Novel amphiphilic diphosphines: Synthesis, x-ray structure, rhodium complexes, use in hydroformylation, and rhodium recycling

Article abstract of DOI:10.1021/om970068w

Three novel amphiphilic diphosphines have been synthesized: bis[2-(phenyl(3-pyridyl)-phosphino)ethyl] ether (POPpy) and bis[2-((4-((diethylammo)methyl)phenyl)phenylphosphino)-ethyl] ether (POPam), based on bis[2-(diphenylphosphino)ethyl] ether (POP), and 4,6-bis[bis(4-((diethylaniino)methyl)phenyl)phosphino]-10,10- dimethylxanthene(xantham), based on 4,6-bis(diphenylphosphino)-10,10-dimethylxanthene (Xantphos). The crystal structure of xantham has been determined. Solution structures of rhodium xantham complexes have been studied using NMR and IR spectroscopy. When POPam is used in the hydroformylation of oct-1-ene, a linear/branched ratio (1/b) of 7.3 is achieved (88% nonanal) without isomerization to oct-2-ene. The use of POPpy results in an activity that is twice as high, a higher 1/b-ratio of 9, and 0.7% isomerization. When xantham is used in the hydroformylation of hex-1-ene, oct-1-ene, and dodec-1-ene (80 °C, 20 bar of syngas, toluene), 1/b-ratios of 50 have been achieved (94% nonanal) together with 4% of isomerized octenes. The pH-dependent distribution characteristics of the free ligands have been determined, and it is shown that POPpy, POPam, and xantham can be extracted for 40% from a Et₂O solution into an aqueous solution of H₂SO₄ at pH values of, respectively, 2.5, 5, and 5. Rhodium recycling experiments using POPam and xantham show that rhodium can be extracted into an aqueous layer at pH 5-5.5 for more than 99.95%, as determined by ICP-AES analysis, which allows complete separation of the product aldehydes and catalyst. After neutralization of the aqueous phase containing protonated POPam with an aqueous solution of NaHCO₃, only 65% of the rhodium can be re-extracted into fresh toluene. In the case of xantham, 98% of the rhodium is re-extracted. Pressurizing the recovered rhodium and excess xantham to 20 bar at 80 °C resulted in regeneration of the original active hydride complex, as judged by the retention of activity of up to 86%. POPpy has proved inappropriate for the recycling of rhodium.

Full text of DOI:10.1021/om970068w

Organometallics 1997, 16, 3027-3037
3027
Novel Am p h ip h ilic Dip h osp h in es: Syn th esis, X-r a y
Str u ctu r e, Rh od iu m Com p lexes, Use in
Hyd r ofor m yla tion , a n d Rh od iu m Recyclin g
Armin Buhling, Paul C. J . Kamer, and Piet W. N. M. van Leeuwen*
University of Amsterdam, Van’t Hoff Research Institute, Inorganic Chemistry,
Nieuwe Achtergracht 166, 1018 WV Amsterdam, The Netherlands
J aap W. Elgersma
University of Amsterdam, Amsterdam Institute for Molecular Studies, Laboratory for
Analytical Chemistry, Nieuwe Achtergracht 166, 1018 WV Amsterdam, The Netherlands
Kees Goubitz and J an Fraanje
University of Amsterdam, Amsterdam Institute for Molecular Studies, Department of
Crystallography, Nieuwe Achtergracht 166, 1018 WV Amsterdam, The Netherlands
Received J anuary 30, 1997X
Three novel amphiphilic diphosphines have been synthesized: bis[2-(phenyl(3-pyridyl)-
phosphino)ethyl] ether (POPpy) and bis[2-((4-((diethylamino)methyl)phenyl)phenylphosphino)-
ethyl] ether (POPam), based on bis[2-(diphenylphosphino)ethyl] ether (POP), and 4,6-
bis[bis(4-((diethylamino)methyl)phenyl)phosphino]-10,10-dimethylxanthene (xantham), based
on 4,6-bis(diphenylphosphino)-10,10-dimethylxanthene (Xantphos). The crystal structure
of xantham has been determined. Solution structures of rhodium xantham complexes have
been studied using NMR and IR spectroscopy. When POPam is used in the hydroformyl-
ation of oct-1-ene, a linear/branched ratio (l/b) of 7.3 is achieved (88% nonanal) without
isomerization to oct-2-ene. The use of POPpy results in an activity that is twice as high, a
higher l/b-ratio of 9, and 0.7% isomerization. When xantham is used in the hydroformylation
of hex-1-ene, oct-1-ene, and dodec-1-ene (80 °C, 20 bar of syngas, toluene), l/b-ratios of 50
have been achieved (94% nonanal) together with 4% of isomerized octenes. The pH-
dependent distribution characteristics of the free ligands have been determined, and it is
shown that POPpy, POPam, and xantham can be extracted for 40% from a Et2O solution
into an aqueous solution of H2SO4 at pH values of, respectively, 2.5, 5, and 5. Rhodium
recycling experiments using POPam and xantham show that rhodium can be extracted into
an aqueous layer at pH 5-5.5 for more than 99.95%, as determined by ICP-AES analysis,
which allows complete separation of the product aldehydes and catalyst. After neutralization
of the aqueous phase containing protonated POPam with an aqueous solution of NaHCO3,
only 65% of the rhodium can be re-extracted into fresh toluene. In the case of xantham,
98% of the rhodium is re-extracted. Pressurizing the recovered rhodium and excess xantham
to 20 bar at 80 °C resulted in regeneration of the original active hydride complex, as judged
by the retention of activity of up to 86%. POPpy has proved inappropriate for the recycling
of rhodium.
In tr od u ction
First of all, a highly efficient method for recovering the
expensive rhodium catalyst from the high-boiling alde-
hydes is imperative. Secondly, a highly selective rho-
dium catalyst is required that can limit the formation
of branched aldehydes and suppress isomerization of the
feedstock alkenes. Ideally, isomerization should be
absent since the isomerized alkenes build up in the
reactor as a consequence of feed recycle.
The concept of biphasic catalysis has emerged as an
important method for achieving easy separation and
reuse of homogeneous metal catalysts.9-15 The solubil-
ity of higher alkenes in water, however, is limited, which
Since the discovery that rhodium phosphine
complexes1-3 enable hydroformylation of propene at
much lower temperatures and pressures than cobalt,4
they have replaced cobalt catalysts in nearly all major
plants producing lower aldehydes, despite the higher
price of rhodium.5-7 Apart from the isononyl alcohol
process by Mitsubishi Chemical Company,8 no major
commercial process for the hydroformylation of longer
chain alkenes exploits rhodium phosphine complexes.
The development of such a process faces two challenges.
X Abstract published in Advance ACS Abstracts, J une 1, 1997.
(1) Osborn, J . A.; Young, J . F.; Wilkinson, G. J . Chem. Soc. 1965,
(5) Bahrmann, H.; Bach, H. W. In Ullmann’s Encyclopedia of
Industrial Chemistry; Verlagsgesellschaft mbH: Weinheim, Germany,
1991; Vol. A18 pp 321-326.
(6) Cornils, B. In Hydroformylation: New Syntheses with Carbon
Monoxide; J . Falbe, Ed.; Springer Verlag: Berlin, 1980.
(7) Weissermel, K.; Arpe, H.-J . Industrielle Organische Chemie;
Verlagsgesellschaft mbH: Weinheim, Germany, 1988.
(8) Onada, T. CHEMTECH 1993, 23(Sept), 34.
17.
(2) Slaugh, L. H.; Mullineaux, R. D. (Shell) U.S. Patent 3 239 566,
1966.
(3) Evans, D.; Osborn, J . A.; Wilkinson, G. J . J . Chem. Soc. A 1968,
3133.
(4) Yagupski, G.; Brown, C. K.; Wilkinson, G. J . J . Chem. Soc. A
1970, 1392.
S0276-7333(97)00068-X CCC: $14.00 © 1997 American Chemical Society

3028 Organometallics, Vol. 16, No. 13, 1997
Buhling et al.
would result in low reaction rates due to the approxi-
mate first-order dependency of the hydroformylation
reaction in alkene concentration.16-18 We focused on a
known catalyst recycling strategy, analogous to that of
the Kuhlmann process,6 using amphiphilic phos-
phines.19-23 Thus, the hydroformylation is performed
under conventional homogeneous conditions in an or-
ganic medium at high alkene concentrations. After
(partial) conversion of the alkene, the catalyst is ex-
tracted into an acidic aqueous phase, thus allowing
separation of the organic products, and re-extracted into
a fresh organic phase upon neutralization of the aqueous
phase. Previously, we investigated this system using
a series of functionalized triphenylphosphines24,25 and
succeeded in the almost quantitative recovery of rho-
dium. Amphiphilic ligands based on 2,2′-bis(diphenyl-
phosphinomethyl)-1,1′-biphenyl (BISBI) formed highly
active and selective rhodium catalysts, but the rhodium
recycling experiments were less successful.26
very flexible chelating POP ligand also provides an
improved linearity in the hydroformylation of 1-decene,
1,9-decadiene, and 1,5-hexadiene when compared to
PPh3.32 Moreover, the observed isomerization of the
starting alkenes is extremely low. Brown et al. sug-
gested that POP has the capacity to coordinate to the
active rhodium hydride in both a bis(equatorial) and
equatorial-axial fashion.33 1,5-Bis(diphenylphosphino)-
pentane can easily lead to cyclometallation giving a
Rh(III) complex, as reported by Shaw et al. for 1,5-bis-
[bis(tert-butyl)phosphino]pentane,34,35 which is pre-
vented by the central oxygen atom in POP and Xant-
phos.
Phosphino ethers are potentially both oxygen- and
phosphorus-donor ligands. Complexes have been re-
ported in which POP only functions as a cis-coordinated
phosphorus bidentate ligand, such as in [(POP)-
NiCl2],36-38 [(POP)Rh(COD)][ClO4],39 and [(POP)Pd-
(CH3CN)2][BF4]2.40 However, POP can also serve as a
terdentate ligand with additional oxygen coordination,
as found in [(POP)Pd{η1-(σ-alkenyl)}][PF6],41 [(POP)Pd-
(PEt3)][BF4]2,40 [(POP)Rh(H)2(MeOH)][BF4],33 and [(POP)-
Rh(CO)][PF6].42 All of these complexes contain metal
cations. In neutral, stable rhodium phosphine hydride
species, we expect that coordination of oxygen is not
significant, although it might play a role in transient
species.
As part of our search for more effective catalysts, we
decided on functionalisation of two other interesting
ligands for the hydroformylation reaction, namely Xant-
phos, developed in our research group,27 and bis[2-
(diphenylphosphino)ethyl] ether (POP). The rigid Xant-
Here, we present a study of the novel amphiphilic
ligands xantham, POPpy, and POPam. Their coordina-
phos ligand induces the highest linearity in the
hydroformylation reaction known so far, while the
degree of isomerization is low. Extremely high selectiv-
ity is also obtained with diphosphite ligands,28-31 but
these systems are not suitable for aqueous work-up. The
(9) Herrmann, W. A.; Kohlpaintner, C. W. Angew. Chem., Int. Ed.
Engl. 1993, 32, 1524.
(10) Herrmann, W. A.; Kohlpaintner, C. W.; Manetsberger, R. B.;
Bahrmann, H.; Kottmann, H. J . Mol. Catal. 1995, 97, 65.
(11) Darensbourg, D. J .; J oo´, F.; Kannisto, M.; Katho´, A.; Reiben-
spies, J . H.; Daigle, D. J . Inorg. Chem. 1994, 33, 200.
(12) Fell, B.; Papadogianakis, G. J . Prakt. Chem. 1994, 336, 591.
(13) Fell, B.; Papadogianakis, G.; Konkol, W.; Weber, J .; Bahrmann,
H. J . Prakt. Chem. 1993, 335, 75.
tion behavior in some rhodium complexes is discussed.
These ligands are active and highly selective in the
hydroformylation reaction. Additionally, xantham en-
ables an efficient rhodium recycling.
(14) Ding, H.; Hanson, B. E.; Bartik, T.; Bartik, B. Organometallics
1994, 13, 3761.
(15) Horva´th, I. T.; Ra´bai, J . Science 1994, 266, 72.
(16) Cavalieri d’Oro, P.; Raimondi, L.; Pagani, G.; Montrasi, G.;
Gregorio, G.; Andreetta, A. Chim. Ind. (Milan) 1980, 62, 572.
(17) van Rooy, A.; de Bruijn, N. H.; Roobeek, K. F.; Kamer, P. C. J .;
van Leeuwen, P. W. N. M. J . Organomet. Chem. 1995, 507, 69.
(18) Gerritsen, L. A.; Klut, W.; Vreugdenhil, M. H.; Scholten, J . J .
F. J . Mol. Catal. 1980, 9, 265.
(28) van Rooy, A.; Kamer, P. C. J .; van Leeuwen, P. W. N. M.;
Goubitz, K.; Fraanje, J .; Veldman, N.; Spek, A. L. Organometallics
1996, 15, 835.
(29) Billig, E.; Abatjoglou, A.; Bryant, D. R. (Union Carbide) U.S.
Patent 4 769 49 8, 1988.
(19) To´th, I.; Hanson, B. E.; Davis, M. E. J . Organomet. Chem. 1990,
396, 363.
(20) Bayo´n, J . C.; Real, J .; Claver, C.; Polo, A.; Ruiz, A. J . Chem.
Soc., Chem Commun. 1989, 1056.
(21) Kurtev, K.; Ribola, D.; J ones, R. A.; Cole-Hamilton, D. J .;
Wilkinson, G. J . J . Chem. Soc., Dalton Trans. 1980, 55.
(22) Nagel, U. Angew. Chem. 1984, 96, 425.
(30) Cuny, G. D.; Buchwald, S. L. J . Am. Chem. Soc. 1992, 155, 2066.
(31) J ohnson, J . R.; Cuny, G. D.; Buchwald, S. L. Angew. Chem.,
Int. Ed. Engl. 1995, 34, 1760.
(32) Goodall, B. L.; Grotenhuis, P. A. M.; van Leeuwen, P. W. N. M.
GB Pat. Appl. 8722460, 1988.
(33) Brown, J . M.; Cook, S. J .; Kent, A. G. Tetrahedron 1986, 42,
5104.
(23) Andreetta, A.; Barberis, G.; Gregorio, G. Chim. Ind. (Milan)
1978, 60, 887.
(24) Buhling, A.; Kamer, P. C. J .; van Leeuwen, P. W. N. M. J . Mol.
Catal. 1995, 98, 69.
(25) Buhling, A.; Elgersma, J . W.; Kamer, P. C. J .; van Leeuwen,
P. W. N. M. J . Mol. Catal. A: Chem. 1997, 116, 297.
(26) Buhling, A.; Nkrumah, S.; Elgersma, J . W.; Kamer, P. C. J .;
van Leeuwen, P. W. N. M. J . Chem. Soc., Dalton Trans. 1996, 2143.
(27) Kranenburg, M.; van der Burgt, Y. E. M.; Kamer, P. C. J .; van
Leeuwen, P. W. N. M.; Goubitz, K.; Fraanje, J . Organometallics 1995,
14, 3081.
(34) Shaw, B. L. Adv. Chem. Ser. 1982, 196, 101.
(35) Crocker, C.; Errington, R. J .; Markham, R.; Moulton, C. L.;
Odell, K. J .; Shaw, B. L. J . Am. Chem. Soc. 1980, 102, 4373.
(36) Dapporto, P.; Sacconi, L. J . Chem. Soc. A 1971, 1914.
(37) Dapporto, P.; Sacconi, L. J . Am. Chem. Soc. 1970, 92, 4133.
(38) Greene, P. T.; Sacconi, L. J . Chem. Soc. A 1970, 866.
(39) Thewissen, D. H. M. W.; Timmer, K.; Noltes, J . G.; Marsman,
J . W.; Laine, R. M. Inorg. Chim. Acta 1985, 97, 143.
(40) Steffey, B. D.; Miedaner, A.; Maciejewski-Farmer, M. L.;
Bernatis, P. R.; Herring, A. M.; Allured, V. S.; Carperos, V.; DuBois,
D. L. Organometallics 1994, 13, 4844.

Novel Amphiphilic Diphosphines
Sch em e 1. Syn th esis of Liga n d s
Organometallics, Vol. 16, No. 13, 1997 3029
F igu r e 1. 31P{1H} NMR: spectrum for HRh(4)(CO)(PPh3).
found when we used a large excess of phenol and higher
temperatures (80-90 °C). The excess phenol could be
easily removed by azeotropic distillation with mesitylene
providing pure phosphinite 9 in 98% yield. 4,6-
Dibromoxanthene 10 was prepared to cleanly generate
the dilithio compound with 2 equiv of n-BuLi.27,48-51
This compound did not react with the coproduct 1-bro-
mobutane below 20 °C and smoothly substituted the
phenoxy group in 9 at lower temperatures.
Rh od iu m Com p lexes. To establish the coordination
chemistry of the novel ligands and the possible forma-
tion of P-N chelated complexes, solution structures of
rhodium complexes were studied by means of NMR and
IR spectroscopy.45,52 Exchange of 1 equiv of xantham
with HRh(CO)(PPh3)3 results in quantitative formation
of HRh(xantham)(CO)(PPh3). The 31P{1H} NMR spec-
trum, as shown in Figure 1, is typical of a structure in
which both PPh3 and xantham are equatorially coordi-
nated, and it reveals the expected Rh-P and P-P
coupling constants.27 The additional fine structure in
the spectrum is caused by second-order effects owing
to the relatively small difference in chemical shifts of
PPh3 and xantham, as confirmed by simulation.
The relatively small H-P coupling constants in the
1H NMR spectrum indicate coordination of the phos-
phines in a tris(equatorial) fashion. For the analogous
Xantphos complex, it was shown that the rigid confor-
mation of Xantphos, which is coordinated in the equato-
rial plane, leads to the inequivalence of the methyl
groups in the xanthene bridge and the ortho hydrogens
of the phenyl groups on the diphenylphosphino moieties.
This was also observed for the xantham complex. In
addition, the protons in the benzylic diethylamino
groups are diastereotopic. The methyl protons of the
ethyl groups give a multiplet, while the methylene
protons are split into two overlapping multiplets. The
methylene protons in the benzylic amine bridges give
two separate resonances: an AB pattern and a singlet,
both with the intensity of four protons. The AB pattern,
which reveals geminal coupling constants of 14.3 Hz,
can be accounted for by the “axial” positions that two
of the phenyl groups assume in the rhodium complex,
according to molecular modeling. This renders the two
protons on the benzylic carbon diastereotopic. Formally,
the benzylic hydrogens, situated on the “equatorial”
Resu lts a n d Discu ssion
Liga n d Syn th esis. POPpy was synthesized accord-
ing to Scheme 1. 3-Pyridyllithium was synthesized43,44
and isolated as a yellow solid. Direct addition to PhPCl2
gave a mixture of PhPCl2, Ph(3-pyridyl)PCl (4), and
PPh(3-pyridyl)2. The use of a pyridylzinc intermediate45
gave a similar isolated yield, though this procedure was
more laborious. NaPPh(3-pyridyl), obtained by the slow
reduction45 of 4, could be used in situ to react with bis-
(2-chloro)ethyl ether. Higher yields were obtained upon
isolation and subsequent deprotonation of the secondary
phosphine 5. POPam was readily synthesized from the
secondary phosphine oxide 6 and bis(2-iodo)ethyl ether.
Dioxide 7 was readily reduced, and the pure diastereo-
meric mixture of POPam was obtained, albeit in low
yield (42%) due to the formation of many decomposition
products.
The synthesis of xantham starts with the highly
water-sensitive phosphinous amide 8. The common
procedure to cleave the P-N bond with dry HCl to give
the corresponding chlorophosphine46 would lead to
protonation of the benzylic amino groups by HCl.
Therefore, we replaced the diethylamide group by an
aryloxy group, which is also a potential leaving group.
This was done with phenol in the absence of the usually
required equivalent of acid.47 Useful reaction rates were
(47) J uge, S.; Stephan, M.; Laffitte, J . A.; Genet, J . P. Tetrahedron
Lett. 1990, 31, 6357.
(48) Hillebrand, S.; Bruckmann, J .; Kru¨ger, C.; Haenel, M. W.
Tetrahedron Lett. 1995, 36, 75.
(49) Haenel, M. W.; Fieseler, H.; J akubik, D.; Gabor, B.; Goddard,
R.; Kru¨ger, C. Tetrahedron Lett. 1993, 34, 2107.
(50) Haenel, M. W.; J akubik, D.; Rothenberger, E.; Schroth, G.
Chem. Ber. 1991, 124, 1705.
(51) Schwarz, E. B.; Knobler, C. B. J . Am. Chem. Soc. 1992, 114,
10775.
(41) Kataoke, Y.; Tsuji, Y.; Matsumoto, O.; Ohashi, M.; Yamagata,
T.; Tani, K. J . Chem. Soc., Chem. Commun. 1995, 2099.
(42) Alcock, N. W.; Brown, J . M.; J effery, J . C. J . Chem. Soc., Dalton
Trans. 1976, 583.
(43) Rubinsztajn, S.; Zeldin, M.; Fife, W. K. Synth. React. Inorg. Met.-
Org. Chem. 1990, 20, 495.
(44) Parham, W. E.; Piccirilli, R. M. J . Org. Chem. 1977, 42, 257.
(45) Budzelaar, P. H. M.; Frijns, J . H. G. Organometallics 1990, 9,
1222.
(46) To´th, I.; Hanson, B. E.; Davis, M. E. Organometallics 1990, 9,
675.
(52) Lo Schiavo, S.; Rotondo, E.; Bruno, G.; Faraone, F. Organo-
metallics 1991, 10, 1613.

3030 Organometallics, Vol. 16, No. 13, 1997
Buhling et al.
phenyl groups and appearing as a singlet, are also
diastereotopic, but they are accidentally degenerate.
In the 13C{1H} NMR spectrum of the 13CO-enriched
HRh(xantham)(CO)(PPh3) complex, the 13CO ligand
gives an apparent double quartet at 205.9 ppm. The
multiplet hydride signal in the 1H NMR spectrum
revealed a C-H coupling of 37.5 Hz, which is consistent
with values found in other trans-13CO hydride com-
plexes.53,54
Bubbling CO through a solution of the HRh(xan-
tham)(CO)(PPh3) complex resulted in partial formation
of HRh(xantham)(CO)2, the presumed catalytically ac-
tive species under hydroformylation conditions. Also in
this complex, xantham is bis(equatorially) coordinated
as judged by the neat doublet in the 31P{1H} NMR
spectrum and the small H-P coupling constant. The
hydride signal unveils a H-Rh coupling constant of 6.6
Hz.
The 13C{1H} NMR spectrum of the crude 13CO-
enriched HRh(xantham)(CO)2 reveals only one CO
resonance as a double triplet, owing to coupling with
rhodium and two phosphorus atoms. The hydride signal
also reveals coupling with two equivalent 13CO ligands.
Low-temperature NMR showed that the observed reso-
nance is a time-averaged signal of an axial and equato-
rial CO which are rapidly exchanging. The CO signal
loses its multiplicity at -60 °C and is very broad at -80
°C. Apparently, the complex is subject to fluxional
processes even at low temperatures. This was previ-
ously reported for the analogous complex of BISBI.54
Unlike the flexible BISBI, the rigid xantham ligand does
not allow Berry pseudorotations55 to occur. A “turnstile”
transformation, as reported by Meakin et al.,56,57 is more
likely since the P-Rh-P entity can then maintain its
integrity throughout the operation. The relatively low
value for the P-C coupling constant can be accounted
for by averaging a negative apical J P-C of about -30
Hz and a positive J P-C of about +10 Hz.58 The same
applies for the low C-H coupling constant. Xantham
has been shown to give similar complexes as Xantphos
and, hence, does not engage in P-N chelated rhodium
complexes.
tham is coordinated in a bidentate fashion as no 31P
signals are observed without coupling to 103Rh. Appar-
ently, xantham is reluctant to form the expected
Rh(acac)(xantham) complex, analogous to the known
Rh(acac)(PPh3)2 complex.59,60 The natural bite angle of
xantham (about 115°) may be the driving force for the
formation of a five-coordinate rhodium species incorpo-
rating one CO ligand. The bite angle in this complex,
presumably possessing a square pyramidal or distorted
trigonal bipyramidal structure, is still somewhat
wrenched, which can explain the relatively low Rh-P
coupling constant. According to the 1H NMR spectrum,
the acetylacetonate ligand is symmetrically coordinated,
thus, in a bidentate fashion. The spectrum further
shows that the methyl and benzylic amino groups and
the ortho phenyl protons in xantham are equivalent,
which suggests that the rhodium complex is involved
in rapid fluxional processes. At lower temperatures,
line broadening is indeed observed, leading to a very
broad singlet at -80 °C. NMR measurements at lower
temperatures have not been performed.
Bubbling 13CO through the solution containing the
complex resulted in the formation of a complex which
reveals a double triplet in the 13C NMR spectrum at
191.5 ppm. The Rh-C and P-C coupling constants of,
respectively, 77.0 and 12.1 Hz are typical of a structure
which possesses an apical CO facing two phosphorus
atoms in an equatorial plane. Upon standing under an
atmosphere of 13CO, other complexes including
Rh(acac)(13CO)2 (doublet at 184.6 ppm, J Rh-C ) 72.4 Hz)
are formed.
The Rh(acac)(CO)(xantham) complex is thermally
stable and was not converted to the Rh(acac)(xantham)
complex by heating the solution at 80 °C for 2 h59,60 or
by carefully evacuating the solution, as judged by the
unchanged IR and 31P NMR spectra.
Rh(acac)(CO)PP complexes, in which PP functions as
a bidentate ligand, have been recently reported for
diphosphites61 but, to our knowledge, have never been
reported for diphosphines.
Cr ysta l Str u ctu r e of Xa n th a m . The crystal struc-
ture of the xantham ligand is shown in Figure 2. When
compared with that of Xantphos,27 it is clear that the
benzylic amino groups, although located at the para
positions, cause a significant distortion of the overall
structure. In contrast with Xantphos, the flat xanthene
backbone of xantham does not possess a mirror plane
(through O, C7, C14, C15). Due to steric congestion,
the diphenylphosphino moieties point into different
directions. Hence, the energetically favorable confor-
mation in which two of the phenyl groups are aligned
(“π-stacking”), as was observed in Xantphos, is lost. Also,
the distance between the phosphorus atoms has in-
creased from 4.080 for Xantphos to 4.383 Å for xantham.
According to molecular modeling calculations,27 this
results in a somewhat larger bite angle for xantham
(115°) than for Xantphos (111.7°) in the rhodium
complexes.
The catalytically active HRh(CO)2(xantham) complex
can also be prepared under hydroformylation conditions
from Rh(acac)(CO)2 and xantham under syngas pres-
sure. The rhodium acetylacetonato complex formed
prior to syngas admission was also studied with NMR
and IR spectroscopy.
Upon addition of 1 equiv of xantham to Rh(acac)(CO)2,
immediate evolution of CO is observed and a yellow
solution is formed. The IR spectrum exhibits a very
strong band at 1968 cm-1, which is indicative of a Rh-
CO species. The 31P NMR spectrum at room tempera-
ture reveals a broad doublet at 11.7 ppm with a Rh-P
coupling constant of 90.9 Hz. The complex concerned
is most likely Rh(acac)(CO)(xantham), in which xan-
(53) Brown, J . M.; Kent, A. G. J . Chem. Soc., Perkin Trans. 2 1987,
1597.
(54) Casey, C. P.; Whiteker, W. T.; Melville, M. G.; Petrovich, L.
M.; Gavney, J . A.; Powell, D. R. J . Am. Chem. Soc. 1992, 114, 5535.
(55) Berry, R. S. J . Chem. Phys. 1960, 32, 933.
(56) Meakin, P.; Muetterties, E. L.; J esson, J . P. J . Am. Chem. Soc.
1972, 94, 5271.
(57) Meakin, P.; J esson, J . P.; Tebbe, F. N.; Muetterties, E. L. J .
Am. Chem. Soc. 1971, 93, 1797.
(58) Brown, J . M.; Kent, A. G. J . Chem. Soc., Chem. Commun, 1982,
723.
Ca ta lysis. Liga n d Effects. The new diphosphines
have been tested in the rhodium-catalyzed hydroformyl-
ation of oct-1-ene. The HRh(CO)2PP complex is formed
(59) Trzeciak, A. M.; Zio´lkowski, J . J . J . Organomet. Chem. 1992,
429, 239.
(60) Bonati, F.; Wilkinson, G. J . Chem. Soc. 1964, 3156.
(61) Buisman, G. J . H.; Vos, E. J .; Kamer, P. C. J .; van Leeuwen, P.
W. N. M. J . Chem. Soc., Dalton Trans. 1995, 409.

Novel Amphiphilic Diphosphines
Organometallics, Vol. 16, No. 13, 1997 3031
number of available â-hydrogen atoms and the in-
creased steric hindrance.
It can be seen that the rhodium catalyst derived from
xantham and Xanthphos gives rise to an even higher
yield of linear aldehyde owing to both a high l/b-ratio
and a relatively low activity for isomerization. This is
ascribed to the well-defined template structure of both
rigid ligands which preferentially occupy two equatorial
sites in the catalytically active rhodium hydride owing
to their relatively large natural bite angle. This geom-
etry leads to a higher proportion of linear aldehyde
formation as compared to geometries with axial-
equatorial chelates.27,53 Xantphos and xantham give
similar results in the hydroformylation of oct-1-ene. The
reaction rate of the xantham-derived catalyst, however,
is somewhat higher despite the electron-donating effect
of the four aminomethyl groups, which has a negative
effect on the reaction rate.62,63 The different geometry
of xantham, as shown for the free ligand in the X-ray
structure, in the active rhodium hydride is held respon-
sible for a faster replacement of a CO molecule by an
alkene.
Hydroformylation of hex-1-ene and dodec-1-ene using
xantham as a ligand shows very similar results.
Con d ition s. The influence of temperature on the
hydroformylation of oct-1-ene is shown in Table 2. As
expected, the turn-over frequencies considerably de-
crease with lower temperature, whereas the l/b-ratios
slightly increase. The isomerization activities also
decrease, especially in the case of xantham, resulting
in the production of 96.9% nonanal. At 100 °C, POPpy
gives rise to an increased proportion of 2/3-octenes. This
is in sharp contrast with POPam for which no isomer-
ization can be detected, not even at 100 °C. When
xantham is used at 100 °C, the selectivity with regard
to both isomerization and linearity only slightly de-
creases while the reaction rate increases to synthetically
attractive values. This effect was also found for Xant-
phos and was ascribed to the rigidity of the diphosphines
stabilizing chelated complexes, even at elevated tem-
peratures. In contrast, ligand systems such as BISBI
and PPh3 give rise to rhodium complexes with a less
rigidly defined geometry, which leads to a significant
increase in isomerization at higher temperatures.26
A few preliminary experiments have been performed
to study the effect of the partial pressures of CO and
H2 for the catalyst system containing POPam (Table 3).
Variation of the partial pressure of CO (entries 1 and
2) has a moderate effect on the turn-over frequencies.
This seems to indicate that the hydroformylation of oct-
1-ene using POPam only has a small negative order in
CO concentration and does not involve a simple rate
equation comprising merely CO replacement by alkene.
The variation in H2 concentration (entry 3) suggests that
the last step in the hydroformylation cycle, hydro-
genolysis of the acyl rhodium complex, is not rate
determining. Remarkably, the selectivity of the hydro-
formylation reaction is not affected by variations in the
composition of the used syngas. Even at CO pressures
as low as 5 bar, POPam fully suppresses isomerization.
Distr ibu tion Ch ar acter istics of th e Fr ee Ligan ds.
For functionalized triphenylphosphines and BISBI de-
F igu r e 2. Molecular structure (ORTEP plot) of xantham.
For clarity hydrogen atoms are omitted.
in situ from Rh(acac)(CO)2 and a 10-fold excess of the
diphosphine. Table 1 lists the hydroformylation results
for the novel ligands, as obtained under standard
conditions, and also includes the results for unmodified
POP and Xantphos, which have been used under the
same conditions for the sake of comparison. As can be
seen in Table 1, POP and its amphiphilic derivatives
give rise to mainly linear aldehydes (88-89%). Al-
though the selectivity for linear aldehydes is moderate
as compared to Xantphos, virtually no isomerization is
observed. Until now, the coordination chemistry of
rhodium catalysts modified with POP under hydro-
formylation conditions has not been thoroughly inves-
tigated, but it is tempting to speculate on the reasons
for the absence of isomerization products. It is reason-
able to assume that the coordinated POP ligand is
capable of fast intramolecular occupation of formed
vacant sites in the rhodium complexes, since the phos-
phorus bidentate ligand can span a wide range of bite
angles; POP may act as a hemilabile terdentate ligand
in an unsaturated transient species, in which the
bridging oxygen atom also coordinates (Figure 3). In
this way, â-H-elimination might be suppressed. The
steric bulk around the rhodium center in the alkene
hydride intermediate and the preference for bite angles
around 110° may explain the increased regioselectivity
in comparison with that of PPh3.
The amino groups in POPam have no effect on the
catalytic properties since both selectivity and activity
are similar to those obtained with unmodified POP. The
catalyst derived from POPpy, however, is almost twice
as fast as the one derived from POP. This has been
observed earlier for pyridyl-modified arylphosphines24
and has been ascribed to the electron-withdrawing
capacity of the pyridyl ring. π-Accepting ligands result
in a weaker bonding of CO molecules to the rhodium
center, thereby facilitating alkene coordination. In
addition, owing to the higher electrophilicity, the rho-
dium alkyl species reacts less readily with CO. This
results in an increased tendency toward â-H-elimination
and thus to some isomerization (0.7%). This in turn
partly accounts for the low yields of branched aldehyde
relative to that found for POP and POPam; the branched
alkyl species is more prone to undergo â-H-elimination
than the linear alkyl species because of the higher
(62) Moser, W. R.; Papite, C. J .; Brannon, D. A.; Duwell, R. A. J .
Mol. Catal. 1987, 41, 271.
(63) Unruh, J . D.; Christenson, J . R. J . Mol. Catal. 1982, 14, 19.

3032 Organometallics, Vol. 16, No. 13, 1997
Ta ble 1. Hyd r ofor m yla tion u n d er Sta n d a r d Con d ition sa
Buhling et al.
selectivity (%)
l-aldehyde
ligand
POP
alkene
time (h)
conversion (%)b
isomersc
b-aldehyde
l/bd
TOFe
oct-1-ene
1
4.5
20
1
4.5
20
1
4.5
21
1
4.5
24
1
4.5
24
1
4.5
24
1
4.8
19.4
67.0
9.1
34.0
88.0
5.0
21.5
71.5
3.8
16.0
61.6
4.1
16.9
67.9
4.3
17.7
71.0
4.0
0.0
0.0
0.0
0.8
0.6
0.7
0.0
0.0
0.0
3.9
3.9
3.9
3.9
4.0
4.0
3.9
3.9
3.8
4.1
4.2
4.2
87.7
88.0
88.2
89.3
89.5
89.3
88.0
88.0
88.0
94.2
94.1
94.1
94.2
94.1
94.1
94.2
94.1
94.2
94.1
94.0
94.0
12.3
12.0
11.8
9.9
7.1
7.3
7.5
9.0
9.0
8.9
7.3
7.3
7.3
239
215
168
449
372
207
251
238
178
179
170
123
196
181
137
205
189
143
193
179
134
POPpy
oct-1-ene
oct-1-ene
oct-1-ene
oct-1-ene
hex-1-ene
dodec-1-ene
9.9
10.0
12.0
12.0
12.0
1.9
2.0
2.0
1.9
1.9
1.9
1.9
2.0
2.0
1.8
1.8
1.8
POPam
xantphos
xantham
49
48
46
50
49
49
49
48
48
51
52
51
4.5
24
16.6
66.9
a
b
Conditions: 20 bar of H2/CO (1:1), 80 °C, toluene (20 mL), [L] ) 17 × 10-4 M, [Rh] ) 1.7 × 10-4 M, [alkene] ) 0.84 M. Percentage
of 1-alkene converted. c Percentage of 2-, 3-, and 4-alkenes formed. Linear/branched ratio. e Turn-over-frequency in mol or aldehyde per
d
mol of Rh per h, averaged over the time given.
Ta ble 2. Va r ia tion of Rea ction Tem p er a tu r e in th e Hyd r ofor m yla tion of Oct-1-en ea
selectivity (%)
ligand
temp (C)
60
time (h)
conversion (%)
isomers
l-aldehyde
b-aldehyde
l/b
TOF
POPpy
4
56
0.2
0.7
2
5.7
64.9
9.5
31.5
64.0
5.1
0.4
0.3
1.5
1.3
1.4
0.0
0.0
0.0
0.0
0.0
0.0
1.4
1.4
3.5
3.4
4.3
4.2
4.1
89.7
89.7
88.3
88.4
88.3
88.8
88.4
88.5
86.1
86.1
86.3
96.7
96.9
94.6
94.7
93.6
93.6
93.7
9.7
9.8
9.2
9.1
8.7
8.6
8.6
7.9
7.6
7.7
6.2
6.2
6.3
71
58
2350
2250
1567
4
37
26
1250
1066
866
10
9
30
26
866
830
732
100
10.2
10.3
10.3
11.2
11.6
11.5
13.9
13.9
13.7
1.9
1.8
1.9
1.9
2.1
2.2
2.2
POPam
40
60
70
4
3.0
64
0.5
2.2
4.5
24
72
48
96
1
33.2
12.5
42.0
78.0
4.4
13.7
29.5
50.7
18.0
34.5
68.5
100
xantham
40
60
51
53
51
51
44
43
43
100
2
4.5
a
Conditions: 20 bar of H2/CO (1:1), toluene (20 mL), [L] ) 17 × 10-4 M, [Rh] ) 1.7 × 10-4 M, [oct-1-ene] ) 0.84 M. See Table 1 for
definitions.
pyridyl)2.25 The recorded extraction curves for POPam
and xantham are almost the same and closely resemble
that of PhP(C6H4CH2NEt2)2.25 Both ligands are com-
pletely located in the organic layer at pH 7, but are
extracted for more than 60% at pH 4.5. Extraction is
complete at the relatively high pH of 2.5.
F igu r e 3. Possible role of the bridging oxygen in the
prevention of â-H-elimination.
Rh od iu m Recyclin g Exp er im en ts. Table 4 shows
the results of the recycling experiments in which the
novel ligands are used. The rhodium contents of the
aqueous and organic layers were analyzed by induc-
tively coupled plasma atomic emission spectrometry
(ICP-AES). The recycled rhodium was reused in a
second hydroformylation run, and the observed reaction
rate as measured by the turn-over frequency (TOF) was
compared with that of the TOF in the first run. Since
the hydroformylation reaction is first-order in oct-1-ene
concentration, the quotient of both turn-over frequencies
is a measure of the recovery of catalytically active
rhodium and is referred to as the retention of activity
(RA).25,26
rivatives, the pH-dependant distribution characteristics
of the free amphiphilic ligand have been shown to be
comparable to that of the corresponding rhodium com-
plex.25,26 The distribution coefficient D is defined as
CH2O
D )
× 100%
CH2O + Corg
In Figure 4, the D-pH plots for POPpy, POPam, and
xantham are depicted. POPpy is located mostly in the
Et2O layer at pH values of 4-7. Extraction of the ligand
into the aqueous phase occurs at pH 3 and is complete
at pH 1. The curve closely resembles that of PhP(3-

Novel Amphiphilic Diphosphines
Organometallics, Vol. 16, No. 13, 1997 3033
Ta ble 3. Va r ia tion s in P (CO) a n d P (H2) in th e Hyd r ofor m yla tion of Oct-1-en e w ith P OP a m a
selectivity (%)
entry
1
PH
PCO
time (h)
conversion (%)
isomers
l-aldehyde
b-aldehyde
l/b
TOF
2
10
10
5
5
0.6
4.5
20
1
4.5
20
1
3.6
25.9
80.4
4.4
18.6
66.4
4.8
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
88.0
88.1
87.9
88.0
88.1
87.7
87.6
88.1
88.1
12.0
11.9
12.1
12.0
11.9
12.3
12.4
11.9
11.9
7.3
7.4
7.3
7.3
7.4
7.1
7.0
7.2
7.2
297
288
201
222
207
166
241
229
178
2
3
30
10
4.5
19
20.6
67.6
a
Conditions: 80 °C, toluene (20 mL), [L] ) 17 × 10-4 M, [Rh] ) 1.7 × 10-4 M, [oct-1-ene] ) 0.84 M. See Table 1 for definitions.
Ta ble 4. Resu lts of th e Recyclin g Exp er im en ts; Rh od iu m Mea su r em en ts by ICP -AES a n d Reten tion of
Activitya
rhodium content (µg)
first organic aqueous
new organic Rh recoveryb Rh balancec R. A.d
entry
1
ligand
conditions
layer
layer
layer
(%)
(%)
(%)
POPpy
6 acidic extractions at pH 2.5
neutralization to pH 7
30
1010
183
15
99
13
2
3
4
5
6
POPpy
8 acidic extractions at pH 3
neutralization to pH 6-6.5
6 acidic extractions at pH 5
neutralization to pH 7
8 acidic extractions at pH 5-5.5
neutralization to pH 6.5-7
7 acidic extractions at pH 5
neutralization to pH 7
30
996
489
432
37
216
707
17
59
65
97
98
100
97
16
51
58
66
86
POPam
POPam
xantham
xantham
< 0.7e
<0.7e
<0.7e
<0.7e
791
99
1194
1193
100
99
9 acidic extractions at pH 5-5.5
neutralization to pH 6.5-7
26
a
b
The 95% confidence interval of the mean measured values is (4.5% for contents >10 µg, otherwise (10%. Rhodium recovered in
the new organic layer as percentage of the total amount measured. c Rhodium mass balance defined as the total amount of rhodium
measured as a percentage of the starting amount () 1235 µg). Retention of activity (see text). e The detection limit is 0.7 µg (based upon
d
3σ of the blank solution).
done under milder conditions, acidic extraction at pH 3
and subsequent neutralization to pH 6-6.5 (entry 2),
gave similar results. Evidently, POPpy is inapt for
rhodium reuse, as observed earlier for pyridyl-modified
PPh3 and BISBI ligands.25,26
Six extractions at pH 5 of the catalytic mixture of
POPam resulted in a bright-yellow aqueous layer and
a colorless toluene layer having a rhodium contamina-
tion below the detection limit (entry 3). Upon neutral-
ization to pH 7, the aqueous layer became turbid and
toluene extractions recovered only 59% of the total
amount of rhodium charged. No improvement was
observed when milder conditions were applied (entry
4).
Successive extraction of the catalytic mixture of
xantham at pH 5, at which free xantham is extracted
for more than 40% (Figure 4), effected the quantitative
extraction of rhodium (entry 5). Neutralization and
successive toluene extractions rendered an aqueous
phase contaminated with only 36 µg Rh. Although the
overall rhodium recovery was almost quantitative (97%),
the retention of activity amounted to only 66%. How-
ever, a substantial elevation to 86% was observed (entry
6) when the acidic titration procedure was done at pH
5-5.5. This result equals the success of PhP(C6H4CH2-
NEt2)2.25 To study the rhodium species involved through-
out the rhodium recycling procedure using xantham, a
small scale experiment was done at higher concentra-
tions and monitored by IR and NMR spectroscopy. The
organic reaction medium in the autoclave, which was
shown to contain the previously characterized HRh-
(xantham)(CO)2 complex, was extracted with a solution
of H2SO4 in D2O. 31P NMR analysis of the deep yellow
F igu r e 4. Extraction curves of POPpy (b), POPam (0) and
xantham (2).
Optimum conditions for the recycling procedure were
established in previous experiments25,26 using the am-
phiphilic derivatives of PPh3 and BISBI. The catalytic
mixture of POPpy was titrated with an aqueous H2SO4
solution of pH 1.8 to an effective pH of 2.5, although at
pH 3, the aqueous layer had already turned slightly
yellow. These observations are in agreement with the
extraction curve depictured in Figure 4. Six extractions
at pH 2.5 rendered a combined bright-yellow aqueous
layer and a colorless toluene layer, which contained only
30 µg of rhodium (2.4% of the total rhodium amount,
entry 1). Upon neutralization to pH 7, the aqueous
phase remained yellow. Analysis confirmed that the
majority of the rhodium amount remained in the aque-
ous phase. The recovered toluene phase accordingly
showed a low retention of activity (13%). An experiment

3034 Organometallics, Vol. 16, No. 13, 1997
Buhling et al.
according to literature procedures. For column chromatogra-
phy, both silica gel 60 (Merck, 230-400 mesh) and aluminium
oxide (activated neutral, 50-200 micron, Acros) were used. 1H
NMR (300 MHz, TMS as standard), 31P NMR (121.5 MHz,
H3PO4 as standard), and 13C NMR (75.5 MHz, TMS as
standard) spectra were measured on a Bruker AMX 300
spectrometer in CDCl3 unless otherwise stated. IR spectra
were recorded on a Nicolet 510 FT-IR spectrophotometer.
Melting points were determined on a Gallenkamp MFB-595
apparatus. GCMS was measured on a HP 5890/5971 spec-
trometer. For exact mass determination, a J EOL J MS-SX/
SX102A spectrometer was used. Hydroformylation reactions
were carried out in a homemade 200 mL stainless steel
autoclave. Gas chromatographic analyses were run on a Carlo
Erba GC 6000 Vega Series apparatus (split/splitless injector;
J &W Scientific; DB1 30m column; film thickness 3.0 µm;
carrier gas 70 kPa He; FID detector) equipped with a HP 3396
integrator. Syngas 3.0 was purchased from UCAR. The
1-alkenes were freshly filtered over a short column of alumi-
num oxide before use to remove hydroperoxides (activated
neutral, 50-200 micron, Acros). UV spectra were measured
on a Varian Cary 4 single-beam apparatus. The pH values
were measured on a Corning 240 pH meter. ICP-AES meas-
urements were done using a sequential J arrell Ash upgraded
(Model 25) Atomscan model 2400 ICP scanning monochroma-
tor. Sulphuric acid calibration solutions of rhodium were
prepared by dilution of a rhodium standard (RhCl3 in 20% HCl)
from J ohnson Matthey. Experimental conditions of the ICP-
AES analysis are described in ref. 25. Elemental analysis was
performed by the Department of Micro-Analyses at the Uni-
versity of Groningen.
aqueous layer revealed the presence of several unknown
rhodium species. Apart from the sharp resonance
assigned to free ligand, several very broad signals
between -10 and +20 ppm were observed. After
neutralization, the aqueous phase was extracted with
toluene-d8. The 31P spectrum of this layer showed the
presence of several phosphorus containing compounds.
The IR spectrum showed bands at 1970 and 1988 cm-1
,
which were also found for the active hydride complex.
This experiment conclusively showed that the catalyti-
cally active hydride completely decomposes upon acidic
extraction in the absence of syngas.
Con clu sion
Rhodium catalysts modified with the novel am-
phiphilic diphosphines are highly active and selective.
The use of POPam ensures complete suppression of
isomerization of the substrate alkene. Although unable
to fully recycle rhodium, POPam enables complete
separation of the rhodium catalyst, to the extent of
99.95%, from the organic reagents by simple acidic
extraction. Xantham is very promising for the use as a
ligand in a rhodium catalyzed hydroformylation process
of higher alkenes. This ligand gives rise to a rhodium
catalyst that is as selective and even somewhat more
active as the catalyst derived from the hydrophobic
Xantphos ligand. Moreover, rhodium and ligand can
be recycled by extraction and re-extraction at nearly
neutral pH, which implies a minimum amount of salt
waste. For commercial application in a commodity
aldehyde producing process, a rhodium recovery of
99.98% is mentioned in the literature.64 Rhodium and
excess xantham can be extracted from the organic
reagents by simple acidic extraction to the extent of
99.95%, while the overall rhodium recovery amounts to
98%. Given the basic experimental setup we used, these
are promising figures. The 86% retention of activity is
high but far from satisfactory from an industrial point
of view. The spectroscopic study showed that the active
hydride decomposes upon recycling. The majority of the
eventually recycled rhodium species, whose structures
have not yet been elucidated, act as precursors for the
active catalyst. Avoiding irreversible decomposition
could be envisaged by performing all recycling steps
under syngas pressure. Finally, the present recovery
procedure in combination with the xantham ligand may
already be applied to the batchwise hydroformylation
of fine chemicals.
NMR simulation was done with Softshell NMR software.67
The following labeling was used in the NMR assignments of
the POP-related compounds: PPh (Ci-p), P(Ph-amino) (Ci′-p′),
and P(3-pyridyl) (C2-6). For the xantham-related compounds,
the following labeling was used in the NMR assignments and
nomenclature:
Ch lor op h en yl(3-p yr id yl)p h osp h in e (4). A 2.5 M solu-
tion of n-butyllithium in hexane (0.150 mol, 62.3 mL) was
cooled to -78 °C. A solution of 3-bromopyridine (0.150 mol,
14.45 mL) in Et2O (120 mL) was added in 1 h, and stirring
was continued for 1 h. The yellow suspension was quickly
filtered and washed first with cold hexanes (-78 °C, 2 × 35
mL) and then with hexanes at room temperature (50 mL). The
dark yellow solid was suspended in THF (200 mL). In a second
flask, a solution of dichlorophenylphosphine (0.30 mol, 40.5
mL) in Et2O (100 mL) was cooled to -78 °C and the suspension
was added in 1 h in portions through a glass joint. The
reaction mixture was then allowed to warm to room temper-
ature overnight. The solvents were evaporated, and the brown
residue was distilled at reduced pressure, giving a forerun of
dichlorophenylphosphine and the product (bp 110 °C, 0.09
mmHg). The yellowish liquid product contained 5 wt %
lithium chloride, which sublimed from the residue during
distillation. Yield: 33% (0.050 mol, 11.0 g).
Exp er im en ta l Section
Gen er a l Con sid er a tion s. All reactions were carried out
in flame-dried glasswork using standard Schlenk techniques
under an argon atmosphere. Toluene, THF, and diethyl ether
were distilled from sodium/benzophenone; CH2Cl2 was dried
over P2O5 and distilled from CaH2. All solvents used in the
preparation or handling of phosphines were degassed prior to
use. Solvents and reagents were distilled prior to use. Mesit-
ylene and methyl iodide were freshly distilled from sodium
and anhydrous CaCl2, respectively. All chemicals were pur-
chased from Acros Chimica or Aldrich Chemical Co. Bis(2-
iodoethyl) ether,65 bis[2-(diphenylphosphino)ethyl] ether,39
Cl2PNEt2,66 and (4-bromobenzyl)diethylamine24 were prepared
1H NMR: δ 8.72 (m, 1H, H2), 8.56 (d, 1H, J ) 4.8 Hz, H6),
7.78 (dist. t, 1H, J ) 7.9 Hz, H4), 7.57 (m, 2H, aromatic), 7.37
(m, 3H, aromatic), 7.24 (m, 1H, H5). 31P{1H} NMR: δ 75.4.
13C NMR: δ 152.3 (d, J ) 31.7 Hz, C2), 151.2 (C6), 140.0 (d, J
) 17.4 Hz, C4), 137.7 (d, J ) 31.7 Hz, Ci), 136.2 (d, J ) 38.5
Hz, C3), 132.3 (d, J ) 25.7 Hz, Co), 131.5 (Cp), 129.5 (d, J )
(64) Kuntz, E. G. CHEMTECH 1987, 17, 570.
(65) Metzger, L.; Kuhnen, O. P. Z. Naturforsch. B 1957, 12, 28.
(66) Issleib, K.; Seidel, W. Chem. Ber. 1959, 92, 2681.
(67) Budzelaar, P. H. M. geNMR 3.5M; Ivorysoft: Amsterdam, The
Netherlands.

Novel Amphiphilic Diphosphines
Organometallics, Vol. 16, No. 13, 1997 3035
7.6 Hz, Cm), 124.3 (br s, C5). Exact mass (FAB): 186.0456 (M
- Cl)+ (calcd for C11H9NP, 186.0473).
solution of NH4Cl (75 mL) was added. The aqueous layer was
extracted with CH2Cl2 (2 × 50 mL), and the combined organic
phases were evaporated. The resulting yellow oil was purified
by column chromatography (silica gel, 55% THF/40% toluene/
5% NEt3). The resulting oil was dissolved in an aqueous 0.05
M solution of H2SO4 (100 mL) for further purification. This
solution was washed with CH2Cl2 (3 × 50 mL) and then
neutralized with a saturated aqueous solution of NaHCO3.
Extraction with CH2Cl2 (2 × 75 mL) and evaporation of the
organic phases gave a colorless oil. Yield: 69% (34.5 mmol,
9.9 g).
P h en yl(3-p yr id yl)p h osp h in e (5). Small pieces of sodium
(137 mmol, 3.15 g) and ca. 20 mg of 4,4′-di-tert-butylbiphenyl
were added to a solution of chlorophenyl(3-pyridyl)phosphine
4 (66.6 mmol, 14.7 g) in THF (120 mL). The initially colorless
mixture was refluxed until all the sodium had reacted, which
took approximately 5 days. Then the dark red reaction
mixture was quenched with water (40 mL) at 0 °C and Et2O
(50 mL) was added. The organic layer was separated and
evaporated. The resulting dark red residue was azeotropically
dried with toluene (3 × 50 mL) and distilled at reduced
pressure (bp 106 °C, 0.20 mmHg). Yield: 53% (6.6 g, 35.3
mmol) of a colorless liquid.
1H NMR: δ 8.70 (m, 1H, H2), 8.54 (d, 1H, J ) 5.6 Hz, H6),
7.73 (m, 1H, H4), 7.48 (m, 2H, aromatic), 7.34 (m, 3H,
aromatic), 7.22 (m, 1H, H5), 5.22 (d, 1H, J P-H ) 220.4 Hz, HP).
31P NMR: δ -49.6 (d, J P-H ) 220 Hz). 13C NMR: δ 155.4 (d, J
) 18.1 Hz, C2), 150.9 (C6), 142.6 (d, J ) 15.9 Hz, C4), 135.5 (d,
J ) 17.4 Hz, Co), 134.5 (d, J ) 9.1 Hz, Ci), 132.7 (d, J ) 15.1
Hz, C3), 130.4 (Cp), 130.2 (d, J ) 6.8 Hz, Cm), 125.0 (br s, C5).
1H NMR: δ 8.07 (d, 1H, J P-H ) 480.5 Hz, HP), 7.75-7.45
(m, 9H, aromatic), 3.59 (s, 2H, CH2N), 2.51 (q, 4H, J ) 7.1
Hz, CH2CH3), 1.03 (t, 6H, J ) 7.1 Hz, CH3). 31P NMR: δ 22.1
(dm, J P-H ) 479 Hz). 13C NMR: δ 133.0 (Cp), 132.1 (d, J )
101.2 Hz, Ci+i′), 131.3 (d, J ) 3.8 Hz, Co/o′), 131.1 (d, J ) 4.5
Hz, Co/o′), 130.7 (Cp′), 129.7 (d, J ) 12.8 Hz, Cm/m′), 129.4 (d, J
) 12.8 Hz, Cm/m′), 57.9 (CH2N), 47.4 (CH2CH3), 12.3 (CH3).
GCMS m/ e 287 (M+), 272 (M+ - CH3), 215 (M+ - N(CH2-
CH3)2). Exact mass (FAB): 286.1337 (M - H)+ (calcd for
C
17H21ONP, 286.1361).
B i s [2 -((4 -((d i e t h y l a m i n o )m e t h y l )p h e n y l )p h e n -
Bis[2-(p h en yl(3-p yr id yl)p h osp h in o)et h yl]
E t h er
ylp h osp h in o)eth yl] Eth er Dioxid e (7). A solution of 6 (9.5
mmol, 2.72 g) in THF (25 mL) was slowly added to a
suspension of sodium hydride (10.0 mmol, 0.30 g) in THF (10
mL) at room temperature. The resulting yellow solution was
added to a solution of bis(2-iodoethyl) ether (4.7 mmol, 1.54 g)
in THF (40 mL) at -78 °C in 1/2 h. The reaction mixture was
allowed to warm to room temperature overnight. It was then
poured into an aqueous phosphate buffer solution (pH 7, 200
mL), and Et2O (50 mL) was added. The phases were sepa-
rated, and the aqueous layer was washed with CH2Cl2 (2 ×
80 mL). The combined organic phases were dried on Na2SO4
and evaporated. The resulting yellow oil was purified by
column chromatography (silica gel, 60% THF/35% toluene/5%
NEt3). For further purification, water (100 mL) and an
aqueous 0.1 M solution of H2SO4 was added at 0 °C until the
pH was 2-3 and the oil dissolved. This solution was washed
with CH2Cl2 (3 × 60 mL) and toluene (60 mL) and neutralized
with a saturated aqueous solution of NaHCO3. The aqueous
layer was then extracted with CH2Cl2 (3 × 60 mL). The
organic phases were dried on Na2SO4 and evaporated, result-
ing in a colorless glassy oil. Yield 56% (2.64 mmol, 1.7 g).
1H NMR: δ 7.75-7.52 (m, 9H, aromatic), 7.50-7.41 (m, 9H,
aromatic), 3.65 (q, 4H, J P-H ) 8.3, J H-H ) 8.3 Hz, CH2O), 3.59
(s, 4H, CH2N), 2.50 (q + m, 12H, J ) 7.1 Hz, CH2CH3 + CH2P),
1.03 (t, 12H, J ) 7.1 Hz, CH3). 31P{1H} NMR: δ 29.9. 13C
NMR: δ 132.7 (d, J ) 99.7 Hz, Ci+i′), 131.6 (Cp), 130.5 (d, J )
3.8 Hz, Co/o′), 130.4 (d, J ) 3.0 Hz, Co/o′), 130.2 (Cp′), 128.9 (d,
J ) 12.1 Hz, Cm/m′), 128.4 (d, J ) 12.1 Hz, Cm/m′), 63.9 (CH2O),
56.9 (CH2N), 46.6 (CH2CH3), 30.4 (d, J ) 70.7 Hz, CH2P), 11.3
(CH3).
B i s [2 -((4 -((d i e t h y l a m i n o )m e t h y l )p h e n y l )p h e n -
ylp h osp h in o)eth yl] Eth er (P OP a m , 2). Dioxide 7 (2.64
mmol, 1.7 g) was dissolved in acetonitrile (80 mL), and
triethylamine (15.9 mmol, 2.2 mL) was added. Trichlorosilane
(15.9 mmol, 1.6 mL) was added dropwise at 0 °C. The
heterogeneous reaction mixture was stirred for 2 h at room
temperature. The clear reaction mixture obtained was con-
centrated to ca. 20 mL, and toluene (20 mL) and an aqueous
20% solution of KOH (30 mL) were added at 0 °C. The turbid
organic phase was separated, and the aqueous phase was
washed with Et2O (20 mL). The combined organic phases were
evaporated, and the resulting brown oil was purified by column
chromatography (silica gel, 20% THF/75% toluene/5% NEt3).
Yield 42% of a yellowish oil (1.1 mmol, 0.67 g).
(P OP p y, 1). A solution of phenyl(3-pyridyl)phosphine, 5 (4.9
mmol, 0.92 g), in THF (10 mL) was cooled to -78 °C. A 2.5 M
solution of n-butyllithium in hexane (4.9 mmol, 2.0 mL) was
added in 30 min, and stirring was continued at room temper-
ature for 1 h. A solution of bis(2-chloroethyl) ether (2.45 mmol,
0.29 g) in THF (15 mL) was added dropwise at -78 °C to the
resulting deep red-brown solution. The reaction mixture was
then allowed to warm to room temperature overnight. After
the reaction mixture was concentrated, an aqueous 4 M
solution of NaOH (10 mL) was added. The aqueous layer was
extracted with CH2Cl2 (3 × 20 mL). The combined organic
phases were evaporated, and the resulting oil was purified by
column chromatography (silica gel, 10% MeOH/CH2Cl2).
Yield: 69% of a yellowish oil (1.7 mmol, 0.76 g).
1H NMR: δ 8.61 (m, 2H, H2), 8.53 (d, 2H, J ) 4.7 Hz, H6),
7.64 (dist t, 2H, J ) 7.0 Hz, H4), 7.50-7.35 (m, 10H, aromatic),
7.23 (m, 2H, H5), 3.52 (q, 4H, J P-H ) 7.6, J H-H ) 7.6 Hz, CH2O),
2.34 (dist t, 4H, J P-H < 3, J H-H ) 7.4 Hz, CH2P). 31P{1H}
NMR: δ -26.8. 13C NMR: δ 152.9 (d, J ) 22.6 Hz, C2), 149.3
(C6), 139.7 (d, J ) 15.1 Hz, C4), 136.4 (d, J ) 12.1 Hz, Ci/3),
134.7 (d, J ) 18.1 Hz, Ci/3), 132.6 (d, J ) 19.6 Hz, Co), 128.9
(Cp), 128.5 (d, J ) 6.8 Hz, Cm), 123.2 (C5), 67.6 (d, J ) 22.8
Hz, CH2O), 28.2 (d, J ) 13.0 Hz, CH2P). Exact mass (FAB):
445.1553 (M + H) (calcd for C26H27N2OP2, 445.1599), Anal.
Calcd for C26H26N2OP2: C, 70.26; H, 5.90; N, 6.30. Found: C,
69.46; H, 6.15; N, 6.19.
P h en ylp h osp h in ic a cid eth yl ester was prepared from
dichlorophenylphosphine, ethanol, and water by the method
of Emmick.68 Yield: 86% (literature yield 59%), bp 91-92 °C,
0.3 mmHg. 31P NMR: δ 25.1 (d, J P-H ) 563 Hz). 1H NMR: δ
7.77-7.71 (m, 2H, aromatic), 7.54 (d, J P-H ) 561.3 Hz), 7.53-
7.41 (m, 3H, aromatic), 4.13 (m, 2H, CH2), 1.34 (t, 3H, J ) 7.0
Hz).
[4-((-Dieth ylam in o)m eth yl)ph en yl]ph en ylph osph in e ox-
id e (6). A solution of (4-bromobenzyl)diethylamine (25 mmol,
6.05 g) in THF (60 mL) was added dropwise to magnesium
(25.5 mmol, 0.625 g), which was preactivated by 1,2-dibromo-
ethane, allowing a gentle reflux. The reaction mixture was
refluxed for an additional period of 3 h. In another flask, a
solution of phenylphosphinic acid ethyl ester (25 mmol, 4.25
g) in THF (80 mL) was cooled to -78 °C. A 2.5 M solution of
n-butyllithium in hexane (25 mmol, 10.0 mL) was added in 1
h, resulting in a light yellow solution. After the mixture was
1
stirred at room temperature for /2 h, the phosphide anion was
1H NMR: δ 7.43-7.30 (m, 18H, aromatic), 3.56 (s, 4H,
CH2N), 3.49 (q, 4H, J P-H ) 7.9, J H-H ) 7.9 Hz, CH2O), 2.52
(q, 8H, J ) 7.1 Hz, CH2CH3), 2.33 (dist t, 4H, J P-H < 5, J H-H
) 8.0 Hz, CH2P), 1.04 (t, 12H, J ) 7.1 Hz, CH3). 31P{1H} NMR:
δ -22.6. 13C NMR: δ 140.7 (Cp′), 138.3 (d, J ) 12.8 Hz, Ci/i′),
135.8 (d, J ) 12.1 Hz, Ci/i′), 132,4 (d, J ) 18.9 Hz, Co/o′), 128.7
added at 0 °C to the Grignard reagent in 1/2 h. The reaction
mixture was allowed to warm to room temperature overnight.
The solvents were evaporated, and a saturated aqueous
(68) Emmick, T. L.; Letsinger, R. L. J . Am. Chem. Soc. 1968, 90,
3459.

3036 Organometallics, Vol. 16, No. 13, 1997
Buhling et al.
(d, J ) 6.8 Hz, Cm/m′), 128.3 (Cp), 128.2 (d, J ) 6.8 Hz, Cm/m′),
67.8 (d, J ) 24.9 Hz, CH2O), 57.0 (CH2N), 46.6 (CH2CH3), 28.7
(d, J ) 12.7 Hz, CH2P), 11.6 (CH3). Exact mass (FAB):
613.3429 (M + H) (calcd for C38H51N2OP2, 613.3477). Anal.
Calcd for C38H50N2OP2: C, 74.48; H, 8.23; N, 4.57. Found: C,
74.52; H, 8.25; N, 4.48.
4,6-Bis[bis(4-((dieth ylam in o)m eth yl)ph en yl)ph osph in o]-
10,10′-d im eth ylxa n th en e (xa n th a m , 3). A 2.5 M solution
of n-butyllithium in hexane (6.15 mmol, 2.46 mL) was cooled
to -25 °C. A solution of 10 (3.0 mmol, 1.1 g) in Et2O (35 mL)
was added in 30 min, and stirring was continued at +15 °C
for another 30 min. Then a solution of 9 (6.0 mmol, 2.68 g) in
Et2O (20 mL) was added dropwise at -20 °C. The reaction
mixture was allowed to warm to room temperature overnight
and poured in an aqueous 0.15 M solution of H2SO4 (100 mL).
The aqueous phase was separated, washed with CH2Cl2 (3 ×
50 mL) and toluene (50 mL), and neutralized with a saturated
aqueous solution NaHCO3. Extraction with CH2Cl2 (4 × 50
mL) and evaporation of the combined organic phases resulted
in a yellow oil. This crude oil was further purified by flash
column chromatography (silica gel, 55% EtOAc/40% hexane/
5% NEt3). Crystallization from acetonitrile gave white crystals
suitable for crystal structure determination. Yield: 74% (2.2
mmol, 2.0 g)
1H NMR: δ 7.39 (d, 2H, J ) 7.4 Hz, H3), 7.25-7.17 (m, 16H,
P(C6H4)), 6.95 (d, 2H, J ) 7.6 Hz, H2), 6.59 (d, 2H, J ) 7.1 Hz,
H1), 3.57 (s, 8H, CH2N), 2.55 (q, 16H, J ) 7.1 Hz, CH2CH3),
1.66 (s, 6H, CH3), 1.07 (t, 24H, J ) 7.1 Hz, CH2CH3). 31P NMR:
δ -18.9. 13C NMR: δ 152.2 (t, J ) 9.8 Hz, C12), through-space
P-P coupling constant g45 Hz, 139.7 (C4′), 135.5 (t, J ) 6.0
Hz, C1′), 133.7 (t, J ) 10.6 Hz, C2′), 131.8 (C3), 129.5 (C11), 128.5
(br s, C3′), 126.3 (t, J ) 10.2 Hz, C4), 126.0 (C1/2), 122.9 (C1/2),
57.1 (CH2N), 46.6 (CH2CH3), 34.2 (C10), 31.9 (CCH3), 11.6
(CH2CH3). Mp 286-288 °C. Anal. Calcd for C59H76N4OP2:
C, 77.09; H, 8.34; N, 6.09. Found: C, 76.99; H, 8.43; N, 6.03.
HRh (xa n th a m )(CO)(P P h 3) (11). This complex was pre-
pared in situ by stirring a solution of xantham (40 mg, 0.044
mmol) and HRh(CO)(PPh3)3 (40 mg, 0.044 mmol) in benzene-
d6 (1 mL) overnight at 25 C.
Bis[4-((d ieth yla m in o)m eth yl)p h en yl]p h osp h in ou s Di-
eth yla m id e (8). A solution of (4-bromobenzyl)diethylamine
(25 mmol, 6.05 g) in THF (50 mL) was added to magnesium
(26 mmol, 0.63 g), which was activated by 1,2-dibromoethane,
at such a rate that the reaction mixture refluxed gently. After
the addition, the reaction mixture was refluxed for an ad-
ditional 2 h. At -78 °C, a solution of dichlorophosphonous
diethylamide, Cl2PNEt2 (12.5 mmol, 2.17 g), in THF (10 mL)
was added in 30 min. The reaction mixture was allowed to
warm to room temperature overnight. The THF was evapo-
rated and hexane (35 mL) was added. The suspension was
filtered, and the solid was washed with another 20 mL of
hexane. The filtrate was evaporated, resulting in a viscous
yellow oil, which was sufficiently pure for further synthesis.
Yield: 92% (11.5 mmol, 4.92 g).
1H NMR: δ 7.40-7.28 (m, 8H, aromatic), 3.58 (s, 4H, CH2N),
3.12-3.02 (dq, 4H, J PH ) 9.4, J H-H ) 7.1 Hz, CH2NP), 2.54 (q,
8H, J ) 7.1 Hz, CH2CH3), 1.06 (t, 12H, J ) 7.1 Hz, CH3), 0.94
(t, 6H, J ) 7.1 Hz, CH3CH2NP). 31P NMR: δ 61.1. 13C NMR:
δ 138.5 (d, J ) 13.6 Hz, C1′), 138.3 (C4′), 131.6 (d, J ) 20.4 Hz,
C2′), 128.5 (d, J ) 6.0 Hz, C3′), 57.0 (CH2N), 46.6 (CH2CH3),
44.1 (d, J ) 15.1 Hz, CH2NP), 14.3 (d, J ) 3.3 Hz, CH3CH2-
NP), 11.5 (CH3).
B is [4-(N ,N -d ie t h y la m in o m e t h y l)p h e n y l]p h e n o x y -
p h osp h in e (9). Phenol (67 mmol, 6.3 g) was azeotropically
dried three times with 25 mL of toluene. A solution of 8 (16.8
mmol, 7.2 g) in THF (50 mL) and mesitylene (60 mL) was
added, and the reaction mixture was stirred at 80 °C overnight.
The solvents were distilled from the reaction mixture at 90
°C. The excess phenol was removed by repeated codistillation
with mesitylene (5 × 30 mL). The resulting yellow oil was
free of phenol and 100% pure. Yield: 98% (16.5 mmol, 7.39
g).
1H NMR: δ 7.54 (approximate t, 4H, J ) 7.7 Hz, aromatic),
7.38 (approximate d, 4H, J ) 7.4 Hz, aromatic), 7.27 (ap-
proximate t, 2H, J ) 8.5 Hz, aromatic), 7.13 (m, 2H, aromatic),
7.01 (t, 1H, J ) 7.3 Hz, aromatic), 3.58 (s, 4H, CH2N), 2.53 (q,
8H, J ) 7.1 Hz, CH2CH3), 1.05 (t, 12H, J ) 7.1 Hz, CH3). 31P
NMR: δ 111.0. 13C NMR: δ 157.2 (d, J ) 9.8 Hz, Ci), 140.8
(C4′), 139.1 (d, J ) 16.6 Hz, C1′), 130.5 (Cm), 130.4 (d, J ) 22.6
Hz, C2′), 129.0 (d, J ) 6.8 Hz, C3′), 119.6 (Cp), 118.7 (d, J )
10.6 Hz, Co), 56.9 (CH2N), 46.4 (CH2CH3), 11.2 (CH3).
1H NMR: (C6D6) δ 7.91 (m, 4H, aromatic), 7.81 (m, 2H,
aromatic), 7.71 (apparent q, J ) 4.1 Hz, 4H, aromatic), 7.61
(m, 4H, aromatic), 7.46 (m, 8H, aromatic), 7.22 (m, 6H,
aromatic), 7.11 (m, 25H, aromatic), 6.98 (m, 12H, aromatic),
1
6.85 (m, 2H, aromatic), AB (3.61 (d, J H-H ) 14.3 Hz) + 3.52
(d, 1J H-H ) 14.3 Hz), 4H, CH2N), 3.43 (s, 4H, CH2N), 2.48 (dm,
16H, J ) 7.1 Hz, CH2CH3), 1.49 (s, 3H, CH3), 1.32 (s, 3H, CH3),
1.03 (m, 24H, J ) 7.1 Hz, CH2CH3), -8.95 (dt, 1H, J Rh-H
)
1.5 Hz, J HP ) 12.2 Hz, J H-P′ ) 19.2 Hz, RhH). 31P NMR:
(C6D6): δ 42.2 (dt, J Rh-P′ ) 167.5 Hz, J P-P′ ) 131.6 Hz, PPh3),
23.8 (dd, J Rh-P ) 148.1 Hz, J P-P′ ) 131.6 Hz, xantham). IR
(ν(cm-1), C6D6): 1996.6 (s), 1915.1 (w). Exact mass (FAB):
1314.5 (M + H) (calcd for C78H93N4O2P3Rh, 1314.5).
HRh (xa n th a m )(CO)2 (12). A gentle stream of CO was led
through a solution of HRh(xantham)(CO)(PPh3) in benzene-
d6, which was prepared as described above, in an NMR tube
for 30 min. No quantitative conversion, about 70%, could be
achieved due to the presence of excess PPh3.
4,6-Dib r om o-10,10-d im et h ylxa n t h en e (10). 9.9′-Di-
methylxanthene (4.8 mmol, 1.0 g), TMEDA (14 mmol, 2.09
mL), and Et2O (45 mL) were cooled to -78 °C. A 1.3 M
solution of sec-butyllithium in hexane (14 mmol, 10.8 mL) was
added, and the reaction mixture was stirred at room temper-
ature for 20 h. A solution of bromine (15.9 mmol, 0.82 mL) in
pentane (10 mL) was added dropwise at -78 °C, after which
the reaction mixture was allowed to warm to room tempera-
ture overnight. Then a 20% aqueous solution of sodium
bisulphite (20 mL) was added. The organic phase was
separated, and the aqueous phase was washed with Et2O (20
mL). The combined organic phases were washed with the
sulphite solution (15 mL) and water (30 mL), dried over
MgSO4, and evaporated. The orange oil was purified by flash
column chromatography (silica gel, 3% EtOAc/hexane), result-
ing in a yellow oil which solidified overnight. Yield: 70% (3.3
mmol, 1.22 g).
1H NMR (C6D6): δ 7.86-6.76 (aromatic), 3.43 (s, 8H, CH2N),
2.53 (q, 16H, J ) 6.9 Hz, CH2CH3), 1.46 (s, 6H, CH3), 1.07 (t,
24H, J ) 7.0 Hz, CH2CH3), -8.46 (dt, 1H, J Rh-H ) 6.6 Hz, J HP
) 16.8 Hz, RhH). 31P NMR (C6D6): δ 18.7 (d, J Rh-P ) 126.1
Hz). IR (ν(cm-1), C6D6): 1989.9 (s), 1969.6 (s), 1937.9 (s).
HRh (xa n th a m )(13CO)(P P h 3). 13CO-enriched 11 was pre-
pared in situ by bubbling a stream of 13CO very gently into a
solution of 11 in an NMR tube for 10 s, after which the tube
was sealed and thoroughly shaken. Low-temperature NMR:
experiments were done with toluene-d8 as the solvent.
13C NMR (C6D6) δ 205.9 (dq, J Rh-C ) 54.4 Hz, J P-C ) 10.7
Hz, CO). 1H NMR (C6D6): δ -8.93 (mp, J H-C ) 37.5 Hz, J Rh-H
) 1.5 Hz, J HP ) 12.2 Hz, J H-P′ ) 19.1 Hz, RhH).
HRh (xa n th a m )(13CO)2. 13CO-enriched 12 was prepared
analogously to 13CO-enriched 11 by the addition of 13CO for 5
min.
13C NMR (C6D6): δ 200.1 (dt, J Rh-C ) 64.9 Hz, J P-C ) 10.6
Hz, CO). 1H NMR (C6D6): δ -8.45 (mp, J H-C ) 10.0 Hz, J Rh-H
) 6.4 Hz, J HP ) 16.6 Hz, RhH).
1H NMR: δ 7.49 (dd, 2H, J ) 7.9, 1.4 Hz, H3), 7.36 (dd, 2H,
J ) 7.9, 1.4 Hz, H1), 7.00 (t, 2H, J ) 7.9 Hz, H2), 1.63 (s, 6H,
CH3). 13C NMR: δ 147.1 (C12), 131.7 (C11), 131.2 (aromatic),
124.7 (aromatic), 124.2 (aromatic), 110.8 (C4), 35.1 (C10), 31.7
(CCH3). GCMS m/ e 368 (M+), 353 (M+ - CH3), 273 (M+
CH3, -Br). Mp 53-54 °C.
-
Rh (xa n th a m )(CO)(a ca c). Xantham (35 mg, 0.038 mmol)
and Rh(acac)(CO)2 (9 mg, 0.035 mmol) were dissolved in

Novel Amphiphilic Diphosphines
Organometallics, Vol. 16, No. 13, 1997 3037
Ta ble 5. Selected Wa velen gth s a n d Cor r esp on d in g
Extin ction Coefficien ts
benzene-d6 (0.8 mL) and stirred for 15 min. This complex was
enriched with 13CO by bubbling stream of 13CO a very gently
into the solution in an NMR tube for 2 min. Low-temperature
NMR experiments were done with toluene-d8 as the solvent.
X-r a y Cr ysta l Str u ctu r e Deter m in a tion of Xa n th a m .
A crystal with approximate dimensions of 0.35 × 0.50 × 0.50
mm was used for data collection on an Enraf-Nonius CAD-4
diffractometer with graphite-monochromated Cu KR radiation
and ω-2θ scan. A total of 10 996 unique reflections was
measured within the range 0 e h e 14, -19 e k e 19, -21 e
l e 19. Of these, 6153 were above the significance level of 2.5
organic solution
aqueous solution
10-4 ꢀ/dm3
10-4 ꢀ/dm3
phosphine
λ/nm
mol-1 cm-1
λ/nm
mol-1 cm-1
POPpy
POPam
xantham
250
252
262
2.21
2.13
2.74
240
270
273
2.24
0.86
2.08
19 mL of toluene under an atmosphere of argon. The autoclave
was pressurized with syngas (CO:H2 ) 1:1) to 20 bar, and the
temperature was raised to 80 °C in approximately 1 h.
Subsequently, oct-1-ene (20 mmol, 3.14 mL) and decane (3
mmol, 0.6 mL) as internal standard were added under pres-
sure. Samples were taken which were quenched with tri-
phenylphosphite to deactivate the catalyst and analyzed by
gas chromatography.
Distr ibu tion Mea su r em en ts of th e F r ee Liga n d s. The
method for measuring the distribution characteristics of the
free ligands have been described in a previous paper.26 Table
5 lists the extinction coefficients of the ligands in Et2O and in
water.
Rh od iu m Recyclin g Exp er im en ts. Experimental details
concerning the procedure of a typical titration experiment and
rhodium analysis by ICP-AES were given elsewhere.26 In the
NMR/IR experiment, a small vessel was filled with Rh(acac)-
(CO)2 (0.027 mmol, 8.5 mg), xantham (0.11 mmol, 0.104 g),
and toluene (2.5 mL) and placed in the autoclave. After 2 h
under 20 bar of syngas at 80 °C, the reaction mixture was
cooled to 10 °C and siphoned into a flask. Characterization
of this organic mixture revealed the presence of the HRh-
(xantham)(CO)2 complex (31P δ ) 18.9 ppm, J Rh-P ) 127.6 Hz;
IR 1988, 1970, and 1939 cm-1). It was extracted into D2O by
titration with H2SO4 (pH 1) and neutralized with NaHCO3 in
the presence of toluene-d8. The 31P NMR spectrum of the
toluene-d8 layer contained the following signals: δ 27.0 (s,
0.06), 24.4 (d, 0.46, J ) 8.7 Hz), 11.4 (br m, 0.05), 7.6 (double
m, 0.2, J ) 165 Hz), -0.4 (double m, 0.2, J ) 152 Hz), -18.6
(xantham, 1.0), -22.7 (d, 0.4, J ) 9.7 Hz).
σ (I). The maximum value of (sin θ)/λ was 0.63 Å-1
. Two
reference reflections (033h, 300) were measured hourly and
showed no decrease during the 34 h collecting time. Unit-cell
parameters were refined by a least-squares fitting procedure
using 23 reflections with 80 < 2θ < 82°. Corrections for
Lorentz and polarization effects were applied. The structure
was solved by direct methods. After isotropic refinement, some
of the ethyl groups had rather high temperature factors. It
was possible to distribute most of these atoms over two half-
occupied positions which remained isotropic during the entire
refinement (C26, C56, C57, and C59). No attempts were made
to determine the hydrogen atoms of the disordered atoms. The
hydrogen atoms were calculated. Full-matrix least-squares
refinement on F, anisotropic for the non-hydrogen atoms and
isotropic for the hydrogen atoms keeping the latter fixed at
their calculated positions with a temperature factor of U )
0.10 Å2, converged to R ) 0.080, Rw ) 0.085, (∆/σ)max ) 0.86,
S ) 0.500(5). A weighting scheme w ) [14.8 + 0.0055*(σ(Fobs))2
+ 0.00004/(σ(Fobs))]-1 was used. The secondary isotropic
extinction coefficient69,70 refined to ext ) 0.073(8). A final
difference Fourier map revealed a residual electron density
between -0.5 and 0.8 e Å-3
. Scattering factors were taken
from Cromer and Mann (1968).71,72 The anomalous scattering
of P was taken into account. All calculations were performed
with XTAL,73 unless stated otherwise.
Hyd r ofor m yla tion . In a typical experiment, the autoclave
was filled with a mixture of a 4 mM solution of Rh(acac)(CO)2
in toluene (0.004 mmol, 1 mL), the ligand (0.04 mmol), and
(69) Zachariasen, W. H. Acta Crystallogr. 1967, A23, 558.
(70) Larson, A. C. In The inclusion of Secondary Extinction in Least-
Squares Refinement of Crystal Structures. Crystallographic Computing;
Ahmed, F. R., Hall, S. R., Huber, C. P., Eds.; Munksgaard: Copen-
hagen, 1969; pp 291-294.
Su p p or tin g In for m a tion Ava ila ble: Tables of crystal
data and collection parameters, atomic coordinates, bond
lengths, bond angles, thermal parameters, and H atom coor-
dinates and NMR spectra of HRH(xantham)(CO)(PPh3), HRh-
(xantham)(CO)2, HRH(xantham)(13CO)(PPh3), HRh(xantham)-
(13CO)2, Rh(xantham)(CO)(acac), and Rh(xantham)(13CO)(acac)
(25 pages). Ordering information is given on any current
masthead page.
(71) Cromer, D. T.; Mann, J . B. Acta Crystallogr. 1968, A24, 321-
324.
(72) International Tables for X-ray Crystallography; Kynoch Press:
Birmingham, 1974; Vol. IV; pp 55.
(73) XTAL3.2 Reference Manual; Hall, S. R., Flack, H. D., Stewart,
J . M., Eds.; Universities of Western Australia, Geneva, and Mary-
land: 1992.
OM970068W