diphenylphos-
phino-substituted parent ligand 1. Especially the high
activities of ligands 3 and 5 are striking. High selectivi-
ties for linear aldehyde are obtained with ligands 2 and
In the hydroformylation of trans 2-octene both ligands
2 and 5 displayed a high activity and selectivity toward
the formation of the linear aldehyde. The hydroformy-
lation of 1-octene, formed in situ by isomerization, is
highly favored over the hydroformylation of the large
excess of 2-octene, and no hydrogenation is observed.
The high reaction temperature and the low syn gas
pressure are very important for the catalytic perfor-
mance of the ligands. In a preliminary experiment with
ligand 5 at 100 °C and 10 bar CO/H2 (1:1) a lower
regioselectivity (l:b ratio was 6.3) and a lower activity
4
-6, although they are not as high as the one observed
for the parent compound 1.
Since for ligand 2 the rhodium species present under
hydroformylation conditions is the [(diphosphine)Rh-
(CO)2]2 complex (vide supra), hydroformylation experi-
ments were also performed under a 5-fold higher
hydrogen pressure. Under syn gas atmosphere the
catalytically inactive dimer complexes are in equilibri-
um with the monomeric rhodium hydride com-
(tof was 60) were obtained.
plexes.1
7,19,20
Increased hydrogen pressure will shift the
The high selectivity of ligands 2 and 5 is even more
equilibrium in favor of the mononuclear hydridorhodium
complexes, and thus this can enhance the hydroformy-
lation rate supposing the monomer-dimer equilibrium
influences the hydroformylation rate.
pronounced in the hydroformylation of trans 4-octene.
Although three consecutive double-bond isomerizations
have to precede hydroformylation, selectivities for linear
aldehyde still exceed 80%.
Kin etics of 13CO Dissocia tion fr om (Diph osph in e)-
1
3
Rh ( CO)2H Com p lexes. The hydroformylation results
unambiguously show that the introduction of the phos-
phacycles on the xanthene backbone result in an
increase of the hydroformylation rate. To determine in
which steps in the hydroformylation cycle the introduc-
tion of the phosphacycles causes the rate enhancement,
we determined the CO dissociation rates for the
2
For ligand 2, a rate enhancement of approximately
0% was observed at 50 bar of H2. The higher hydrogen
5
(diphosphine)Rh(CO) H complexes containing ligands 4
pressure also results in a slightly decreased selectivity
for linear aldehyde formation.
and 5. The CO dissociation rate constants k can be
1
13
12
obtained by exchanging CO for CO in the (diphos-
1
3
9
phine)Rh( CO)2H complexes. The CO dissociation
proceeds via a dissociative mechanism and consequently
obeys simple first-order kinetics. The rate constants k1
(
19) Brown, C. K.; Wilkinson, G. J . Chem. Soc. A 1970, 2753-2764.
(20) Moser, W. R.; Papile, C. J .; Brannon, D. A.; Duwell, R. A.;
Weininger, S. J . J . Mol. Catal. 1987, 41, 271-292.

New Phosphacyclic Diphosphines
Organometallics, Vol. 18, No. 23, 1999 4771
F igu r e 5. Representative difference IR spectrum (a) and kinetic data (b and c) for the 13CO dissociation from (POP-
1
3
13
xantphos)Rh( CO)
2
H (16- CO) in the presence of unlabeled CO at 40 °C.
1
3
Ta ble 6. Kin etics of CO Dissocia tion of
(
can, therefore, be derived from eq 1 and eq 2.
Dip h osp h in e)Rh ( CO)
2
H Com p lexesa
1
3
1
3
-
d[(diphosphine)Rh( CO) H)]/dt )
ligand
phosphacycle
phosphorine
P(CO) (bar)
r2
k1 (h-1
)
2
1
3
4
4
5
5
25
25
25
25
0.987
0.987
0.992 1188 ( 29
0.990 1171 ( 23
288 ( 8
266 ( 7
k [(diphosphine)Rh( CO) H)] (1)
1
2
1
3
phenoxaphosphine
ln[(diphosphine)Rh( CO) H)] )
2
1
3
-
k t + ln[(diphosphine)Rh( CO) H)] (2)
a
Reaction conditions: [(diphosphine)Rh(13CO) H] ) 2.00 mM
2
1
2
0
in cyclohexane, P(13CO) ) 1 bar, P(H2) ) 4 bar, T ) 40 °C,
diphosphine/Rh ) 5. Values for k1 are least-squares fit of lines
from ln[Rh] vs time over the first 15 s (ligand 4) or over the first
1
3
The (diphosphine)Rh( CO)2H complexes were pre-
pared in situ from Rh(acac)(CO)2 and diphosphine under
an atmosphere of 13CO/H2 (1:4). Complex formation was
4
s (ligand 5), calculated using TableCurve 2D.38
1
3
usually complete within 2 h. The exchange of CO for
1
2
13
CO in the (diphosphine)Rh( CO)2H complexes was
analogue. In the IR spectrum only the ee complex isomer
is visible (vide supra). The exponential decay of the
intensity of the carbonyl absorption at 1948 cm in time
monitored by rapid-scan HP IR spectroscopy at 40 °C.
The 13CO/ CO exchange was initiated by adding a large
12
-1
1
2
excess of CO (20-fold or more). The carbonyl absorp-
is displayed in Figure 5b. The solvent cyclohexane used
in the experiment is responsible for the remaining
background absorption of 0.67 au. The linear plot of the
natural logarithm of the complex concentration (ln[Rh])
vs time (Figure 5c) is obtained by converting the
background corrected absorbance in the complex con-
centration. The negative of the slope of this line is the
first-order rate constant k1 (eq 2).
-
1
tions of the complexes at approximately 1945 cm were
taken to calculate the concentrations of the complexes.
These absorption bands are assigned to one of the CO
vibrations of the ee isomer of the complexes. Represen-
tative kinetic data of the experiments with ligand 4 and
5
are shown in Figure 5, and the observed rate con-
stants, k1, are listed in Table 6.
The representative difference IR spectrum displayed
in Figure 5a for one of the experiments with (POP-
The decay of the carbonyl resonances of the (diphos-
1
3
phine)Rh( CO)2H complexes in time follows simple
first-order kinetics in all experiments. Plots of the ln-
1
3
13
xantphos)Rh( CO)2H (16- CO) clearly shows the con-
1
3
12
13
version of the CO-labeled complex in to its CO
[(diphosphine)Rh( CO)2H)] vs time are linear for at

4
772 Organometallics, Vol. 18, No. 23, 1999
van der Veen et al.
least 2 half-lives. Comparison of the rate constants k1
obtained for ligands 4 and 5 with the ones obtained
previously for other xantphos ligands shows that the
CO dissociation rate for ligand 4 is in the same range
as other ligands. The CO dissociation rate for ligand 5,
however, proves to be 4-6 times as high.
agreement with earlier studies, in which phosphacyclic
ligands have been applied in hydroformylation. Hayashi
and co-workers have reported on the hydroformylation
using dibenzophospholes and dibenzophospholyl-modi-
9
2
3-25
fied diphosphine ligands.
Two derivatives of the
BISBI ligand substituted with dibenzophospholyl groups
were developed independently at Eastman Kodak and
2
6,27
Discu ssion
Hoechst.
The phenoxaphosphine moiety has been
introduced into ligands for rhodium-catalyzed hydro-
formylation by Hobbs and Knowles. They studied the
The (diphosphine)Rh(CO)2H complexes of the phos-
phacyclic xantphos ligands are very similar to those of
ligand 1 and other xantphos-type ligands. Only the
28
asymmetric hydroformylation of vinyl acetate with
several DIOP derivatives, including one substituted
with two 2,8-dimethylphenoxaphosphino groups. The
general characteristic found for all these phosphacyclic
ligands is that they lead to more active catalysts than
their diphenylphosphino-substituted parent compounds.
To the best of our knowledge, the phosphorine and
phenothiaphosphine phosphacycles have not been em-
ployed in catalysis so far.
(
diphosphine)Rh(CO)2H complex of ligand 7 behaves
somewhat differently. Unlike the other complexes the
diphosphine)Rh(CO)2H complex of ligand 7 does not
(
display fluxional behavior at room temperature on the
NMR time scale. The introduction of the phenoxaphos-
phino substituents and the tert-butyl-groups on the
diphenyl ether backbone evidently increases the barrier
for interconversion of equatorial and apical coordinated
phosphines significantly. The relatively large rhodium-
proton coupling constants and the relatively small
rhodium-phosphorus coupling constants indicate that
only the ea complex isomer is formed for ligand 7. The
relatively small phosphorus-proton constant, however,
suggests that the complex is distorted to some extent.
Why for ligand 2 not the (diphosphine)Rh(CO)2H com-
plex but the [(diphosphine)Rh(CO)2]2 dimer complex is
formed exclusively under hydroformylation conditions
is unclear. This may be the result of the decreased
phosphine basicity (vide infra) or the wide natural bite
angle. For diphosphines with rigid backbones and wide
natural bite angles often an increased tendency for the
formation of dimeric rhodium carbonyl complexes is
observed.7
The rate enhancement that is induced by the phos-
phacycles may be explained by a combination of elec-
tronic and steric effects. Hayashi and co-workers have
suggested that the rate acceleration observed for the
dibenzophospholyl moiety can be ascribed to the higher
electronic “elasticity” of the phosphacycle compared to
2
4
diphenylphosphine. The larger electronic “elasticity”
is attributed to the extended conjugation of the ligand
π-system, which supposedly buffers the charge fluctua-
tion on rhodium along the different steps in the hydro-
formylation cycle more efficiently. Following this rea-
soning the high activity of the phenoxaphosphino moiety
can be explained in a similar fashion. Like the diben-
zophosphole ring, the phenoxaphosphine ring is more
or less flat (the dihedral angle between the two outer
,9
2
2
rings of the phosphacycle is only 15° ), leading also to
an extension of the conjugation of the ligand π-system.
A more common electronic explanation can be found
in the phosphine basicity of the different phosphacycles.
The increase of the reaction rate with decreasing
phosphine basicity is generally observed in rhodium-
catalyzed hydroformylation, also for the xantphos-type
The IR spectra of the (diphosphine)Rh(CO)2H com-
plexes show that for ligands 4-6 the carbonyl absorp-
tions are shifted to higher wavenumbers compared to
those of the complex of ligand 1. This indicates that the
incorporation of phosphorus in six-membered rings
leads to a decrease in phosphine basicity. The lower
electron density on the metal results in a decrease in
back-bonding of the rhodium to the carbonyl ligands,
and hence, higher CO stretching frequencies are ob-
served. This seems to be a common characteristic of the
medium-sized unsaturated phosphacycles, since for the
phospholes it has been reported previously that they are
1
4,20,29
ligands.
The phosphine basicity cannot explain,
however, the order of reactivity between the phosphacy-
clic ligands. Although the least basic diphosphine, ligand
5
, displays the highest activity, the order of reactivity
between ligands 3, 4, and 6 is the opposite of what can
be expected based on the carbonyl frequencies of the
2
1
weaker donor ligands than their noncyclic analogues.
(diphosphine)Rh(CO)2H complexes. The relatively mod-
The low phosphine basicity of the phenoxaphosphine
est activity of ligand 6 may be explained, however, by
the higher flexibility of the phenothiaphosphine ring
system or by partial coordination of the sulfur atoms to
ring system may be explained by the resonance struc-
tures that can be drawn for this heterocycle.22 Another
effect of the lower phosphine basicity in combination
with the wider natural bite angles is an increase of the
ee:ea isomer ratio of the (diphosphine)Rh(CO)2H com-
(
(
21) Neibecker, D.; Reau, R. J . Mol. Catal. 1989, 53, 219-227.
22) Mann, F. G.; Millar, I. T.; Powell, H. M.; Watkin, D. J . J . Chem.
1
4
plexes. This is clearly demonstrated by comparing the
NMR and IR spectra of the complexes for ligands 4-6
with those of the complex for ligand 1.
Soc., Perkin Trans. 2 1976, 1383-1385.
(23) Tanaka, M.; Ikeda, Y.; Ogata, I. Chem. Lett. 1975, 1115-1118.
(24) Hayashi, T.; Tanaka, M.; Ogata, I. J . Mol. Catal. 1979, 6, 1-9.
(25) Hayashi, T.; Tanaka, M.; Ikeda, Y.; Ogata, I. Bull. Chem. Soc.
J pn. 1979, 52, 2605-2608.
26) Devon, T. J .; Phillips, G. W.; Puckette, T. A.; Stavinoha, J . L.;
Vanderbilt, J . J . (to Eastman Kodak) U.S. Patent 5.332.846, 1994
Chem. Abstr. 1994, 121, 280879].
27) Bahrmann, H.; Lappe, P.; Herrmann, W. A.; Albanese, G. P.;
Manetsberger, R. B. (to Hoechst AG) EP 646.588, 1994 [Chem. Abstr.
(
[
(
1
995, 123, 112408d].
28) Hobbs, C. F.; Knowles, W. S. J . Org. Chem. 1981, 46, 4422-
427.
29) Unruh, J . D.; Christenson, J . R. J . Mol. Catal. 1982, 14, 19-
(
4
The enhancement of the hydroformylation rate ob-
served for ligands 2-6 compared to ligand 1 is in
(
34.

New Phosphacyclic Diphosphines
Organometallics, Vol. 18, No. 23, 1999 4773
rhodium, resulting in catalyst deactivation. The results
of CO dissociation experiments show that the high
activity of ligand 5, which places it among the most
active diphosphines known, can be attributed to the
higher rate constant for CO dissociation. The fact that
the 4-fold faster CO dissociation compared to ligand 4
only leads to approximately a 2-fold increase in hydro-
formylation rate is partially explained by the concomi-
tant higher rate of isomerization observed for ligand 5.
Furthermore, the electronic factors that promote the CO
dissociation can adversely affect the consecutive steps
in the hydroformylation cycle.
of the ligands, which are comparable, and the molecular
models of the ligands that predict that the steric
differences between these ligands are small. The low
selectivity obtained with ligand 3 is considerably out of
the range obtained for the other xantphos ligands. Since
the steric crowding in the rhodium complexes containing
ligand 3 is expected to be considerably larger than that
in the complexes of the other ligands, it is likely that at
some point in the hydroformylation cycle dissociation
of one of the phosphine arms occurs. This would explain
not only the low selectivity but also the relatively high
activity observed for ligand 3. The low activity observed
for ligand 7 is in good agreement with the observations
by NMR and IR spectroscopy that only the ea isomer of
the (diphosphine)Rh(CO)2H complex is formed for this
ligand.
It is remarkable that ligand 2 is also very active in
the hydroformylation. Although the increased activity
can also be explained by decreased phosphine basicity,
the catalyst resting state for ligand 2 is the (inactive)
Very few rhodium catalysts have been reported that
will hydroformylate internal alkenes at acceptable rates
and even fewer that will produce linear aldehydes
[
(
(diphosphine)Rh(CO)2]2 dimer complex and not the
diphosphine)Rh(CO)2H complex. The hydroformylation
experiments at increased hydrogen pressure show that
the reaction of dimer complex with hydrogen is not rate
limiting, since the 5-fold higher pressure resulted in an
increase of the activity of only 50%.
3
0-32
preferentially.
Considerable progress in this field
has been made by using bulky diphosphites as modify-
3
3,34
ing ligands.
The intrinsically low long-term stability
of phosphites, however, warrants the development of
new efficient catalysts for the hydroformylation of
internal alkenes. To the best of our knowledge ligands
The higher activity of the phosphacyclic xantphos
derivatives 2-6 compared to the parent ligand 1 is
explained by the wider natural bite angles of the former
ligands, in addition to the lower phosphine basicity.
2
and 5 constitute the first rhodium diphosphine
catalysts for the selective hydroformylation of internal
aldehydes to linear aldehydes. The excellent perfor-
mance of these ligands in the hydroformylation of
internal octenes can be explained by both the low
phosphine basicity and the wide natural bite angles of
the ligands. The low phosphine basicity leads to the
required high isomerization and hydroformylation ac-
tivity, while the large natural bite angle induces the
high selectivity for linear aldehyde formation. By choice
of temperature and pressure the proportion between
isomerization and hydroformylation activity and here-
with the selectivity for linear aldehydes can be tuned.
Contrary to the hydroformylation of 1-alkenes, the
hydroformylation of internal alkenes toward linear
aldehydes is favored by high temperature and low syn
gas pressure. The absence of high isomerization activi-
ties for ligands 4 and 6 makes these ligands less suited
for the selective hydroformylation of internal alkenes.
Especially diphosphines with bite angles close to 120°
prove to be very active ligands.7
-9
The difference in
activity between ligands 5 and 7 demonstrates that the
rigidity of the ligand backbone is crucial for obtaining
highly active catalysts. The rigid xanthene-based ligand
5
is more than 15 times as active as ligand 7, which
contains the more flexible diphenyl ether backbone.
Upon the introduction of the phosphacyclic moieties,
the very high selectivity for linear aldehyde formation,
obtained for the ligand 1, is somewhat reduced. Despite
wider natural bite angles, lower selectivities are ob-
served for ligands 2-7 compared to ligand 1. For ligands
2
and 5 the lower selectivity for linear aldehyde forma-
tion is the result of an increased isomerization rate,
while for ligands 3, 4, 6, and 7 it is caused by a reduced
l:b ratio. These results, however, are consistent with the
model that the ligand backbone determines the regio-
selectivity in the hydroformylation reaction by control-
ling the orientation of the phenyl substituents of the
diphosphines.14 In ligands 2-7 the orientation of the
phenyl substituents of the diphosphines is fixed by the
rigid phosphacycles. This results in decreased steric
hindrance for the alkene entering the rhodium coordi-
nation sphere and, hence, in lower linearities. In this
context it is important to note that the high l:b aldehyde
ratios obtained for ligands 2 and 5 are not caused by
more selective formation of the linear alkyl rhodium
species, but are the result of an increased tendency of
the branched alkyl rhodium species to form 2-octene.
Decreasing phosphine basicity results in increasing
electrophilicity of the rhodium center, which in turn
leads to a higher reactivity of the rhodium species
Con clu sion
New phosphacyclic diphosphines with wide natural
bite angles have been synthesized and a short and
efficient route toward the synthesis of 10-chlorophen-
oxaphosphine and 10-chlorophenothiaphosphine has
been developed. Upon the introduction of the phos-
phacyclic moieties to the xanthene backbone, the ee:ea
isomer ratio of the (diphosphine)Rh(CO) H complexes
2
increases. This is the combined result of the lower
phosphine basicity and the wider natural bite angles of
(30) van Leeuwen, P. W. N. M.; Roobeek, C. F. J . Organomet. Chem.
1
983, 258, 343-350.
31) van Rooy, A.; Orij, E. N.; Kamer, P. C. J .; van Leeuwen, P. W.
N. M. Organometallics 1995, 14, 34-43.
32) Breit, B.; Winde, R.; Harms, K. J . Chem. Soc., Perkin Trans. 1
(
2
9
toward CO dissociation and â-hydrogen elimination.
Both effects were observed for ligand 5.
(
1
997, 2681-2682.
(33) Billig, E.; Abatjoglou, A. G.; Bryant, D. R. (to Union Carbide)
The selectivities for the linear aldehyde observed for
ligands 2 and 4-6 in the hydroformylation of 1-octene
are approximately the same. This is in good agreement
with the natural bite angles and the flexibility ranges
EP 213, 639, 1987 [Chem. Abstr. 1987, 107, 7392r].
34) Burke, P. M.; Garner, J . M.; Tam, W.; Kreutzer, K. A.;
(
Teunissen, A. J . J . M.; Snijder, C. S.; Hansen, C. B. (to DSM/Du Pont
de Nemours) WO 97/33854, 1997 [Chem. Abstr. 1997, 127, 294939r].

4
774 Organometallics, Vol. 18, No. 23, 1999
van der Veen et al.
the ligands. Very remarkably, for the dibenzophospho-
lyl-substituted xantphos ligand the [(diphosphine)Rh-
was purchased from Cambridge Isotope Laboratories. Gas
chromatographic analyses were run on an Interscience HR GC
Mega 2 apparatus (split/splitless injector, J &W Scientific, DB1
(CO)2]2 dimer complex is formed exclusively under
3
0m column, film thickness 3.0 mm, carrier gas 70 kPa He,
FID detector) equipped with a Hewlett-Packard data system
Chrom-Card).
,7-Di-ter t-b u t yl-9,9-d im et h yl-4,5-b is(d ip h en ylp h os-
hydroformylation conditions. The phosphacyclic ligands
give very active and selective catalysts for the hydro-
formylation of 1-octene, which is also explained by the
decreased phosphine basicity and the wider natural bite
angles. The very high selectivity obtained employing the
diphenylphosphino-substituted parent ligands, however,
is slightly reduced. These results confirm the idea that
the ligand backbone determines the regioselectivity of
the reaction by controlling the orientation of the phenyl
substituents of the diphosphines.
The phenoxaphosphino-substituted xantphos ligand
proves to be one of the most active diphosphine ligands
known so far. The high activity of this ligand compared
to the other xantphos ligands can be ascribed to an
increased rate of CO dissociation from the (diphosphine)-
Rh(CO)2H complex, as determined by the rapid-scan IR
experiments. The dibenzophospholyl- and phenoxaphos-
phino-substituted xantphos ligands constitute the first
rhodium-diphosphine catalysts for the selective linear
hydroformylation of internal alkenes. This area, which
seemed hardly accessible before, may now be explored
for a whole new range of applications for rhodium-
diphosphine-catalyzed hydroformylation.
(
2
p h in o)xa n th en e (Di-t-bu xa n tp h os, 1). At -65 °C 4.9 mL
of n-butyllithium (2.5 M in hexanes, 12 mmol) was added
dropwise to a stirred solution of 2.67 g of 8 (5.6 mmol) in 75
mL of THF. The resulting beige suspension was stirred for 1
h. Next, a solution of 2.2 mL of chlorodiphenylphosphine (12
mmol) in 10 mL of hexanes was added, and the reaction
mixture was slowly warmed to room temperature overnight.
The reaction mixture was diluted with 25 mL of ethyl acetate
and hydrolyzed with 50 mL of a 1:1 mixture of brine and dilute
hydrochloric acid. The water layer was removed, and the
4
organic layer was dried with MgSO . The solvents were
removed in vacuo, and the residual yellow oil was washed with
hexanes and crystallized from ethanol/THF. Yield: 3.12 g of
1
white crystals (81%). Mp: 194-195 °C. H NMR (CDCl
3
): δ
4
J (H,H) ) 2.2 Hz, 2H; H1,8), 7.24 (m, 20H; phenyl),
3,6
) 7.38 (d,4
6.53 (bd, J (H,H) ) 2.1 Hz, 2H; H ), 1.68 (s, 6H; CH
3
), 1.11
): δ ) -16.2. C-
3
1
1
13
(s, 18H; tert-butyl). P{ H} NMR (CDCl
3
1
{
H} NMR (CDCl ): δ ) 150.7 (t, J (P,C) ) 19.1 Hz; CO),
3
2
,7
through-space P-P coupling ) 25.2 Hz, 145.4 (C ), 137.0 (t,
J (P,C) ) 13.1 Hz; PC), 134.2 (t, J (P,C) ) 20.4 Hz; PCCH),
1
1
29.7 (C3 ), 129.1 (CC ), 128.3 (d, J (P,C) ) 6.0 Hz; PCCHCH),
,6
9
4,5
28.3 (PCCHCHCH), 124.9 (t, J (P,C) ) 18.9 Hz; C ), 123.2
1
,8
9
9
(C
), 35.1 (C ), 34.7 (C(CH
3
)
3
), 32.4 (C CH
IR (KBr, cm ): 2962 (s), 1477 (m), 1424 (s), 1264 (s), 1248
m), 744 (s), 696 (s). Anal. Calcd for C47 : C, 81.71; H,
.01. Found: C, 81.78; H, 6.69.
4,5-Bis(9-d iben zo[b,d ]p h osp h olyl)-2,7-d i-ter t-bu tyl-9,9-
3 3 3
), 31.5 (C(CH ) ).
Exp er im en ta l Section
-1
(
7
H48OP
2
Com p u ta tion a l Deta ils. The molecular mechanics calcula-
tions were performed using the CAChe WorkSystem version
3
5
4
.0, on a Apple Power MacIntosh 950, equipped with two
d im eth ylxa n th en e (DBP -xa n tp h os, 2). At -65 °C 8.9 mL
of n-butyllithium (2.5 M in hexanes, 22 mmol) was added
dropwise to a stirred solution of 4.85 g of 8 (10.1 mmol) in 250
mL of THF. The resulting beige suspension was stirred for 2
h. Next, a solution of 4.86 g of 9 (22.2 mmol) in 100 mL of
THF was added, and the reaction mixture was slowly warmed
to room temperature overnight. The reaction mixture was
diluted with 250 mL of ethyl acetate and hydrolyzed with 50
mL of a 1:1 mixture of brine and dilute hydrochloric acid. The
water layer was removed, and the organic layer was dried over
CAChe CXP coprocessors. Calculations were carried out
similarly to the method described by Casey and Whiteker,6
using a Rh-P bond length of 2.315 Å. Minimizations were done
via the block-diagonal Newton-Raphson method, allowing the
structures to converge fully with a termination criterion of a
-
1
-1
rms factor of 0.0001 kcal mol
Å
or less.
Gen er a l P r oced u r e. All reactions were carried out using
standard Schlenk techniques under an atmosphere of purified
argon. Toluene and TMEDA were distilled from sodium, THF
and diethyl ether from sodium/benzophenone, and hexanes
from sodium/benzophenone/triglym. Methanol, ethanol, and
4
MgSO . The solvents were removed in vacuo, and the residual
white powder was washed with hexanes and crystallized from
dichloromethane. Yield: 4.64 g of a white microcrystalline
dichloromethane were distilled from CaH
2
. Chemicals were
purchased from Acros Chimica and Aldrich Chemical Co.
1
1
1
compound (67%). Mp: 330 °C (dec). H NMR (CDCl
3
): δ ) 8.35
9
-Chlorodibenzo[b,d]phosphole, 5-chloro-1,2,3,4-tetraphen-
3
4
1
ylphosphole, and 10-chlorodibenzo[b,e]phosphorine12 were
1
1
(dd, J (H,H) ) 7.5 Hz, J (H,H) ) 1.6 Hz, 4H; DBP-H ), 7.97
3 4 3
(
d, J (H,H) ) 7.5 Hz, 4H; DBP-H ), 7.49 (dt, J (H,H) ) 7.5
prepared according to literature procedures. Silica gel 60 (230-
4
2
3
Hz, J (H,H) ) 1.4 Hz, 4H; DBP-H ), 7.41 (dt, J (H,H) ) 7.4
4
00 mesh) purchased from Merck was used for column
4
3
4
Hz, J (H,H) ) 1.3 Hz, 4H; DBP-H ), 7.38 (d, J (H,H) ) 2.4
chromatography. Melting points were determined on a Gal-
lenkamp MFB-595 melting point apparatus in open capillaries
and are uncorrected. NMR spectra were obtained on a Bruker
1
,8
4
Hz, 2H; H ), 6.76 (dt, J (H,H) ) 2.3 Hz, J (P,H) ) 2.5 Hz, 2H;
3
,6
H
), through-space P-P coupling ) 37.8 Hz, 1.69 (s, 6H; Me),
3
1
1
1
3
3
1
13
1.12 (s, 18H; t-Bu). P{ H} NMR (CDCl
3
): δ ) -20.8. C-
): δ ) 151.0 (t, J (P,C) ) 19.6 Hz; CO), 146
(C ), 144.1 (PCC), 142.5 (t, J (P,C) ) 4.5 Hz; PC), 131.9 (t,
AMX 300 spectrometer. P and C spectra were measured
1
1
1
13
{ H} NMR (CDCl
3
H decoupled. TMS was used as a standard for H and
NMR and H PO for P NMR. Mass spectroscopy was mea-
3 4
C
2
,7
3
1
4
9
2
J (P,C) ) 26.4 Hz; DBP-C ), 129.5 (CC ), 128.7 (DBP-C ),
sured on a J EOL J MS-SX/SX102A. Elemental analyses were
carried out on an Elementar Vario EL apparatus. Standard
infrared spectra were recorded on a Nicolet 510 FT-IR spec-
trophotometer; rapid-scan (4.5 spectra per second) measure-
ments were performed using a Bio-Rad FTS-60A spectropho-
tometer. HP-IR spectra were measured using a 20 mL
homemade stainless steel autoclave equipped with mechanical
stirring and ZnS windows. Hydroformylation reactions were
carried out in a 200 mL homemade stainless steel autoclave.
3
3,6
1,8
1
1
3
3
27.5 (t, J (P,C) ) 3.0 Hz; DBP-C ), 126.5 (C ), 124.5 (C ),
4,5 1 9
24.5 (m, J (P,C) ) 25.7 Hz; C ), 121.6 (DBP-C ), 35.1 (C ),
9
36
-1
4.9 (C(CH
049 (m), 2962 (s), 1460 (m), 1426 (s), 1362 (m), 1264 (s), 1109
: C, 82.19; H,
3 3 3 3 3
) ), 33.2 (C CH ), 31.6 (C(CH ) ). IR (KBr, cm ):
(
m), 748 (s), 722 (m). Anal. Calcd for C47
H
44OP
2
6
.46. Found: C, 81.66; H, 6.40.
2
,7-Di-ter t-bu tyl-9,9-dim eth yl-4,5-bis(5-1,2,3,4-tetr aph en -
ylp h osp h olyl)xa n th en e (TP P -xa n tp h os, 3). At -65 °C 3.1
mL of n-butyllithium (2.5 M in hexanes, 7.8 mmol) was added
dropwise to a stirred solution of 1.50 g of 8 (3.1 mmol) in 50
Syn gas (CO/H
from Air Liquide. D
2
, 1:1, 99.9%) and CO (99.9%) were purchased
1
3
2
was purchased from Hoekloos. CO (99%)
(
35) CAChe Scientific Inc., 18700 N.W. Walker Rd., Building 92-
(36) Nelson, J . H.; Affandi, S.; Gray, G. A.; Alyea, E. C. Magn. Reson.
Chem. 1987, 25, 774-779.
0
1, Beaverton, OR 97006.

New Phosphacyclic Diphosphines
Organometallics, Vol. 18, No. 23, 1999 4775
mL of THF. The resulting beige suspension was stirred for 2
h. Next, a suspension of 3.30 g of 5-chloro-1,2,3,4-tetraphen-
and 36 mL of TMEDA (0.24 mol) in 600 mL of diethyl ether/
hexanes (1:2), and the reaction mixture was stirred overnight
at room temperature. The resulting orange suspension was
cooled to -60 °C, and a solution of diisopropylphosphorami-
dous dichloride in 50 mL of hexanes was added. The reaction
mixture was slowly warmed to room temperature overnight.
The solvents were removed in vacuo, and the residue was
extracted with hexanes. The combined extracts were evapo-
rated to dryness, and the residue was dissolved in 30 mL of
phosphorus trichloride and refluxed overnight. The excess
phosphorus trichloride and the formed diisopropylphosphor-
amidous dichloride were distilled off, and light yellow crystals
1
1
ylphosphole (7.81 mmol) in 75 mL of THF was added, and
the reaction mixture was slowly warmed to room temperature
overnight. The clear dark red reaction mixture was concen-
trated, and the residual brown foam was dissolved in 100 mL
of DCM and washed with 25 mL of water. The water layer
was removed, and the organic layer was dried over MgSO .
4
The solvent was removed in vacuo and the resulting orange
paste purified by flash column chromatography (40% DCM in
light petroleum ether (v:v), R
upon concentration of the eluent. Yield: 1.22 g of light yellow
f
) 0.55). The product crystallized
1
crystals (36%). Mp: 204-206 °C. H NMR (CDCl
3
): δ ) 7.26
of 10 were obtained after sublimation (110 °C/0.1 mbar).
4
1,8
3
1
(
d, J (H,H) ) 2.7 Hz, 2H; H ), 7.21 (d, J (H,H) ) 7.3 Hz, 8H;
Yield: 12.5 g (54%). Mp: 90-91 °C. H NMR (CDCl
3
): δ )
4
3,6
3
4
3
CH), 7.15 (quar., J (H,H) ) 2.4 Hz, J (P,H) ) 2.4 Hz, 2H; H ),
.10 (m, 12H; CH), 6.99 (m, 8H; CH), 6.92 (m, 4H; CH), 6.84
7.83 (ddd, J (H,H) ) 7.6 Hz, J (H,H) ) 1.7 Hz, J (P,H) ) 10.4
1,8 3 4
7
Hz, 2H; H ), 7.58 (dt, J (H,H) ) 7.8 Hz, J (H,H) ) 1.7 Hz,
3,6 3 4,5 3
3
1
1
(
m, 8H; CH), 1.42 (s, 6H; Me), 1.23 (s, 18H; t-Bu). P{ H }
2H; H ), 7.36 (d, J (H,H) ) 8.2 Hz, 2H; H ), 7.29 (tt, J (H,H)
4 4 2,7 31
1
3
1
NMR (CDCl
3
): δ ) -0.5. C{ H} NMR (CDCl
3
): δ ) 148.8,
) 7.4 Hz, J (H,H) ) 1.3 Hz, J (P,H) ) 1.3 Hz, 2H; H ). P-
1 13 1
2
1
1
1
1
3
48.6 (t, J (P,C) ) 12 Hz; CO), 145.7, 138.4, 137.2 (t, J (P,C) )
{ H} NMR (CDCl
3
): δ ) 34.5. C{ H} NMR (CDCl ): δ )
3
3,6
2
1,8
6.6 Hz; CP), 130.9, 130.7 (CH), 129.8 (t, J (P,C) ) 8.3 Hz; CH),
28.5 (CH), 128.0 (CH), 127.8 (CH), 126.6 (CH), 126.2 (CH),
153.9 (CO), 135.1 (d, J (P,C) ) 42.3 Hz, C ), 133.4 (C ), 123.8
3
2,7
4,5
1
(d, J (P,C) ) 12.8 Hz, C ), 118.2 (C ), 117.9 (d, J (P,C) )
9
23.9 (CH), 117.0 (t, J (P,C) ) 6.8 Hz), 35.5 (C ), 34.9 (C(CH
3
)
3
),
25.7 Hz, PC). Anal. Calcd for C12
Found: C, 61.19; H, 3.90.
8
H ClOP: C, 61.43; H, 3.44.
9
-1
3 2 3
1.6 (C(CH ) ), 30.9 (C CH ). IR (KBr, cm ): 3057 (m), 2962
(
6
s), 1599 (m), 1479 (m), 1427 (s), 1261 (s), 744 (m), 712 (m),
96 (s). Anal. Calcd for C79 : C, 86.62; H, 6.26. Found:
C, 86.34; H, 6.27.
,7-Di-ter t-bu t yl-9,9-d im et h yl-4,5-bis(10-d iben zo[b,e]-
2,7-Di-ter t-bu tyl-9,9-d im eth yl-4,5-bis(10-p h en oxa p h os-
p h in o)xa n th en e (P OP -xa n tp h os, 5). At -65 °C 3.7 mL of
n-butyllithium (2.5 M in hexanes, 9.2 mmol) was added
dropwise to a stirred solution of 2.0 g of 8 (4.2 mmol) in 50
mL of THF. The resulting beige suspension was stirred for 2
h. Next, a solution of 2.2 g of 10 (9.2 mmol) in 20 mL of toluene
was added, and the reaction mixture was slowly warmed to
room temperature overnight. The reaction mixture was diluted
with 70 mL of dichloromethane and hydrolyzed with 25 mL
of a 1:1 mixture of brine and dilute hydrochloric acid. The
water layer was removed, and the organic layer was dried over
H68OP
2
2
p h osp h or in o)xa n th en e (P CP -xa n tp h os, 4). At -65 °C 3.2
mL of n-butyllithium (2.5 M in hexanes, 8.0 mmol) was added
dropwise to a stirred solution of 1.75 g of 8 (3.64 mmol) in 50
mL of THF. The resulting beige suspension was stirred for 2
h and slowly warmed to 0 °C. Next 6.2 mL of zinc chloride
(1.3 M in diethyl ether, 8.0 mmol) was added. After another
1
2
hour a solution of 1.86 g of 10-chlorodibenzo[b,e]phosphorine
(8.0 mmol) in 25 mL of THF was added, and the reaction
4
MgSO . The solvent was removed in vacuo, and the resulting
mixture was slowly warmed to room temperature. After 2 days,
the reaction mixture was concentrated. The residual white
paste was dissolved in 150 mL of benzene and hydrolyzed with
beige powder was washed with hexanes and crystallized from
toluene. Yield: 2.23 g of a white microcrystalline compound
1
3
(75%). Mp: 336-338 °C. H NMR (CDCl ): δ ) 8.18 (t, J (H,H)
3
3
1
5
0 mL of a 1:1 mixture of brine and dilute hydrochloric acid.
The water layer was removed, and the organic layer was dried
over MgSO . The solvent was removed in vacuo, and the
) 7.2 Hz, J (P,H) ) 14.4 Hz, 4H; PP-H ), through-space P-P
3 4
coupling > 65 Hz, 7.40 (dt, J (H,H) ) 7.7 Hz, J (H,H) ) 1.3
3
1,8
3
4
Hz, 4H; PP-H ), 7.26 (s, 2H; H ), 7.24 (d, J (H,H) ) 8.2 Hz,
4 3 2
resulting slightly colored powder was washed with hexanes
and crystallized from ethanol/THF. Yield: 2.31 g of a white
4H; PP-H ), 7.17 (t, J (H,H) ) 7.3 Hz, 4H; PP-H ), 6.67 (s,
,6 31 1
2H; H3 ), 1.55 (s, 6H; Me), 1.10 (s, 18H; t-Bu). P{ H} NMR
1
13
1
microcrystalline compound (63%). Mp: 306 °C. H NMR
(CDCl
3
): δ ) -69.9. C{ H} NMR (CDCl ): δ ) 155.8 (PP-
CO), 149.6 (t, J (P,C) ) 21.3 Hz; CO), 145.1 (C ), 135.5 (t,
3
4
2,7
(
H
C
6
D
6
): δ ) 7.86 (m, 4H; CH), 7.49 (d, J (H,H) ) 2.4 Hz, 2H;
1
,8
4
3,6
1
3
9
), 7.20 (quar., J (H,H) ) 2.5 Hz, J (P,H) ) 2.5 Hz, 2H; H ),
.01 (m, 12H; CH), 3.83 (AB-system: δ ) 3.99, δ ) 3.67,
), 1.53 (s, 6H; Me), 1.08 (s, 18H;
): δ ) -36.2. C{ H} NMR (C
δ ) 153.6 (d, J (P,C) ) 20 Hz; CO), 146.2, 141.3 (t, J (P,C) )
.1 Hz; CCH ), 137.2 (t, J (P,C) ) 11 Hz; PC), 133.3 (t, J (P,C)
29 Hz; PCCH), 131.4 (CH), 130.7, 126.5 (CH), 124.6 (d,
J (P,C) ) 43.1 Hz; PP-C ), 130.5 (PP-C ), 128.8 (CC ), 128.5
3,6 1,8
7
Α
Β
(C ), 125.7 (t, J (P,C) ) 29.0 Hz; PP-PC), 123.8 (C ), 123.3
2 4,5 4
2
J (H
A
,H
B
) ) 16.6 Hz, 4H; CH
2
(t, J (P,C) ) 11.1 Hz; PP-C ), 118.3 (C ), 117.4 (PP-C ), 34.4
3
1
1
13
1
9
9
t-Bu). P{ H} NMR (C
6
D
6
6
D
6
):
(C ), 34.1 (C(CH
3
)
3
), 32.7 (C CH
3
), 30.9 (C(CH
3
)
3
). IR (KBr,
2
-1
cm ): 3064 (m), 2962 (s), 1583 (m), 1461 (m), 1426 (s), 1265
9
2
(s), 1224 (m), 883 (m), 758 (m). Anal. Calcd for C47
C, 78.53; H, 6.17. Found: C, 78.47; H, 5.86.
44 3 2
H O P :
)
1
4,5
9
J (P,C) ) 24 Hz; C ), 124.3 (CH), 40.5 (CH
2
), 35.7 (C ), 34.8
). IR (KBr, cm-1): 3056
w), 2964 (s), 1461 (m), 1441 (m), 1423 (s), 1263 (s), 1244 (s),
53 (s). Anal. Calcd for C49 : C, 82.32; H, 6.77. Found:
C, 81.92; H, 6.69.
10-Ch lor op h en th ia p h osp h in e (11). This compound was
prepared similarly to compound 10. Yield: 6.60 g of light
yellow crystals (45%) after sublimation (100-120 °C/0.2 mbar).
9
3 3 3 3 3
(C(CH ) ), 32.3 (C CH ), 31.6 (C(CH )
(
7
1
3
H
48OP
2
Mp: 121-135 °C. H NMR (CDCl ): δ ) 7.97 (dd, J (H,H) )
3 1,8 3
3
7.5 Hz, J (P,H) ) 10.1 Hz, 2H; H ), 7.65 (d, J (H,H) ) 7.2
4
,5
31
1
Diisop r op ylp h osp h or a m id ou s Dich lor id e. At 0 °C 75
mL of diisopropylamine (0.57 mol) is added dropwise to a
vigorously stirred solution of 25 mL of phosphorus trichloride
Hz, 2H; H ), 7.42 (m, 4H; CH). P{ H} NMR (CDCl
3
): δ )
54.4. 13C{ H} NMR (CDCl
1
): δ ) 137.8 (d, J (P,C) ) 7.6 Hz,
2
3
2
1,8
1
SC), 134.6 (d, J (P,C) ) 46.1 Hz, C ), 133.0 (d, J (P,C) ) 28.7
4
,5
3,6
3
(0.29 mol) in 250 mL of diethyl ether. The reaction mixture is
Hz, PC), 131.0 (C ), 128.2 (C ), 127.2 (d, J (P,C) ) 12.8 Hz,
2,7
warmed to room temperature and stirred for another 2.5 h.
The white suspension is filtered, and the residual salts are
washed with plenty of diethyl ether. The filtrate is concen-
C
).
2,7-Di-ter t-bu tyl-9,9-dim eth yl-4,5-bis(10-ph en oth iaph os-
p h in o)xa n th en e (P SP -xa n tp h os, 6). At -65 °C 2.6 mL of
n-butyllithium (2.5 M in hexanes, 6.4 mmol) was added
dropwise to a stirred solution of 1.40 g of 8 (2.91 mmol) in 30
mL of THF. The resulting beige suspension was stirred for 2
h. Next, a solution of 1.61 g of 11 (6.41 mmol) in 20 mL of
THF was added, and the reaction mixture was slowly warmed
to room temperature overnight. The reaction mixture was
diluted with 40 mL of benzene and hydrolyzed with 20 mL of
trated and distilled at reduced pressure. Yield: 46 g of a
1
colorless oil (79%). Bp: 86 °C (20 mbar). H NMR (CDCl
3
): δ
3
3
)
3.93 (sept., J (H,H) ) 6.3 Hz, 2H; CH), 1.28 (d, J (H,H) )
31 1
6
.9 Hz, 12H; CH
3 3
). P{ H} NMR (CDCl ): δ ) 170.2.
1
0-Ch lor op h en oxa h osp h in e (10). At 0 °C 95 mL of
n-butyllithium (2.5 M in hexanes, 0.24 mol) was added
dropwise to a solution of 16.8 g of diphenyl ether (98.7 mmol)

4
776 Organometallics, Vol. 18, No. 23, 1999
van der Veen et al.
3
a 1:1 mixture of brine and dilute hydrochloric acid. The water
layer was removed, and the organic layer was dried over
6.71 (bt, J (H,H) ) 7.2 Hz, J (P,H) ) 7.2 Hz, 4H; CH), 6.64 (m,
4H; CH), 5.97 (dd, 3J (H,H) ) 8.3 Hz, J (P,H) ) 4.1 Hz, 2H;
3
1
1
MgSO
4
. The solvents were removed in vacuo, and the residual
CH), 0.69 (s, 18H; CH
C{ H} NMR (CD
157.5 (CO), 145.0 (CC(CH
3
). P{ H} NMR (CD
3
C
6
D
5
): δ ) -65.8.
1
3
1
2
yellow oil was washed with hexanes and crystallized first from
toluene and then from THF. Yield: 0.96 g of a white micro-
3
C
6
D
5
): δ ) 157.7 (d, J (P,C) ) 14.3 Hz; CO),
2
3
)
3
), 136.9 (d, J (P,C) ) 37.8 Hz;
1
1
crystalline compound (44%). Mp: 307 °C. H NMR (223 K,
PCCH), 131.2 (CH), 130.4 (d, J (P,C) ) 26.4 Hz; PC), 128.9
3
4
2
CDCl
; CH), 7.75 (s, 1H
CH), 7.12 (m, 4HA/B; CH), 6.94 (m, 7HA/B; CH), 6.67 (t, J (H,H)
3
): δ ) 8.02 (dd, J (P,H) ) 15.6 Hz, J (H,H) ) 1.9 Hz,
(CH), 127.2 (CH), 123.9 (d, J (P,C) ) 7.6 Hz; PCCH), 118.7 (d,
2
1
H
A
A
; CH), 7.53 (m, 7HA/B; CH), 7.26 (m, 2H
A
;
3 3
J (P,C) ) 5.3 Hz; PC), 118.6 (CH), 118.0 (CH), 34.0 (C(CH ) ),
3
-1
3 3
31.4 (C(CH ) ). IR (KBr, cm ): 3066 (m), 2959 (s), 1583 (m),
3
4
)
7.4 Hz, 2H
Hz, 2H ; CH), 6.21 (t, J (H,H) ) 7.4 Hz, 2H
; Me), 7.71 (s, 2.5H ; Me), 1.46 (s, 9H ; t-Bu), 1.16 (s, 7.5H
t-Bu), 1.11 (s, 9H ; t-Bu). H NMR (333 K, CDCl ): δ ) 7.62
A
; CH), 6.30 (dd, J (H,H) ) 7.4 Hz, J (H,H) ) 2.9
1460 (s), 1474 (m), 1430 (s), 1264 (s), 1226 (s), 759 (s). Anal.
Calcd for C44
6.04.
(Di-t-bu xa n tp h os)Rh (CO)
(acac) (2.6 mg, 10 µmol) and 1 (6.2 mg, 11 µmol) in 1.0 mL of
was pressurized with 20 bar of CO/H (1:1) and stirred
for 2 h at 70 °C. After being cooled to room temperature, the
reaction mixture was depressurized, transferred into a 0.5 cm
3
A
A
; CH), 1.86 (s,
40 3 2
H O P : C, 77.86; H, 5.94. Found: C, 77.68; H,
6
H
A
B
A
B
;
1
A
3
2 2
H (12). A solution of Rh(CO) -
4
(
d, J (H,H) ) 2.3 Hz, 2H; CH), 7.36 (bm, 6H; CH), 6.88 (bs, 12
31 1
H; CH), 7.80 (s, 6H; CH
3
), 1.31 (s, 18H; t-Bu). P{ H} NMR
), -33.7 (d,
). P{ H} NMR (333 K,
): δ ) -27.9 (bs, 2P). C{ H} NMR (333 K, CDCl ): δ
C
6
D
6
2
(
223 K, CDCl
J (P,P) ) 6.7 Hz, 1 P
CDCl
3
): δ ) -19.4 (d, J (P,P) ) 9.7 Hz, 1P
A
3
1
1
A
), -34.4 (s, 0.8P
B
1
3
1
1
3
3
NMR tube, and directly analyzed at atmospheric pressure. H
2
1
6 6
NMR (C D
): δ ) 7.70 (m, 6H; CH), 7.50 (d, 4J (H,H) ) 2.4
1,8
)
C
1
(
146.1 (d, J (P,C) ) 6.0 Hz, CO), 138.0 (d, J (P,C) ) 15 Hz,
4
,5
2
), 135.5, 135.2, 130.2 (d, J (P,C) ) 14.3 Hz, CH), 129.4,
Hz, 2H; H ), 7.04 (m, 2H; CH), 6.91 (m, 12H; CH), 6.69 (td,
3
3
J (H,H) ) 6.0 Hz, 3J (P,H) ) 2.4 Hz, 2H; CH), 1.62 (s, 6H;
27.9 (CH), 127.0 (CH), 126.2 (d, J (P,C) ) 3.0 Hz, CH), 125.7
1
9
1
CH), 120.3 (d, J (P,C) ) 15 Hz, CP), 35.2 (C ), 34.7 (C(CH
3
)
3
),
C(CH
3
)
2
), 1.09 (s, 18H; C(CH
J (P,H) ) 16.5 Hz, 1H; RhH). P{ H} NMR (C
(d, J (Rh,P) ) 126.4 Hz). HP-IR (cyclohexane, carbonyl region,
cm ): 2039 (m, RhCO), 1996 (w, RhCO), 1974 (m, RhCO),
1950 (m, RhCO).
[(DBP -xa n tp h os)Rh (CO)
prepared similarly to compound 12. H NMR (C
(bm, 4H; CH), 7.62 (d, J (H,H) ) 7.8 Hz, 4H; CH), 7.33 (d,
3
)
3
31
), -8.61 (dt, J (Rh,H) ) 6.6 Hz,
9
-1
2
1
3
3.4 (C CH
3
), 31.5 (C(CH
3
)
3
). IR (KBr, cm ): 3051 (m), 2960
6
D
6
): δ ) 21.9
1
(
s), 1453 (m), 1425 (s), 1264 (s), 1244 (m), 113 (m), 752 (s).
Anal. Calcd for C47 C, 75.17; H, 5.91; S, 8.54.
Found: C, 75.06; H, 5.91; S, 8.76.
,4′-Di-ter t-bu tyld ip h en yl Eth er (9). This compound was
-
1
H
44OP
2
S
2
:
4
2 2
] (13). This compound was
3
7
1
prepared using a modified literature procedure. At -10 °C
5 mL of tert-butyl chloride (0.42 mol) is added dropwise to a
6
D
6
): δ ) 7.88
3
4
4
1,8
3
suspension of 32 g of diphenyl ether (0.19 mol) and 3.1 g of
ferric chloride (19 mmol) in 300 mL of DCM. The reaction
mixture is stirred for 2 days at 0-4 °C. The reaction mixture
is washed with dilute hydrochloric acid, water, an aqueous
saturated solution of sodium hydrogen carbonate, and brine.
J (H,H) ) 2.1 Hz, 2H; H ), 7.14 (t, J (H,H) ) 12.3 Hz, 4H;
3 3
CH), 7.01 (t, J (H,H) ) 7.2 Hz, 4H; CH), 6.74 (bd, J (P,H) )
3
,6
5.4 Hz, 2H; H ), 1.48 (s, 6H; C(CH
3
)
2
), 0.98 (s, 18H; C(CH
3
)
6
3
),
):
2
31
1
6
-8.92 (t, J (P,H) ) 12.2 Hz, 1H; RhH). P{ H} NMR (C D
δ ) 1.0 (A
J (P,P) ) -16). HP-IR (cyclohexane, carbonyl region, cm ):
2010 (m, RhCO), 1981 (s, RhCO), 1797 (m, RhCORh), 1768 (s,
RhCORh).
1
2
A′
2
XX′-system: J (Rh,P) ) 149, 2J (Rh,P) ) 10,
3
-1
The organic layer is dried over MgSO
in vacuo, and the residual light yellow oil is distilled under
4
. The solvent is removed
-3
high vacuum. The fraction with bp 110-120 °C (1 × 10 mbar)
was collected and crystallized twice from ethanol. Yield: 25 g
(TP P -xa n tp h os)Rh (CO)
prepared similarly to compound 12. H NMR (C
2
H (14). This compound was
1
1
of white crystals (47%). Mp: 41 °C. H NMR (CDCl
(
3
): δ ) 7.38
6
D
6
): δ ) 7.60
3
3
3
2
d, J (H,H) ) 8.8 Hz, 4H; CH), 7.99 (d, J (H,H) ) 8.8 Hz, 4H;
13 1
(d, J (H,H) ) 7.5 Hz, 2H; CH), 7.47 (d, J (H,H) ) 6.9 Hz, 4H;
1,8
CH), 1.41 (s, 18H; CH
3
). C{ H} NMR (CDCl
3
): δ ) 154.9 (CO),
), 31.4
). IR (KBr, cm ): 2961 (s), 2866 (m), 1503 (s), 1364
H
), 7.36 (m, 4H; CH), 7.18 (m, 6H; CH), 6.86 (m, 28H, CH),
2
1
45.6 (CC(CH
3
)
3
), 126.3 (CH), 118.0 (CH), 34.1 (C(CH
3
)
3
1.31 (s, 6H; C(CH ) ), 1.30 (s, 18H; C(CH ) ), -8.86 (t, J (P,H)
3 2 3 3
-
1
31
1
1
(
(
(
C(CH
3
)
3
6 6
) 12.2, 1H; RhH). P{ H} NMR (C D ): δ ) 24.2 (d, J (Rh,P)
1
-
m), 1296 (m), 1237 (s), 1177 (m), 1109 (m), 829 (s). GC-MS
m/z, rel intensity): 282 (M , 21), 268 (21), 267 (100), 126 (11),
) 137.9 Hz). HP-IR (cyclohexane, carbonyl region, cm ):
+
2028 (m, RhCO), 1980 (m, RhCO), 1947 (vw, RhCO).
1
12 (6), 98 (21). Anal. Calcd for C20
Found: C, 84.76; H, 9.31.
,4′-Di-t er t -b u t yl-2,2′-b is(10-p h e n oxa p h osp h in o)d i-
H
26O: C, 85.05; H, 9.29.
(P CP -xa n tp h os)Rh (CO)
2
H (15). This compound was
1
prepared similarly to compound 12. H NMR (C
6
D ): δ ) 8.19
6
1,8
4
(
1
m, 4H; CH), 7.44 (d, J (H,H) ) 2.1 Hz, 2H; H ), 7.10 (m,
4
2H; CH), 6.96 (bs, 2H; H3 ), 4.08 (s, 4H; CH
,6
), 1.39 (s, 6H;
p h en yl Eth er (P OP -DP Ep h os, 7). At 0 °C 3.5 mL of
n-butyllithium (2.5 M in hexanes, 8.7 mmol) was added
dropwise to a stirred solution of 1.02 g of 4,4′-di-tert-butyl-
diphenyl ether (3.61 mmol) and 1.3 mL of TMEDA (8.7 mmol)
in 25 mL of diethyl ether. The resulting light green suspension
was warmed to room temperature overnight. The reaction
mixture was cooled to -60 °C, and a solution of 2.03 g of 11
2
1
C(CH
3
)
2
), 1.07 (s, 18H; C(CH
3
)
3
31
), -8.92 (td, J (Rh,H) ) 3.9 Hz,
2
1
J (P,H) ) 1.8 Hz, 1H; RhH). P{ H} NMR (C
6
D
6
): δ ) -8.5
1
(
d, J (Rh,P) ) 135.0 Hz). HP-IR (cyclohexane, carbonyl region,
-1
cm ): 2037 (s, RhCO), 2004 (vw, RhCO), 1981 (s, RhCO), 1958
w, RhCO).
P OP -xa n tp h os)Rh (CO)
prepared similarly to compound 12. H NMR (C
(
(
2
H (16). This compound was
1
6
D
6
): δ ) 8.30
(8.7 mmol) in 20 mL of benzene was added. The reaction
4
1,8
(
m, 4H; CH), 7.33 (d, J (H,H) ) 2.1 Hz, 2H; H ), 7.13 (m, 6H;
mixture was slowly warmed to room temperature overnight.
The reaction mixture was diluted with 40 mL of benzene and
hydrolyzed with 20 mL of a 1:1 mixture of brine and dilute
hydrochloric acid. The water layer was removed, and the
organic layer was dried over MgSO . The solvents were
4
removed in vacuo, and the residual yellow foam was washed
with acetonitrile and crystallized from ethanol/toluene.
3
,6
CH), 7.01 (m, 6H; CH), 6.68 (bs, 2H; H ), 1.37 (s, 6H; C(CH
3
)
2
),
1
2
1
.07 (s, 18H; C(CH
10.4 Hz, 1H; RhH). P{ H} NMR (C
J (Rh,P) ) 150.7 Hz). HP-IR (cyclohexane, carbonyl region,
3 3
) ), -9.14 (dt, J (Rh,H) ) 2.4 Hz, J (P,H)
3
1
1
)
6 6
D ): δ ) -29.0 (d,
1
-
1
cm ): 2048 (s, RhCO), 1988 (s, RhCO).
1
3
13
13
1
(P OP -xa n tp h os)Rh ( CO)(CO) H (16- CO). HP C{ H}
1
1
2
Yield: 0.68 g of white crystals (28%). Mp: 187 °C. H NMR
6 6
NMR (323 K, C D ): δ ) 198.7 (dt, J (Rh,C) ) 65.8 Hz, J (P,C)
3
13
(CD
3
C
6
D
5
): δ ) 7.55 (bt, J (H,H) ) 8.1 Hz, J (P,H) ) 8.1 Hz,
) 16.6 Hz; Rh CO).
3
4
4
H; CH), 6.98 (dd, J (H,H) ) 8.2 Hz, J (H,H) ) 1.2 Hz, 4H;
2
(P SP -xa n tp h os)Rh (CO) H (17). This compound was
prepared similarly to compound 12. H NMR (C
3
4
1
CH), 6.87 (dt, J (H,H) ) 7.7 Hz, J (H,H) ) 1.6 Hz, 4H; CH),
6
D
6
): δ ) 8.07
3
3
(
dquar., J (H,H) ) 5.4 Hz, J (P,H) ) 5.4 Hz, 4H; PCCH), 7.44
1,8
(
d, 4J (H,H) ) 2.1 Hz, 2H; H ), 7.31 (m, 6H; CH), 6.95 (t,
J (H,H) ) 7.1 Hz, 4H; CH), 6.87 (dt, J (H,H) ) 7.5 Hz, J (H,H)
) 1.5 Hz, 4H; CH), -9.18 (dt, J (Rh,H) ) 3.1 Hz, J (P,H) )
(
37) Nowick, J . S.; Ballester, P.; Ebmeyer, F.; Rebek, J ., J r. J . Am.
Chem. Soc. 1990, 112, 8902-8906.
38) TableCurve, J andel Scientific.
3
3
4
1
2
(

New Phosphacyclic Diphosphines
Organometallics, Vol. 18, No. 23, 1999 4777
pressed into the autoclave with 2 or 20 bar CO/H (1:1). For
1-octene the reaction was stopped at the 20% conversion level
by adding 0.25 mL of tri-n-butyl phosphite and cooling on ice.
For 2-octene a sample of the reaction mixture was taken after
1 h and quenched with tri-n-butyl phosphite. For 4-octene the
reaction was stopped after 17 h by cooling on ice. Samples of
the reaction mixture were analyzed by temperature-controlled
gas chromatography.
.7 Hz, 1H; RhH). 31P{ H} NMR (C
1
D
): δ ) -1.2 (d, J (Rh,P)
1
7
6
6
2
-
1
)
139.8 Hz). HP-IR (cyclohexane, carbonyl region, cm ):
2
044 (s, RhCO), 2010 (w, RhCO), 1985 (s, RhCO), 1963 (m,
RhCO).
P OP -DP Ep h os)Rh (CO)
prepared similarly to compound 12. HP H NMR (C
(
2
H (18). This compound was
1
6
6
D ): δ )
3
8
.68 (dd, J (P,H) ) 13.0 Hz, J (H,H) ) 6.5 Hz, 2H; CH), 8.58
3
(
dd, J (P,H) ) 13.8 Hz, J (H,H) ) 7.4 Hz, 2H; CH), 7.94 (dd,
3
3
J (P,H) ) 16.1 Hz, J (H,H) ) 7.3 Hz, 2H; CH), 7.68 (bs, 2H;
HP F T-IR Exp er im en ts. In a typical experiment the HP
IR autoclave was filled with 2-5 equiv of ligand, 4 mg of Rh-
3
CH), 6.96 (m, 4H; CH), 6.85 (t, J (H,H) ) 8.1 Hz, 4H; CH),
3
6
(
.79 (dd, J (P,H) ) 13.6 Hz, J (H,H) ) 7.2 Hz, 4H; CH), 5.52
(CO)
2
(acac), and 15 mL of cyclohexane. The autoclave was
(1:1), pressurized to
4
dd, J (P,H) ) 8.5 Hz, J (H,H) ) 4.7 Hz, 2H; CH), 0.88 (s, 18H;
purged three times with 15 bar CO/H
2
1
2
t-Bu), -8.63 (td, J (Rh,H) ) 12.3 Hz, J (P,H) ) 17.3 Hz,
approximately 18 bar, and heated to 80 °C. Catalyst formation
was monitored in time by FT-IR and was usually completed
within 1 h.
Ra p id -Sca n HP F T-IR Exp er im en ts. In a typical experi-
ment the HP IR autoclave was filled with 75 µmol of ligand,
3
1
1
2
J (P,H) ) 12.3 Hz, 1H; RhH). HP P{ H} NMR (C
6
D
6
): δ )
1
1
-
8.7 (d, J (Rh,P) ) 116.5 Hz), -10.3 (d, J (Rh,P) ) 117.3 Hz).
-1
HP-IR (cyclohexane, carbonyl region, cm ): 2014 (vw, RhCO),
2
000 (s, RhCO), 1973 (vw, RhCO), 1962 (vs, RhCO).
Hyd r ofor m yla tion Exp er im en ts. Hydroformylation reac-
30 µmol of Rh(CO)
autoclave was purged four times with 15 bar of H
to 3 bar with 13CO, and consequently to 10 bar with H
2
(acac), and 15 mL of cyclohexane. The
, pressurized
, and
tions were carried out in an autoclave, equipped with a glass
inner beaker, a substrate inlet vessel, a liquid sampling valve,
and a magnetic stirring rod. The temperature was controlled
by an electronic heating mantle. In a typical experiment 50
µmol of ligand was placed in the autoclave, and the system
was evacuated and heated to 50 °C. After 0.5 h the autoclave
2
2
heated at 40 °C for approximately 2 h. Next a gas reservoir
containing 50 bar of unlabeled CO is opened, which resulted
in a total pressure in the autoclave of 30 bar. At the same
-
1
time the rapid-scan FT-IR experiment (270 scans min for a
total acquisition time of 2.5 min) was started.
was filled with CO/H
2
(1:1) and a solution of 10 µmol of Rh-
(
CO) (dpm) in 8.5 mL of toluene. The autoclave was pressur-
2
ized to the appropriate pressure (2 or 16 bar), heated to the
appropriate temperature (80 or 120 °C), and stirred for 1.5 h
to form the active catalyst. Then 1.0 mL of substrate (filtered
over neutral activated alumina to remove peroxide impurities)
and 0.5 mL of the internal standard n-decane were placed in
Ackn owledgm en t. Financial support from the Tech-
nology Foundation (STW) of The Netherlands Organiza-
tion for Scientific research (NWO) is gratefully acknowl-
edged.
the substrate vessel, purged with 10 bar CO/H
2
(1:1), and
OM990523J