Bulky diphosphite-modified rhodium catalysts: Hydroformylation and characterization

Article abstract of DOI:10.1021/om950549k

The rhodium-catalyzed hydroformylation with the diphosphites P[O(2,2′-(4-X-6-Y-C₆H₂) ₂O]-[O(2,2′-(4-X-6-Y-C₆H₂) ₂)OP][O(4-Z-C₆H₄]₂ (X = Y = tert-butyl, Z = H (1), Z = OMe (2), Z = C₆H₅ (3), Z = Cl (4); X = OMe, Y = ferf-butyl, Z = H (5)), [PO(2,2′-(4,6-(teri-C₄H₉)₂C ₆H₂)₂)O]₂R [R = O(CH₂)₂O (6), R = O(CH₂)₃O (7)], [PO(2,2′-(C₆H₄)₂)O]₂R, R = O(2,2′-(4-MeO-6-tert-C₄H₉C₆H ₂)₂O (8), R = O(2,2′(4,6-tert-C₄H₉)₂C₆H ₂)₂O (9)], and [PO(2,2′-(4,6-(tert-C₄H₉)₂-C ₆H₂)₂)O]₂[O(2,2′-(C ₆H₄)₂)O] (10) as ligands is studied with oct-1-ene and styrene as substrates. For oct-1-ene the highest normal to branched ratio obtained is 48 (5). For styrene the product selectivity depends strongly on the reaction temperature; a branched to normal ratio of 19 is found for 7 when T = 40 °C vs a branched to normal ratio of 0.19 for 1 when T = 120 °C. A bulky and bisequatorially (ee) coordinating diphosphite is required to obtain a high regioselectivity for linear aldehydes, while flexible diphosphites or equatorially-axially (ea) coordinating diphosphites lead to an enhancement of the formation of branched aldehydes. The hydroformylation of oct-1-ene has a first-order dependency in the oct-1-ene concentration, the order in CO is approximately -0.65, and the order in H₂ is approximately 0.2. This is consistent with a kinetic scheme in which alkene addition is the rate-determining step. The crystal structures of RhAcac(4) and RhH(CO)₂(4) (11), are presented. The hydrido ligand in 11 could not be located. The structure reveals a distorted TBP with 4 ee coordinated, the P(1)-Rh-P(2) angle being 115.95(9)°. The distortion is indicative of a crowded rhodium center which explains the obtained high linearity of the hydroformylation products.

Full text of DOI:10.1021/om950549k

Organometallics 1996, 15, 835-847
835
Bu lk y Dip h osp h ite-Mod ified Rh od iu m Ca ta lysts:
Hyd r ofor m yla tion a n d Ch a r a cter iza tion
Annemiek van Rooy,† Paul C. J . Kamer,† Piet W. N. M. van Leeuwen,*,†
Kees Goubitz,‡,§ J an Fraanje,‡ Nora Veldman,| and Anthony L. Spek|,
Van ’t Hoff Research Institute, Department of Inorganic Chemistry, University of Amsterdam,
Nieuwe Achtergracht 166, 1018 WV Amsterdam, The Netherlands, Amsterdam Institute
for Molecular Studies, Department of Crystallography, University of Amsterdam,
Nieuwe Achtergracht 166, 1018 WV Amsterdam, The Netherlands, and Bijvoet Center
for Biomolecular Research, Vakgroep Kristal- en Structuurchemie, Utrecht University,
Padualaan 8, 3584 CH Utrecht, The Netherlands
Received J uly 18, 1995X
The rhodium-catalyzed hydroformylation with the diphosphites P[O(2,2′-(4-X-6-Y-C6H2)2O]-
[O(2,2′-(4-X-6-Y-C6H2)2)OP][O(4-Z-C6H4]2 (X ) Y ) tert-butyl, Z ) H (1), Z ) OMe (2), Z )
C6H5 (3), Z ) Cl (4); X ) OMe, Y ) tert-butyl, Z ) H (5)), [PO(2,2′-(4,6-(tert-C4H9)2C6H2)2)O]2R
[R ) O(CH2)2O (6), R ) O(CH2)3O (7)], [PO(2,2′-(C6H4)2)O]2R, R ) O(2,2′-(4-MeO-6-tert-
C4H9C6H2)2O (8), R ) O(2,2′(4,6-tert-C4H9)2C6H2)2O (9)], and [PO(2,2′-(4,6-(tert-C4H9)2-
C6H2)2)O]2[O(2,2′-(C6H4)2)O] (10) as ligands is studied with oct-1-ene and styrene as
substrates. For oct-1-ene the highest normal to branched ratio obtained is 48 (5). For styrene
the product selectivity depends strongly on the reaction temperature; a branched to normal
ratio of 19 is found for 7 when T ) 40 °C vs a branched to normal ratio of 0.19 for 1 when
T ) 120 °C. A bulky and bisequatorially (ee) coordinating diphosphite is required to obtain
a high regioselectivity for linear aldehydes, while flexible diphosphites or equatorially-
axially (ea) coordinating diphosphites lead to an enhancement of the formation of branched
aldehydes. The hydroformylation of oct-1-ene has a first-order dependency in the oct-1-ene
concentration, the order in CO is approximately -0.65, and the order in H2 is approximately
0.2. This is consistent with a kinetic scheme in which alkene addition is the rate-determining
step. The crystal structures of RhAcac(4) and RhH(CO)2(4) (11), are presented. The hydrido
ligand in 11 could not be located. The structure reveals a distorted TBP with 4 ee
coordinated, the P(1)-Rh-P(2) angle being 115.95(9)°. The distortion is indicative of a
crowded rhodium center which explains the obtained high linearity of the hydroformylation
products.
In tr od u ction
and easy to synthesize. Recently we found that the
rhodium catalyst precursor modified by a bulky mono-
phosphite increased the reactivity enormously, but a
moderate regioselectivity was obtained.4 From patents
and recent literature5 it is known that bulky diphos-
phites give rise to very selective hydroformylation cat-
alysts. We suggested earlier that the regioselectivity
in the hydroformylation of unsubstituted 1-alkenes is
Since Wilkinson’s discovery that through addition of
phosphorus ligands rhodium-catalyzed hydroformyla-
tion can be performed at lower temperatures and
pressures,1 much research has been done to develop
ligands that increase the rate and the regioselectivity
of the reaction. During the last several years it has
become apparent that these improvements can be
effected with bidentate ligands, and research has con-
centrated on chelating ligands such as diphosphines,
ether-phosphane ligands, and diphosphinite ligands.2
Chiral diphosphines proved to be applicable in the
rhodium-catalyzed enantioselective hydroformylation of
prochiral substrates.3 A general problem of these
ligands is their sensitivity toward oxidation. Phosphite
ligands, however, are relatively inert toward oxidation
(2) (a) Casey, C. P.; Whitteker, G. T.; Melville, M. G.; Petrovich, L.
M.; Gavney, J . A., J r.; Powell, D. R. J . Am. Chem. Soc. 1992, 114, 5535.
(b) Consiglio, G.; Rama, F. J . Mol. Catal. 1991, 66, 1. (c) Del Zotto,
A.; Costella, L.; Mezzetti, A.; Rigo, P. J . Organomet. Chem. 1991, 414,
109. (d) Trzeciak, A.; Ziolkowski, J . J . J . Organomet. Chem. 1991, 420,
353. (e) Alper, H.; Zhou, J .-Q. J . Chem. Soc., Chem. Commun. 1993,
317. (f) Lindner, E.; Norz, H. Chem. Ber. 1990, 123, 459. (g) Lindner,
E.; Glaser, E. J . Organomet. Chem. 1990, 391, C37. (h) Selke, R. J .
Organomet. Chem. 1989, 370, 241. (i) Valderama, M.; Scott, M.;
Abugoch, L. J . Coord. Chem. 1990, 21, 55. (j) Pottier, Y.; Mortreux,
A.; Petit, F. J . Organomet. Chem. 1989, 370, 333. (k) Baker, M. J .;
Pringle, P. G. J . Chem. Soc., Chem. Commun. 1993, 314. (l) Devon,
T. J .; Phillips, G. W.; Puckette, T. A.; Stavinoha, J . L.; Vanderbilt, J .
J . U.S. Pat. 4 694 109, 1989 (to Eastman Kodak). (m) Unruh, J . D.;
Christenson, J . R. J . Mol. Catal. 1982, 14, 19.
(3) (a) Salomon, C.; Consiglio, G.; Botteghi, C.; Pino, P. Chimia 1973,
27, 215. (b) Stern, R.; Hirschauer, A.; Sajus, L. Tetrahedron Lett. 1973,
35, 3247. (c) Tanaka, M.; Ikeda, Y.; Ogata, I. Chem. Lett. 1975, 1115.
Hayashi, T.; Tanaka, M.; Ikeda, Y.; Ogata, I. Bull. Chem. Soc. J pn.
1979, 52, 2605. (d) Becker, Y.; Eisenstadt, A.; Stille, J . K. J . Org.
Chem. 1980, 45, 2145. (e) Gladiali, S.; Pinna, L. Tetrahedron:
Asymmetry 1990, 1, 693.
† Van’t Hoff Research Institute.
‡ Amsterdam Institute for Molecular Studies.
§ Address correspondence pertaining to the crystallographic study
of RhAcac(4) to this author.
| Bijvoet Center for Biomolecular Research.
Address correspondence pertaining to the crystallographic study
of 11 to this author.
X Abstract published in Advance ACS Abstracts, December 15, 1995.
(1) (a) Slaugh, L. H.; Mullineaux, R. D. U.S. Pat. 3 239 566, 1966.
(b) Eiseman, J . L. U.S. Pat. 3 290 379 1966. (c) Evans, D.; Osborn, J .
A.; Wilkinson, G. J . J . Chem. Soc. A 1968, 3133. Yagupski, G.; Brown,
C. K.; Wilkinson, G. J . J . Chem. Soc. A 1970, 1392. Brown, C. K.;
Wilkinson, G. J . J . Chem. Soc. A 1970, 2753.
(4) (a) Van Rooy, A.; Orij, E. N.; Kamer, P. C. J .; Van den Aardweg,
F.; Van Leeuwen, P. W. N. M. J . Chem. Soc., Chem. Commun. 1991,
1096. (b) Van Rooy, A.; Orij, E. N.; Kamer, P. C. J .; Van Leeuwen, P.
W. N. M. Organometallics 1995, 14, 34.
0276-7333/96/2315-0835$12.00/0 © 1996 American Chemical Society

836 Organometallics, Vol. 15, No. 2, 1996
van Rooy et al.
Ch a r t 1
caused by an equatorial-equatorial (ee) coordination of
both the phosphorus atoms.6 This coordination would
be induced by a natural bite angle of the diphosphite of
approximately 120°, as was hypothesized by Casey et
al. for diphosphine ligands.2a Recently, the first studies
of asymmetric rhodium-catalyzed hydroformylation with
chiral diphosphite ligands were published.7 Although
no data on the mechanistic and kinetic behavior of
catalysts modified by bidentate ligands are as yet
available, it is commonly accepted that the mechanism
for the cobalt-catalyzed hydroformylation as postulated
by Heck and Breslow8 can be applied to ligand-modified
rhodium carbonyl as well. It is also assumed that the
addition of the alkene to the hydrido-rhodium complex
or one of the early migration steps is the rate-limiting
step. Kinetic studies indicate the same for the triph-
enylphosphine-modified system.9 These studies are in
contrast with the persisting assumption10 that the H2
addition is the rate-determining step. This was found
by Wilkinson who studied the rhodium triphenylphos-
phine catalyst at low pressures (<1 bar).1c Under
higher pressures that are more comparable to real
process conditions the alkene addition becomes rate
limiting. Recently Gladfelter concluded that the regi-
oselectivity of the product aldehyde was not determined
during the alkene insertion, but these results were
obtained under mass transfer limiting conditions.11
We started a fundamental study of the influence of
the structure of various diphosphites on the rhodium-
catalyzed hydroformylation reaction. We tested some
of the diphosphites that were recently reported by the
Union Carbide Co. (1-10)5a (Chart 1) as modifying
ligands in the hydroformylation of oct-1-ene and styrene
and compared the rates and product selectivities. With
4 a kinetic study was performed involving the influence
of several reaction parameters on the rate and selectiv-
ity of the hydroformylation of oct-1-ene. We elucidated
the structure of several of the catalysts under process
conditions in order to find a relationship between the
structure and the catalytic and kinetic performance. To
10
gain information on possible intermediates and to
propose a mechanism for the formation of the active
catalyst complex, we studied the reaction of the precur-
sor and the ligands with and without addition of syngas
pressure and characterized the formed complexes spec-
troscopically.
Resu lts a n d Discu ssion
(5) (a) Billig, E.; Abatjoglou, A. G.; Bryant, D. R. U.S. Pat. 4 769
498, 1988 (to Union Carbide). (b) Baker, M. J .; Harrison, K. N.; Orpen,
A. G.; Pringle, P. G.; Shaw, G. J . Chem. Soc., Chem. Commun. 1991,
803. (c) Lorz, P. M. Eur. Pat. 0 472 071 A1, 1991. (d) Cuny, G. D.;
Buchwald, S. L. J . Am. Chem. Soc. 1992, 115, 2066-2068. (e) Moasser,
B.; Gladfelter, W. L. Presented at the 206th ACS National Meeting,
Chicago, IL, 1993; poster 371.
Hyd r ofor m yla tion of Oct-1-en e a n d Styr en e.
Oct-1-en e. The hydroformylation of oct-1-ene with
rhodium catalysts containing the bulky diphosphites
1-5 occurs with a regular reaction rate (average rate
≈ 2500 mol (mol of Rh)-1 h-1, T ) 80 °C, 0.4 mM Rh,
PP/Rh ) 20, 20 mmol of oct-1-ene in 20 mL of toluene;
see Table 1) and a very high selectivity for the linear
aldehyde (up to a normal/branched ratio of 48 for 5).
The amount of isomerized oct-1-ene varied from 0% (2-
4) to 23% (1). Variation of the para substituent on the
phenyl ring on the less bulky phosphorus atom (P2), in
order to modify the electron-donating ability, appears
to be of little or no influence on the rate and the
regioselectivity. Diphosphites 8 and 9, with only bulky
groups on the bridge between the two phosphorus
atoms, give rise to even more selective catalysts (no
branched aldehydes were detected for 8) although the
isomerization rate is high (18 and 26.5%, respectively).
Recently high isomerization rates were also reported by
Gladfelter11 under mass transfer limiting conditions.
The lack of isomerization found with 2-4 is probably a
consequence of a faster CO addition and subsequent
migration compared to â-H elimination under our reac-
tion conditions. Buchwald found no isomerization in the
rhodium-catalyzed hydroformylation of a variety of
(6) Van Leeuwen, P. W. N. M.; Buisman, G. J . H.; Van Rooy, A.;
Kamer, P. C. J . Recl. Trav. Chim. Pays Bas 1994, 1.
(7) (a) Buisman, G. J . H.; Kamer, P. C. J .; Van Leeuwen, P. W. N.
M. Tetrahedron: Asymmetry 1993, 4, 1625. (b) Babin, J . E.; Whiteker,
G. T. W. O. 93 03839, U.S. Pat. 911 518, 1992 (to Union Carbide Corp.).
(c) Buisman, G. J . H.; Vos, E.; Kamer, P. C. J .; Van Leeuwen, P. W. N.
M. J . Chem. Soc., Dalton Trans. 1995, 409. (d) Buisman, G. J . H. Ph.D.
Thesis, University of Amsterdam, 1995.
(8) Heck, R. F.; Breslow, D. S. J . Am. Chem. Soc. 1961, 83, 4023.
(9) (a) Cavalieri d’Oro, P.; Raimondt, L.; Pagani, G.; Montrasi, G.;
Gregorio, G.; Oliveri del Castillo, G. F.; Andreeta, A. Symp. Rhodium
Homogeneous Catal., Proc. 1978; 76-83. (b) Gregorio, G.; Montrasi,
G.; Tampieri, M.; Cavalieri d’Oro, P.; Pagani, G.; Andreetta, A. Chim.
Ind. 1980, 62 (5), 389. Cavalieri d’Oro, P.; Raimondi, L.; Pagani, G.;
Montrasi, G.; Gregorio, G.; Andreetta, A. Chim. Ind. 1980, 62 (7-8),
572. (c) Van Leeuwen, P. W. N. M.; Van Koten, G. Catalysis. An
Integrated Approach to Homogeneous, Heterogeneous and Industrial
Catalysis, 2nd ed.; Moulijn, J . A.; Van Leeuwen, P. W. N. M., Van
Santen, R. A., Eds.; Elsevier: Amsterdam, London, New York, Tokyo,
1995; pp 199-222.
(10) (a) Collman, J . P.; Hegedus, L. S.; Norton, J . R.; Finke, R. G.
Principles and Applications of Organotransition Metal Chemistry, 2nd
ed.; University Science Books: Mill Valley, CA, 1987. (b) Tolman, C.
A.; Faller, J . W. Homogeneous Catalysis with Metal Phosphine
Complexes; Pignolet, L. H., Ed.; Plenum Press: New York and London,
1983; Chapter 2, pp 88-89.
(11) Moasser, B.; Gladfelter, W. L. Organometallics 1995, 14, 3832.

Bulky Diphosphite-Modified Rh Catalysts
Organometallics, Vol. 15, No. 2, 1996 837
Ta ble 1. Resu lts of th e Hyd r ofor m yla tion of
Oct-1-en e w ith Rh (CO)2Aca c a n d Va r iou s
Dip h osp h ites a s Ca ta lyst P r ecu r sor sa
Ta ble 2. Resu lts of th e Hyd r ofor m yla tion of
Styr en e w ith Rh (CO)2Aca c a n d Va r iou s
Dip h osp h ites a s Ca ta lyst P r ecu r sor sa
product distribution (%)
normal brnached
con-
diphos- conversion aldehyde aldehyde
TOFb (mol
diphos- version isomerized
TOFf (mol
phite
(%)
(%)
(%)
(mol of Rh)-1 h-1
)
phite
(%)
oct-1-ene normal branched (mol of RH)-1 h-1
)
1
22
13
13
3
32
25
18
25
31
31
27
49
16
84
73
68
37
49
12
5
51
84
16
27
32
63
51
88
95
84
77
330
47
1
2
3
4
5
26
29
19
21
35
21
27
27
31
37
21
23
b
b
b
16
b
20
7
18
26.5
13
73
96
95
95
82.3
61.5
55
50
82
4
4
5
5
1.7
38.5
25
43
b
2450
2700
2275
3375
1750
11100
1550
130
1c
1d
d
6175 (16% ethylbenzene)
1075 (27% ethylbenzene)
350 (32% ethylbenzene)
320
320
3710
30
1890
47
d
2
3
6
6
7c
7d
8e
9
7c,e
7e
8c
3600
6120
520
16
23
72.1
48
1.4
39
10
a
Conditions: T ) 80 °C, 0.4 mM of Rh(CO)2Acac, diphosphite/
a
Conditions: T ) 80 °C, 0.4 mM of Rh(CO)2Acac, diphosphite/
Rh ) 20, PH ) PCO ) 10 bar, 20 mmol of styrene in 20 mL of
2
Rh ) 20, PH ) PCO ) 10 bar, 20 mmol of 1-octene in 20 mL of
2
b
toluene. Turnover frequencies for the formation of aldehydes
toluene. Not detected. c Diphosphite/Rh ) 2.5.
T ) 40 °C.
b
d
averaged over conversion. c T ) 40 °C. PCO ) 5 bar, PH ) 30
d
e Diphosphite/Rh ) 16. f Averaged over the conversion.
2
bar, T ) 120 °C. e Diphosphite/Rh ) 2.5.
alkenes with 8 as ligand, but these experiments were
performed at room temperature.5d Addition of the
diphosphites with flexible alkyl bridges, 6 and 7, results
for 6 in an increase of the reaction rate with a factor of
3 (11 100 mol (mol of Rh)-1 h-1), but the regioselectivity
decreases enormously (normal/branched (n/b) ) 1.6).
Diphosphite 10, less bulky than 1-5, and in contrast
with 6 and 7 containing a rigid bridging group, gives
both a low product linearity and a low reaction rate.
The rigid diphosphites show rates lower than most
of the monodentate phosphorus ligand systems (e.g.
PPh3, 5000 mol [mol of Rh]-1 h-1, 5% v/v in benzene,
0.5 mM Rh, PPh3/Rh ) 10, 90 °C, 20 bar syngas; tris-
(2-tert-butyl-4-methylphenyl) phosphite, 40 000 mol (mol
of Rh)-1 h-1, 20 mmol of oct-1-ene in 20 mL of toluene,
0.1 mM Rh, P/Rh ) 50 (P ) tris(2-tert-butyl-4-meth-
ylphenyl) phosphite), T ) 80 °C),4b but they react faster
than other bidentate ligands (DPPE (DPPE ) 1,2-
bis(diphenylphosphino)ethane), 100 mol (mol of Rh)-1
h-1, n/b ) 1, 0.4 mM Rh,12 DPPE/Rh ) 10, T ) 80 °C,
20 bar CO/H2; BISBI (BISBI ) 2,2′-bis(diphenylphos-
Rh)-1 h-1 for T ) 40 °C), and we obtained very low
selectivities for the formation of the branched 2-phe-
nylpropanal with the rigid diphosphites 1-4 and 8, the
branched to normal ratio being almost 1. Lowering the
reaction temperature did increase the formation of the
branched aldehyde, but the regioselectivity remained
moderate. Apparently, use of these diphosphites leads,
as also found for oct-1-ene, toward an increase of the
selectivity to the linear aldehydes. Thus, when higher
temperatures were used, a large excess of the linear
aldehydes was now observed; at a reaction temperature
of 120 °C, a selectivity of 84% for the linear phenylpro-
panal was found (hydrogenation not taken into account,
the highest selectivity for the formation of 2-phenylpro-
panal reported so far is 70%13,15).
Due to its specific electronic properties styrene is
known to show a natural preference for forming the
branched (1-phenylethyl)rhodium compound and hence
a high selectivity for the branched aldehyde. For
bidentate ligands high b/n ratio’s were often reported.7,16
However, application of the rigid diphosphites gives rise
to a catalyst with a different selectivity. The pressure-
and temperature-dependent product distribution sug-
gests that the branched rhodium complex, when formed,
undergoes â-H elimination and re-forms the starting
alkene and hydridorhodium complex. From Lazzaro-
ni’s work15 it is known that branched alkylrhodium
intermediates are more sensitive toward â-H elimina-
tion than the linear alkyl complexes; especially for
styrene this difference in sensitivity is large. This is
often suggested to be caused by the stabilization of an
(η3-styryl)rhodium intermediate;10b although these in-
termediates were already demonstrated for several
metals,17 it has not been until recently that the first (η3-
styryl)rhodium complex was isolated.18 Wink16 argued
that in his rhodium diphosphite systems η3-styryl
species do not play a role, although he obtained high
b/n ratios (97% branched aldehyde), which is in concert
with the small size of the unsubstituted bisnaphthol
diphosphites.
phinomethyl)(1,1′-biphenyl)), 850 mol (mol of Rh)-1 h-1
,
n/b ) 80, 1.78 mM Rh, BISBI/Rh ) 2.2, 10 bar, T ) 80
°C, 0.01 mol of oct-1-ene13). The rigid diphosphites react
with equal (BISBI) or higher (DPPE) selectivity to the
linear product. The catalysts with the flexible diphos-
phites react with rates and selectivities equal to those
of the triphenylphosphine-modified system.14 Hydro-
genation of alkenes above the detection limit (0.1%) was
not observed in any of the studied catalyst systems.
Comparison of diphosphites 6 and 7 with 1-5 and 8
and 9 suggests that the rigid and sterically demanding
bridge between the phosphorus atoms is responsible for
the high regioselectivity rather than the large substit-
uents at the phosphorus atoms.
Styr en e. In the hydroformylation of styrene with
ligands 1-4 (Table 2), the reaction rates are generally
a factor of 10 lower than those for oct-1-ene (average
rates of ≈300 mol (mol of Rh)-1 h-1, T ) 80 °C, 20 mmol
of styrene in 20 mL of toluene; around 50 mol (mol of
(12) De Lange, W. G. J .; Kamer, P. C. J .; Van Leeuwen, P. W. N.
M. Unpublished results (8 µmol of Rh(CO)2Acac, 80 µmol of DPPE, 20
mmol of oct-1-ene, 20 bar of CO/H2, 80 °C, in 20 mL of toluene: TOF
) 100 mol (mol of Rh)-1 h-1; n/b ) 1).
(13) Kranenburg, M.; Van der Burgt, Y. E. M.; Kamer, P. C. J .; Van
Leeuwen, P. W. N. M. Organometallics 1995, 14, 3081.
(15) A 65% value was reported by: Lazzaroni, R.; Raffaelli, A.;
Settambolo, R.; Bertozzi, S.; Vituli, G. J . Mol. Catal. 1989, 50, 1.
(16) Kwok, T. J .; Wink, D. J . Organometallics 1993, 12, 1954.
(17) Crascall, L. E.; Litster, S. A.; Redhouse, A. D.; Spencer, J . L. J .
Organomet. Chem. 1990, 394, C35.
(14) Van Rooy, A.; De Bruijn, J . N. H.; Roobeek, K. F.; Kamer, P. C.
J .; Van Leeuwen, P. W. N. M. J . Organomet. Chem. 1995, in press.
(18) Scha¨fer, M.; Mahr, N.; Wolf, J .; Werner, H. Angew. Chem. 1993,
32, 1315.

838 Organometallics, Vol. 15, No. 2, 1996
van Rooy et al.
F igu r e 1. Plot of ln(TOF) vs the reaction time. Condi-
F igu r e 2. Dependency of the reaction rate on the H2
pressure of the hydroformylation of oct-1-ene. Condi-
tions: T ) 60 °C, pCO ) 10 bar, 0.4 mM Rh(CO)2Acac, 4/Rh
) 10, 20 mmol of oct-1-ene in 20 mL of toluene.
tions: T ) 60 °C, PH ) PCO ) 10 bar, 0.4 mM Rh(CO)2-
2
Acac, 4/Rh ) 10, 20 mmol of oct-1-ene in 20 mL of toluene.
As low CO pressures and high temperatures facilitate
â-H elimination, â-H elimination eventually dominates
hydroformylation of the branched alkylrhodium com-
plex. Since formation of the linear (2-phenylethyl)-
rhodium remains irreversible at low pressures and high
temperatures, the product selectivity shifts toward the
normal aldehyde as we indeed observed (see Table 2).
The use of diphosphites 6 and 7 with an aliphatic,
flexible bridging chain as auxiliary ligands resulted in
higher rates (3710 mol (mol of Rh)-1 h-1 for 6) and
higher regioselectivities for the branched aldehydes: a
branched to normal ratio up to 19 for the five-membered
bridge (7). This is in agreement with the preference for
the formation of the branched (1-phenylethyl)rhodium
intermediate when there are no steric constraints.
Cycloh exen e. When we tried to hydroformylate
cyclohexene with the bulky diphosphite-modified cata-
lyst, higher pressures were required to reach any
conversion at all, but even under these conditions
reaction rates remained very low (TOF ) 150 mol (mol
F igu r e 3. Dependency of the reaction rate on the CO
pressure of the hydroformylation of oct-1-ene. Condi-
tions: T ) 60 °C, PH ) 10 bar, 0.4 mM Rh(CO)2Acac, 4/Rh
) 10, 20 mmol of oc2t-1-ene in 20 mL of toluene.
enylphosphine-modified hydroformylation of propene
under real process conditions (i.e. 2-70 bar, 90-110 °C),
found that the reaction rate increased with increasing
alkene concentration and observed a very low order in
the dependencies on both the CO and H2 concentration.
This different order in CO concentration can be ex-
plained by the nature of the phosphorus ligands; in the
triphenylphosphine system either a molecule of CO or
a phosphine ligand can dissociate (depending on the
amount of phosphine ligands coordinated) to allow the
addition of the substrate. At higher CO pressures the
reaction rate hardly decreases implying that a phos-
phine ligand is replaced by the alkene molecule. This
is also supported by a negative order in the PPh3
concentration.20 As follows from the large negative
order in the CO concentration in the diphosphite-
modified system, a CO molecule dissociates after which
the alkene can coordinate. In mononuclear systems
with HRh(CO)4 and HRh(CO)3P as the active catalysts,
the rate expression for 1-alkenes is different from that
for catalysts with diphosphites or PPh3.4b,21 Here, a
zeroth-order dependency in the substrate concentration
and a first-order dependency with respect to the H2
pressure were observed. This means that the rate-
limiting step is now the reaction of the rhodium acyl
complex with H2.
of Rh)-1 h-1, 1, T ) 80 °C, PCO ) PH ) 40 bar). A bulky
2
and monodentate ligand is required to obtain acceptable
rates; e.g., with tris(2-tert-butyl-4-methylphenyl) phos-
phite as a ligand a rate of 512 mol (mol of Rh)-1 h-1
was obtained (T ) 80 °C, PCO ) PH ) 10 bar, initial
2
cyclohexene concentration ) 0.91 M).4b
Kin etic Stu d y. We decided to study the influence
of the reaction parameters with one of the diphosphite
ligands that gave a high selectivity for the linear
aldehyde in the hydroformylation of oct-1-ene. We
collected kinetic data for the reaction of the ligand 4
modified catalyst, and the results are given in Figures
1-3. As can be concluded from the plot of the ln(TOF)
vs reaction time (Figure 1), the reaction is first order
with respect to the oct-1-ene concentration. The rate
constant shows almost no dependency on the H2 pres-
sure (≈0.2) (Figure 2) and a clear negative order in the
CO concentration. The order in CO showed to be -0.65
(Figure 3). This suggests that the first step in the
reaction cycle, the exchange of a CO ligand for a
π-coordinated alkene, is rate-limiting.19
These results are almost equal to the kinetic expres-
sion of the PPh3-modified catalyst. In this active
hydridorhodium complex, at least two phosphine ligands
are coordinated to the rhodium metal. Cavalieri d’Oro,9a,b
one of the few who studied the kinetics of the triph-
Another remarkable result of the diphosphite-modi-
fied catalyst is the decrease in selectivity for the linear
(19) We have not attempted to find an analytic expression for the
rate in view of the large number of steps involved. See ref 4b for
analytical derivations for part of the reaction scheme.
(20) J ardine, F. Polyhedron 1982, 569.
(21) Marko, L. Aspects of homogeneous catalysis; Ugo, R., Ed.;
Reidel: Dordrecht and Boston, 1973; Vol. 2, Chapter 1.

Bulky Diphosphite-Modified Rh Catalysts
Organometallics, Vol. 15, No. 2, 1996 839
Sch em e 1. Hyd r ofor m yla tion vs Isom er iza tion
Ta ble 3. In flu en ce of th e CO P r essu r e on th e n /b
Ta ble 4. 31P NMR Da ta for
Rh (Aca c)(CO)(d ip h osp h ite) a n d
Rh (Aca c)(d ip h osp h ite) (Ben zen e-d 6)
Ra tio a n d th e Am ou n t of Isom er iza tion a
normal
aldehyde
(%)
branched
aldehyde
(%)
δ(31P1)
J RhP1
(Hz)
δ(31P2)
(ppm)
J RhP2
(Hz)
J P1P2
(Hz)
isomerization
(%)
PCO
(ppm)
5
10
20
30
40
87
87
84
78
75
3
6
10
15
19
10
7
6
7
6
Rh(Acac)(CO)(diphosphite)
1
126.2
126.8
127.5
295
296
296
141.0
141.2
141.1
6
6
6
2a
4
Rh(Acac)(diphosphite)
1a
2a
4b
6
134.8
135.7
134.0
144.8
146.7
301
298
296
307
296
130.3
131.6
130.8
309
311
284
110
107
109
a
Conditions: T ) 60 °C, 0.4 mM Rh(CO)2Acac, 4/Rh ) 10, PH
) 10 bar, 20 mmol of 1-octene in 20 mL of toluene.
2
nonanal at increasing CO pressure (see Table 3). The
normal to branched ratio drops from 29 at PCO ) 5 bar
to 4 at PCO ) 40 bar, while the isomerization decreases
from 10% to 6%.
8
a
b
Toluene-d8. CDCl3.
(≈20 ppm) and its rhodium-phosphorus coupling con-
stant of ≈296 Hz. The chemical shift of the noncoor-
dinating phosphorus has changed very slightly com-
pared to the free ligand, and the coupling constant
between P1 and P2 equals that of the free diphosphite
(6 Hz). In the IR spectrum, the two CO stretch
vibrations of Rh(CO)2Acac (2080, 2012 cm-1) are re-
placed by one vibration at 2015 cm-1. Data for several
of these complexes are given in Table 4.
For the nonsymmetric bulky diphosphites only after
heating or evacuating the mixture, the second CO ligand
is eventually substituted, resulting in a 31P NMR
spectrum that is characteristic of a square planar
complex albeit with large coupling constants between
the phosphorus atoms (see Table 4). This coupling
constant is relatively high for a complex with the
phosphorus ligands oriented in a cis fashion. The only
reference we are aware of presents a lower value (70
Hz).22 The IR spectrum shows no CO absorptions
(complex B).
Isomerization of the alkene only occurs when â-H
elimination of the branched alkylrhodium compound
takes place (see Scheme 1); â-H elimination of the linear
alkylrhodium species re-forms the 1-alkene. This â-H
elimination reaction, facilitated at low CO pressures,
converts the branched alkylrhodium into 2-alkenes.
Thus, at low pressures, a substantial amount of the
branched alkylrhodium complex does not proceed to
form the branched aldehydes and consequently (at
incomplete conversions) the n/b ratio increases. How-
ever, the decreasing isomerization rate at high CO
pressures cannot totally account for the loss of selectiv-
ity. The presence of rhodium complexes that contain
only one or even no rhodium-phosphorus bonds can
explain these results as such species are expected to
react with low selectivity and high rate. In addition,
the relatively high rate of these complexes compared to
complexes with bidentate diphosphites can disturb the
reaction kinetics.
Ca ta lyst Ch a r a cter iza tion Exp er im en ts. Rea c-
tion s of Rh (CO)2Aca c. In order to identify the cata-
lyst complex and the possible intermediates during its
formation, we added 1 equiv of 4 to the Rh(CO)2Acac
precursor. Immediately, one CO ligand is substituted
with the least sterically hindered phosphorus atom (A)
For complex RhAcac(4), an X-ray structure was
determined (Figure 4; Tables 5, 6). The figure shows a
square planar surrounded rhodium. The P(1)-Rh-P(2)
angle of 97.5(1)° is large compared to the analogous
triphenyl phosphite complex23 (94.8(2)°) and demon-
(22) Aigen, S.; Van Eldik, R. Organometallics 1987, 6, 1080.
(23) Leipoldt, J . G.; Lamprecht, G. J .; Van Zyl, G. J . Inorg. Chim.
Acta 1985, 96, 31.
as is concluded from the upfield shift of its δ(31P) value

840 Organometallics, Vol. 15, No. 2, 1996
van Rooy et al.
Ta ble 5. Cr ysta llogr a p h ic Da ta for Rh Aca c(4) a n d 11
compound
Rh(acac)(4)
11
Crystal Data
formula
Mn
cryst system
space group
a, Å
C73H95Cl2O8P2Rh‚1.5C2H5OH
1405.41
triclinic
P1h (No. 2)
17.485(4)
18.838(2)
14.220(1)
110.16(1)
107.97(1)
87.98(1)
C70H89Cl2O8P2Rh
1294.23
monoclinic
C2/c (No. 15)
40.632(6)
13.943(3)
23.645(5)
90
92.420(15)
90
13384(4)
1.29
b, Å
c, Å
R, deg
â, deg
γ, deg
V, Å3
4169(1)
1.12
2
D
Z
calc, g cm-3
8
F(000)
1490
30.3 (Cu KR)
0.15 × 0.40 × 0.45
5456
4.3 (Mo KR)
0.50 × 0.25 × 0.15
µ cm-1
cryst size, nm
Data Collection
T, K
250
150
θSET4 range, deg
40-42
10.1-13.9
θ
min, θmax deg
2.5, 65.5
1.5, 23.5
wavelength, Å
scan type
X-ray exposure time, h
linear decay, %
ref refln
1.541 84 (graphite monochr)
ω/2θ
172
0.710 73 (graphite monochr)
ω
28
8
0
11h0, 011h
0:20, -22:22, -16:15
14 136
14 136
0.62-1.67
17 3 3, 6 4 11, 11 5 9
-45:45, 0:15, 0:26
10 633
9873
0.834-1.180
Data set
tot. no. of data
tot. unique data
DIFABS transm range
Refinement
1066
0.087 [9433, I > 2.5σ(I)]
no. of refined params
final R1a [no. of data]
final wR2b [no. of data]
final Rwc [no. of data]
S
773
0.082 [4807, Fo > 4σ(Fo)]
0.164 [9873]
0.116 [9433]
1.04
0.935
ω-1 d
[6.2 + 0.0129(σ(Ro))2 + 0.0002/(σ(Ro))]
σ2(F2) + (0.0509P)2
0.008, 0.742
-0.62, 0.52
(∆/σ)av, (∆/σ)max
0.04, 0.95
-1.4, 1.3
min and max residual density, e Å-3
a
b
2
d
R1 ) ∑||Fo| - |Fc||/∑|Fo|. wR2 ) [∑[w(Fo - Fc2)2]/∑[w(Fo2)2]]1/2
.
c Rw ) [∑[w(||Fo| - |Fc||)2]/∑[w(Fo2)]]1/2
.
P ) (max(Fo2, 0) + 2Fc2)/3.
Ta ble 6. Selected Bon d Len gth s (Å) a n d Bon d
An gles (d eg) of Rh Aca c(4) (Esd ’s in P a r en th eses)
Distances
Rh-P(1)
Rh-P(2)
2.134(3)
2.170(3)
Rh-O(1)
Rh-O(2)
2.053(8)
2.056(8)
Angles
P(1)-Rh-P(2)
P(1)-Rh-O(1)
P(1)-Rh-O(2)
P(2)-Rh-O(1)
P(2)-Rh-O(2)
O(1)-Rh-O(2)
Rh-P(1)-O(3)
Rh-P(1)-O(4)
Rh-P(1)-O(5)
O(3)-P(1)-O(4)
O(3)-P(1)-O(5)
O(4)-P(1)-O(5)
Rh-P(2)-O(6)
97.5(1)
171.5(2)
88.8(3)
86.6(2)
173.8(3)
87.2(3)
117.0(3)
115.0(3)
124.1(3)
103.0(5)
90.8(4)
Rh-P(2)-O(7)
Rh-P(2)-O(8)
O(6)-P(2)-O(7)
O(6)-P(2)-O(8)
O(7)-P(2)-O(8)
P(1)-O(3)-C(6)
P(1)-O(4)-C(12)
P(1)-O(5)-C(18)
P(2)-O(6)-C(32)
P(2)-O(7)-C(46)
P(2)-O(8)-C(60)
Rh-O(1)-C(1)
Rh-O(2)-C(3)
109.7(3)
114.1(2)
98.7(4)
99.2(4)
102.0(4)
129.1(8)
130.3(9)
124.9(7)
140.8(6)
123.4(8)
119.9(8)
129.2(7)
125.8(8)
102.8(5)
129.3(3)
RhAcacP2, where P2 ) bis(phenyl 3,3′,5,5′-tetra-tert-
butyl-1′,1′-biphenyl-2,2′-diyl phosphite) (99.87(3)°).24 The
Rh-P distances (2.1549(9) and 2.1566(9) Å) and the
Rh-O distances (2.070(2) and 2.074(2) Å) in the above
mentioned bis(phosphite)-derived complex are similar
to those in RhAcac(4) (Rh-P(1) ) 2.134(3) and 2.170-
(3) Å, Rh-O(1) ) 2.053(8), Rh-O(2) ) 2.056(8)). The
F igu r e 4. ORTEP plot of RhAcac(4) at the 50% probability
level. Hydrogen atoms and the minor disorder contribution
were left out for clarity.
(24) J ongsma, T.; Challa, G.; Van Leeuwen, P. W. N. M. J .
Organomet. Chem. 1991, 421, 121. Meetsma, A.; J ongsma, T.; Challa,
G.; Van Leeuwen, P. W. N. M. Acta Crystallogr. 1993, C49, 1160.
strates the steric hindrance of the diphosphite. This
enlarged P-Rh-P angle was previously observed for

Bulky Diphosphite-Modified Rh Catalysts
Organometallics, Vol. 15, No. 2, 1996 841
F igu r e 5. 31P[1H] NMR spectrum (300 MHz, benzene-d6)
1
of product formed after pressurizing (PCO/H ) 10 bar) and
F igu r e 6. Hydride region of measured H NMR spectrum
2
heating (T ) 60 °C) Rh(CO)2Acac + 4.
(100 MHz, benzene-d6, lower) and simulated spectrum
(upper) of product formed after pressurizing (PCO/H ) 10
2
O(1)-Rh-O(2) angle of 87.2(3)° is somewhat smaller
than that in the RhAcacP2 complex (89.41(9)°) and in
the triphenyl phosphite analogue (88.8(2)°). The distor-
tion is also noticeable from the various, very large
P-O-C angles that are present. A molecular mechan-
ics study of P-O-C angles of Cr(CO)5L complexes (with
L ) phosphite)25 revealed that the average angles in
the chromium complex are larger than in the free
ligand. They increase with increasing bulkiness of the
phosphite with a maximum of 127.5° for L ) P(O-t-Bu)3.
In RhAcac(4) a very large angle of 140.8(6)° was
determined for P(2)-O(6)-C(32), possibly meaning that
steric hindrance has led to distortions within the
diphosphite. Triphenylphosphine exclusively yields
monophosphine complexes when reacted with Rh(Acac)-
(CO)2, even when a large excess of phosphine is used.26
The chelating effect on coordination of the two phos-
phorus atoms and the high π-acidity, leading to forma-
tion of strong coordination bonds, might explain that
the complex is formed despite the large steric hindrance
of the diphosphite. It also clarifies the extreme condi-
tions required to substitute the second CO ligand of Rh-
(CO)2Acac. One of the tert-butyl groups is disordered
(C(38)-C(41).
bar) and heating (T ) 60 °C) Rh(CO)2Acac + 4.
is observed (Figure 6). Repeating this experiment with
5 equiv of 4 resulted in formation of a mixture of 1 equiv
of the diphosphite complex and 4 equiv of the free
ligand.
We followed this reaction with in-situ infrared spec-
troscopy. After mixing of the diphosphite with the Rh-
(CO)2Acac precursor, instantaneously A was observed.
After pressurization of the mixture with 10 bar of
syngas, a large amount of Rh(CO)2Acac was re-formed,
of which the CO vibrations disappeared on the appear-
ance of two new vibrations in the carbonyl region.
These IR and NMR data point toward a five-coordinated
Rh(I) complex as the final complex containing a hydrido
ligand and the diphosphite coordinated in a bidentate
way. The reported formation of dimeric bimetallic
species at higher temperatures11 was not observed
under our conditions. No infrared signals in the bridg-
ing CO region (1800-1900 cm-1) could be detected, and
both 31P and 1H NMR showed only the signals that could
be assigned to the monomeric rhodium hydride complex.
The active catalyst complex that appeared to be quite
air stable was prepared in the autoclave (see Experi-
mental Section) and was characterized under air. Com-
plexes with the other diphosphites were prepared in this
way, and the obtained spectroscopic data are shown in
Tables 7 and 8. For some of the hydridorhodium
complexes 103Rh NMR was measured as well, and these
data confirmed that each Rh center is coordinated to
two phosphorus atoms.
The symmetric diphosphites give rise to chelating
coordination immediately upon mixing or after stirring
for several minutes. For some of the diphosphites a side
product is obtained that has both phosphorus atoms
coordinated to the rhodium center and that still shows
a CO vibration. These types of complexes, containing
a 5-coordinated rhodium center, were previously de-
scribed by Buisman et al.7c,d
Hyd r id or h od iu m Com p lexes of Non sym m etr ic
Dip h osp h ites. Simulation of the NMR data showed
that the complicated spectra obtained with the nonsym-
metric diphosphites 1-5 are caused by two phosphorus
atoms that have accidentally almost equal chemical
shifts, although they are chemically inequivalent. They
show different hydrogen-phosphorus coupling con-
stants which are in the same order of magnitude as the
difference in the phosphorus chemical shift values. The
values of the 31P chemical shifts point toward an ee
coordination of the diphosphite ligand, but the hydrogen-
phosphorus coupling constants resemble neither a cis
When performing the same experiment under syngas
pressure (in a high-pressure NMR tube, P ) 12 bar, 1:1
CO/H2), we also see formation of A, but after heating of
the mixture, a complex of type B is not observed; the
31P-NMR spectrum now shows two 31P signals with two
different Rh-P coupling constants (243, 226 Hz) indica-
tive of two coordinating phosphorus atoms (see Figure
5) and a TBP surrounding of the rhodium. In the
1
hydride region of the H NMR a second-order multiplet
(25) Caffrey, M. L.; Brown, T. L. Inorg. Chem. 1991, 30, 3907.
(26) Bonati, F.; Wilkinson, G. J . Chem. Soc. A 1964, 3156.

842 Organometallics, Vol. 15, No. 2, 1996
van Rooy et al.
Ta ble 7. NMR Da ta (300.12 MHz) for Rh H(CO)2(d ip h osp h ite) (Ben zen e-d 6) Obta in ed a fter
Sim u la tion (GeNMR)
δ(1H)
δ(103Rh)
diphosphite
δ(31P) (ppm) (ppm) δ(13C) (ppm) (ppm)
coupling consts (Hz)
1
160.6 -10.35 -1073 J RhP1 ) 240, J RhP2 ) 233, J PP ) 250, J HP1 ) -70, J HP2 ) 23, J RhH ) 3
1 (13CO complex) 160.8, 160.6 -10.31 193.7, 191.9
J RhP1 ) 230, J RhP2 ) 234, J PP ) 250, J HP1 ) -73, J HP2 ) 18, J RhH
2.5, |J PC1| ) 8, |J PC2| ) 42, J RhC1 ) 53, J RhC2 ) 67.5
)
2a
3a
4
156.0, 155.7
160.9, 160.7 -10.58
160.5, 159.8 -10.40
J RhP1 ) 234, J RhP2 ) 228, J PP ) 250, J HP1 ) -50, J HP2 ) 96, J RhH ) 3
-1070 J RhP1 ) 235, J RhP2 ) 222, J PP ) 225, J HP1 ) -45, J HP2 ) 93, J RhH ) 3
-1070 J RhP1 ) 237, J RhP2 ) 226, J PP ) 170, J HP1 ) -19, J HP2 ) 70, J RhH ) 3.5
4 (13CO complex) 160.5, 159.8 -10.40 193.5, 191.9
J RhP1 ) 237, J RhP2 ) 226, J PP ) 170, J P1C1 ) -2.5, J P1C2 ) 47, J P2C1
14, J P2C2 ) 38, J RhCl ) 54, J RhC2 ) -65
)
5b
6
7
8
9
162.5, 161.2 -10.61
J RhP1 ) 211, J RhP2 ) 235, J PP ) 115, J HP1 ) 82, J HP2 ) -10, J RhH ) 3
J RhP ) 213, J PH ) 72, J RhH ) 6.5
161.3
171.0
174.5
173.0
151.6
-9.72
-9.80
J RhP ) 226, J PH ) 5, J RhH ) 3.5
-10.81 194.7
-10.30
-10.97
J RhP ) 239, J PH ) 4, J RhH ) 3.5
J RhP ) 234, J PH ) 9, J RhH ) 3
10
J RhP ) 244, J PH ) 4, J RhH ) 5
a
b
Toluene-d8. Acetone-d6.
Ta ble 8. IR F r equ en cies in th e M-H a n d CO
Str etch in g Region of Hyd r id or h od iu m
Dip h osp h ite Com p lexes
frequencies in the hydrido complex are combinations of
two CO ligands and one hydrido ligand. The rhodium-
hydride vibration disappears upon deuterating the
complex as the rhodium-deuteride vibration is situated
in the fingerprint region. The large frequency shift of
the highest energy absorption is indicative of a trans
hydrido-CO relation.28 In solution IR, the rhodium
hydride vibration and the lowest energy CO vibration
overlap resulting in two absorptions. Combining the
spectroscopic experiments, we can conclude that the
catalyst complex contains a chelating diphosphite, a
hydrido ligand, and two CO ligands coordinated to the
rhodium center. Although the coupling constants of the
phosphorus atoms with the hydrido ligand do not lead
to straightforward conclusions, from IR spectroscopy
and the nearly identical 31P chemical shifts and phos-
phorus-rhodium coupling constants, which indicated
two nearly identical phosphorus atoms, we assume that
the diphosphite is bisequatorially coordinated (ee) and
the complex has structure C.
diphosphite
1
ν (cm-1
)
2033, 1995a
2030, 1994, 1988b
1988, 1949c (13CO complex)
2031, 1992a
2
3
4
2031, 1990c
2049, 1966a
2035, 1998, 1990b
2058, 2012b (deuteride complex)
1994, 1957c (13CO complex)
2028, 1998a
5
6
8
2028, 1987a
2074, 2013a
10
2070, 2008a
In acetone-d6. As a Nujol mull. c In benzene-d6.
a
b
nor a trans relation of the phosphorus atoms toward the
hydrido ligand.27 In Table 7 it is also shown that, going
from 1 and 5, the difference in the chemical shift
between the phosphorus atoms increases. For 1 and 4
we have prepared the catalyst complexes with 13CO. The
HRh(1)(13CO)2 spectrum shows two virtual double trip-
lets in the carbonyl region. J RhC1 ) 53 Hz and J RhC2
)
67.5 Hz (|J PC1| ) 8 Hz, |J PC2| ) 42 Hz). From the 31P
NMR data it was concluded that, in this ligand, of all
nonsymmetric ones, the phosphorus atoms have the
smallest difference in chemical shift (<0.1 ppm). Thus,
the two phosphorus-carbon coupling constants give rise
to virtual triplets. On the contrary, HRh(4)(13CO)2
shows a first-order 13C spectrum with two different CO
ligands coupled by two unidentical phosphorus atoms.
In the simulated 31P NMR spectrum the difference in
chemical shift is 0.7 ppm. From these data it becomes
clear that in these complexes the CO ligands do not
exchange their positions. Most likely, on the NMR time
scale the phosphorus ligands retain their positions as
well.
IR spectra of the complexes that could be obtained as
powders were measured as Nujol mulls (see for IR data
Table 8). It was found that three absorptions were
present in the CO-stretching region instead of two
observed with IR in solution. Upon measurement of
DRh(CO)2(4), only two absorptions remained (2058,
2012 cm-1), which were shifted compared to HRh(CO)2-
(4) (2035, 1998, 1990 cm-1). This implies that the three
This structure is confirmed by the X-ray determina-
tion of HRh(CO)2(4)29 (11) (Tables 5, 9, 10; Figure 7).
The crystal structure shows a distorted TBP geometry
around the rhodium with both the phosphorus atoms
in equatorial positions. The P(1)-Rh-P(2) angle is
115.95(9)°. The hydrido ligand could not be found. One
CO ligand is situated equatorially, and the other CO
molecule and the hydrido ligand occupy axial positions.
The angles of the phosphorus atoms with the axial CO
are 92.1(3)° for P(1)-Rh-C and 104.2(3)° for P(2)-Rh-
C, meaning that CO(ax) deviates from the molecular
axis. The rhodium atom is situated out of the equatorial
plane (0.395(1) Å), having moved somewhat toward the
axial CO ligand. This distortion explains the relatively
high hydrogen-phosphorus coupling constants that
were determined by simulation of the NMR spectra;
apparently the H-Rh-P angles are larger than 90°
resulting in larger coupling constants than expected for
(28) Vaska, L. J . Am. Chem. Soc. 1966, 88, 4100.
(29) Van Rooy, A.; Kamer, P. C. J .; Van Leeuwen, P. W. N. M.;
Veldman, N.; Spek, A. L. J . Organomet. Chem. 1995, 494, C15.
(27) Hyde, E. M.; Swain, J . R.; Verkade, J . G.; Meakin, P. J . Chem.
Soc., Dalton Trans. 1976, 1169.

Bulky Diphosphite-Modified Rh Catalysts
Organometallics, Vol. 15, No. 2, 1996 843
of 11, despite the sterically more demanding and also
equatorially coordinating triphenylphosphine ligand. In
the Ir complex the P-Ir-P angle is 117.9(1)°, more
similar to the P-Rh-P angle of 11. Complex 11 has a
more distorted structure than the BISBI compounds
indicative of more steric crowding.
The ee coordination of BISBI to the rhodium follows
from a natural bite angle of approximately 120°. In
order to obtain high regioselectivities ee coordination
and hence a 120° bite angle is required. The structure
of 11 confirms this suggestion.
Hyd r id or h od iu m Com p lexes of Sym m etr ic Di-
p h osp h ites. The symmetric diphosphites 6-9 all give
rise to hydridorhodium complexes that show clear first-
order spectra in which the phosphorus atoms are
equivalent on the NMR time scale at room temperature.
The hydridorhodium complexes derived from 7-9 show
very small P-H coupling constants (4-9 Hz) indicative
of ee-coordinated complexes. On cooling of HRh(CO)2-
(8), the phosphorus atoms lose their time averaged
equivalence, and at T ) 205 K the slow exchange
situation is reached, resulting in a strongly coupled
F igu r e 7. ORTEP plot of 11 at the 30% probability level.
Hydrogen atoms and the minor disorder contribution were
left out for clarity.
Ta ble 9. Selected Bon d Len gth s (Å) a n d Bon d
An gles (d eg) of 11 (Esd ’s in P a r en th eses)
NMR spectrum (31P NMR δ (ppm) 177.3, 175.4, J RhP1
)
Distances
Rh-P(1)
Rh-P(2)
2.255(3)
2.239(3)
Rh-C(69)
Rh-C(70)
1.878(12)
1.938(9)
238, J RhP2 ) 232, J PP ) 205 Hz (acetone-d6)). However,
the P-H coupling constant remains small (not detect-
able as a consequence of broadening of signals at low
temperatures). The existence of two inequivalent phos-
phorus atoms can be explained by hindered rotation of
the bridge between the phosphorus atoms. The two
aromatic rings can rotate freely at higher temperatures,
which results in two equivalent phosphorus atoms. At
low temperatures this process is slowed down and the
two P’s become inequivalent on the NMR time scale.
The room-temperature 13C NMR spectrum of the 13CO
derived hydridorhodium complex of 8 gives rise to a
double triplet in accordance with two equivalent phos-
phorus atoms and two equivalent CO ligands meaning
that here the CO ligands are in fast exchange.
Angles
P(1)-Rh-P(2)
P(1)-Rh-C(69)
P(2)-Rh-C(69)
P(1)-Rh-C(70)
P(2)-Rh-C(70)
C(69)-Rh-C(70)
Rh-P(1)-O(1)
Rh-P(1)-O(2)
Rh-P(1)-O(3)
O(1)-P(1)-O(2)
O(1)-P(1)-O(3)
O(2)-P(1)-O(3)
Rh-P(2)-O(4)
115.95(9)
115.2(4)
118.5(3)
92.1(3)
104.2(3)
105.8(4)
118.1(2)
110.4(2)
124.2(2)
104.6(3)
97.8(3)
Rh-P(2)-O(5)
119.3(2)
117.5(2)
94.6(3)
101.5(3)
99.3(3)
117.3(5)
123.0(4)
117.9(5)
126.5(5)
123.1(5)
124.7(5)
174.2(9)
176.4(8)
Rh-P(2)-O(6)
O(4)-P(2)-O(5)
O(4)-P(2)-O(6)
O(5)-P(2)-O(6)
P(1)-O(1)-C(1)
P(1)-O(2)-C(12)
P(1)-O(3)-C(29)
P(2)-O(4)-C(40)
P(2)-O(5)-C(57)
P(2)-O(6)-C(63)
Rh(1)-C(69)-O(8)
Rh(1)-C(70)-O(7)
98.6(3)
120.2(2)
a complex with the phosphites in a pure cis arrange-
ment. Alternatively, the complex can be seen as having
a distorted tetrahedral geometry with the hydrido
ligand on one of the faces. The Rh-C-O angles show
a slight deviation from 180° (174.2 and 176.4°). The
dihedral angles of the diphenyl rings are 60.6(4)° for
C(29)-C(34) with C(35)-C(40) and 61.9(4)° for C(1)-
C(6) with C(7)-C(12), respectively. One of the tert-butyl
substituents (C(17), C(18), C(19), C(20)) is disordered.
This is the first crystal structure of a hydridorhodium
diphosphite complex reported to our knowledge. In the
square planar complex derived from 4, RhAcac(4), the
Rh-P distances (2.134(3) and 2.170(3) Å) are shorter
than those in 11 (2.239(3) and 2.255(3) Å).
The crystal structure of 11 shows much similarities
with those of the recently crystallographically analyzed
Ir(BISBI)(CO)2H and HRh(CO)(PPh3)(BISBI) (BISBI )
2,2′-bis(diphenylphosphinomethyl)(1,1′-biphenyl)) re-
ported by Casey et al.2a Except for the oxygen vs carbon
atom, BISBI and 4 have similar bridges between both
phosphorus atoms. The hydroformylation of hex-1-ene
with the rhodium catalyst derived from BISBI yields a
normal to branched ratio of 50, which was also ascribed
to bisequatorial coordination of the diphosphine accord-
ing to the authors. The Rh-P distances (2.285(1) and
2.318(1) Å) of HRh(CO)(PPh3)(BISBI) are similar to the
Rh-P distances (2.239(3) and 2.255(3) Å) of 11, and the
P-Rh-P angle is 124.8(1)°, which is larger than that
The hydridorhodium complex of diphosphite 7 does
not form easily, and the precursor requires 8 h instead
of 3 h under syngas pressure. Diphosphite 6 forms the
HRh(CO)2(6) complex very slowly as well (8 h required),
and the reaction is accompanied with much decomposi-
tion as is shown by signals at ≈10 ppm in the 31P NMR,
originating from H-phosphonates formed by hydrolysis.
The structure of this hydrido complex is different,
because, in this complex, the phosphorus atoms are
equatorially axially (ea) coordinated as was concluded
from NMR at low temperatures. At room temperature
the phosphorus atoms are in fast exchange resulting in
one doublet caused by the coupling with the rhodium
1
center. The H NMR hydride region contains a virtual
double triplet with a phosphorus hydride coupling
constant of 72 Hz. We were not able to reach the slow-
1
exchange region by cooling, but in the H NMR at 200
K we did observe a broad doublet (J PH ) 180 Hz; the
signal is too broad to observe other couplings, ∆ν1/2
)
15 Hz) indicative of a trans relation of one phosphorus
atom toward the hydrido ligand. We reacted HRh(CO)-
(PPh3)3 with 6 in order to prepare a more stable
hydrido-rhodium complex,2a HRh(CO)(PPh3)(6) (see
Experimental Section). Indeed a stable complex was
obtained, and from the coupling constants of the hydrido
ligand of 6 with the phosphorus atom from the PPh3
ligand (13 Hz) and different couplings constant of the

844 Organometallics, Vol. 15, No. 2, 1996
van Rooy et al.
Ta ble 10. F in a l Atom ic Coor d in a tes a n d Equ iva len t Isotr op ic Th er m a l P a r a m eter s (Å2) for 11
(Esd ’s in P a r en th eses)a
atom
x
y
z
U(eq)
atom
x
y
z
U(eq)
Rh(1)
Cl(1)
Cl(2)
P(1)
P(2)
O(1)
O(2)
O(3)
O(4)
O(5)
O(6)
O(7)
O(8)
C(1)
C(2)
C(3)
C(4)
C(5)
0.17325(2)
0.16020(6)
0.33688(5)
0.13069(5)
0.16663(5)
0.12871(13)
0.13078(12)
0.09210(12)
0.13918(12)
0.19644(12)
0.15983(12)
0.1584(2)
0.2407(2)
0.1534(2)
0.1765(2)
0.2014(2)
0.2026(2)
0.1767(2)
0.1518(2)
0.1226(2)
0.1043(2)
0.0766(2)
0.0696(2)
0.0871(2)
0.1129(2)
0.1776(2)
0.1500(2)
0.2107(2)
0.1737(3)
0.2310(2)
0.2631(3)
0.2617(5)
0.2407(4)
0.2362(6)
0.2216(4)
0.2257(7)
0.0536(2)
0.0671(2)
0.0489(2)
0.0204(2)
0.0760(2)
0.0733(2)
0.0411(2)
0.18205(5)
0.6768(2)
0.3179(2)
0.0797(2)
0.3279(2)
0.0176(4)
0.0015(4)
0.1109(3)
0.3985(4)
0.4062(4)
0.3340(4)
0.2064(5)
0.0909(7)
-0.0502(6)
-0.0439(6)
-0.1116(6)
-0.1900(6)
-0.1991(6)
-0.1316(5)
-0.1490(6)
-0.2309(6)
-0.2513(6)
-0.1879(6)
-0.1046(6)
-0.0853(5)
0.0293(6)
0.19841(3)
-0.05226(10)
0.15785(12)
0.18458(9)
0.15996(9)
0.1266(2)
0.2339(2)
0.1851(2)
0.1819(2)
0.1671(2)
0.0918(2)
0.3236(3)
0.1943(4)
0.1192(3)
0.0780(3)
0.0796(4)
0.1178(4)
0.1522(4)
0.1536(3)
0.1885(4)
0.1779(3)
0.2081(3)
0.2512(3)
0.2644(3)
0.2294(3)
0.0276(4)
0.0262(4)
0.0292(4)
-0.0266(3)
0.1209(4)
0.0972(8)
0.1398(13)
0.1834(5)
0.0606(7)
0.0915(8)
0.1600(12)
0.1927(4)
0.1470(4)
0.2449(4)
0.1714(4)
0.3143(3)
0.3658(3)
0.3006(4)
0.0312(3)
0.0538(12)
0.0556(10)
0.0211(7)
0.0239(7)
0.0248(17)
0.0211(17)
0.0175(17)
0.0217(17)
0.0260(17)
0.0256(17)
0.054(3)
0.107(4)
0.022(3)
0.028(3)
0.030(3)
0.030(3)
0.027(3)
0.020(2)
0.026(3)
0.019(3)
0.025(3)
0.022(3)
0.022(2)
0.021(3)
0.039(3)
0.046(3)
0.055(4)
0.052(4)
0.030(3)
0.064(6)
0.061(9)
0.063(6)
0.058(9)
0.066(7)
0.077(11)
0.025(3)
0.040(3)
0.036(3)
0.035(3)
0.029(3)
0.041(3)
0.49(4)
C(28)
C(29)
C(30)
C(31)
C(32)
C(33)
C(34)
C(35)
C(36)
C(37)
C(38)
C(39)
C(40)
C(41)
C(42)
C(43)
C(44)
C(45)
C(46)
C(47)
C(48)
C(49)
C(50)
C(51)
C(52)
C(53)
C(54)
C(55)
C(56)
C(57)
C(58)
C(59)
C(60)
C(61)
C(62)
C(63)
C(64)
C(65)
C(66)
C(67)
C(68)
C(69)
C(70)
0.0988(2)
0.0800(2)
0.0633(2)
0.0576(2)
0.0645(2)
0.0763(2)
0.0844(2)
0.0915(2)
0.0690(2)
0.0701(2)
0.959(2)
0.0361(6)
0.1750(6)
0.1387(5)
0.2050(5)
0.3019(6)
0.3351(5)
0.2717(5)
0.3130(6)
0.2944(5)
0.3396(5)
0.4043(5)
0.4266(6)
0.3785(6)
0.0389(6)
-0.0228(6)
-0.0183(6)
0.0559(6)
0.3666(6)
0.3405(6)
0.3528(6)
0.4709(6)
0.3262(6)
0.4232(6)
0.2583(7)
0.2895(6)
0.4999(6)
0.4464(6)
0.5566(6)
0.5739(6)
0.3820(6)
0.4241(6)
0.4058(7)
0.3431(6)
0.3001(6)
0.3212(7)
0.4194(6)
0.4994(6)
0.5782(7)
0.5771(7)
0.4975(7)
0.4170(7)
0.1255(8)
0.2004(7)
0.3337(4)
0.1424(3)
0.0936(3)
0.0512(3)
0.0547(3)
0.1074(3)
0.1521(3)
0.2097(3)
0.2507(3)
0.3035(3)
0.3131(3)
0.2751(3)
0.2234(3)
0.0845(4)
0.1383(4)
0.0395(4)
0.0661(3)
0.0033(3)
-0.0262(4)
-0.0376(3)
0.0199(4)
0.3484(3)
0.3599(4)
0.3294(4)
0.4031(4)
0.2907(3)
0.3040(4)
0.3436(4)
0.2448(4)
0.1636(4)
0.2017(4)
0.1996(4)
0.1599(4)
0.1214(4)
0.1231(3)
0.0598(4)
0.0741(4)
0.0386(4)
-0.0085(4)
-0.0220(4)
0.0129(4)
0.1989(4)
0.2772(4)
0.043(3)
0.018(2)
0.021(3)
0.023(3)
0.024(3)
0.019(2)
0.016(2)
0.017(2)
0.019(3)
0.022(3)
0.022(3)
0.022(3)
0.021(3)
0.026(3)
0.034(3)
0.036(3)
0.032(3)
0.026(3)
0.041(3)
0.039(3)
0.043(3)
0.028(3)
0.042(3)
0.045(3)
0.047(3)
0.024(3)
0.033(3)
0.034(3)
0.036(3)
0.027(3)
0.035(3)
0.040(3)
0.037(3)
0.032(3)
0.031(3)
0.028(3)
0.039(3)
0.041(3)
0.040(3)
0.041(3)
0.034(3)
0.056(4)
0.043(3)
0.1195(2)
0.1166(2)
0.0477(2)
0.0467(2)
0.0647(2)
0.0112(2)
0.0567(2)
0.0243(2)
0.0849(2)
0.0549(2)
0.0444(2)
0.0286(2)
0.0167(2)
0.0606(2)
0.1472(2)
0.1795(2)
0.1382(2)
0.1521(2)
0.2296(2)
0.2511(2)
0.2841(2)
0.2954(2)
0.2738(2)
0.2403(2)
0.1605(2)
0.1432(2)
0.1435(2)
0.1611(2)
0.1792(2)
0.1790(2)
0.2152(3)
0.1635(2)
C(6)
C(7)
C(8)
C(9)
C(10)
C(11)
C(12)
C(13)
C(14)
C(15)
C(16)
C(17)
C(18A)*
C(18B)*
C(19A)*
C(19B)*
C(20A)*
C(20B)*
C(21)
C(22)
C(23)
C(24)
C(25)
C(26)
C(27)
0.1041(6)
0.0814(7)
-0.0296(7)
-0.2596(6)
-0.2233(13)
-0.2011(17)
-0.2860(13)
-0.2989(19)
-0.3537(10)
-0.3457(16)
-0.3340(5)
-0.3986(6)
-0.3969(6)
-0.2929(6)
-0.0461(6)
-0.1112(6)
-0.0045(7)
a
1
U(eq) )
/ of the trace of the orthogonalized U. Starred atom sites have a population less than 1.0.
3
hydrido with the phosphorus atoms of 6 (-191 and 13
Hz) it became clear that also in this complex diphosphite
6 is ea coordinated. Likewise, in HRh(CO)(PPh3)(4) the
diphosphite was found to coordinate in an ee way as in
the dicarbonyl complexes (small hydrogen-phosphorus
coupling constants: 20, 9, 4 Hz).
structure as intermediate. Meakin31 described an ex-
change process which involves a tetrahedral surround-
ing of the metal. The flexible ea-coordinating diphos-
phites may be able to coordinate with varying P-Rh-P
angles, giving rise to fast exchange, while the diphos-
phites with rigid biphenyl backbones show less flex-
ibility.
Str u ctu r e-Selectivity Rela tion sh ip s. In this last
section we will try to explain the catalyst behavior in
relation to the structural information obtained. First
we have to ascertain that under the highly diluted
conditions of the catalytic experiments the same hydrido
species are formed that we observed with NMR and IR
spectroscopy. From the work of J ongsma32 on the bulky
monophosphite tris(2-tert-butyl-4-methylphenyl) phos-
It appears that at room temperature in the ee-
coordinated complexes of the nonsymmetric diphosphi-
tes the phosphorus atoms and the CO ligands retain
their positions in the TBP surrounding rhodium (on the
NMR time scale), while the ea complexes easily undergo
structural isomerization, as in the HRh(CO)2(6) com-
pound and several analogous complexes published by
Buisman.7c,d From his work it was concluded that
ligands with a preference for ea coordination always
lead to more fluxional complexes than those leading to
ee coordination, even if these ea ligands are more rigid;
ea diphosphite complexes are intrinsically more flux-
ional. This can be explained by the flexibility of the
ligands, i.e. the feasible P-Rh-P angles, upon coordi-
nation.
(31) Meakin, P.; Muetterties, E. L.; J esson, J . P. J . Am. Chem. Soc.
1972, 94, 5271.
(32) J ongsma, T. Ph.D. Thesis, Rijksuniversiteit Groningen, 1992;
Chapter 6.
(33) (a) Beurskens, P. T.; Admiraal, G.; Beurskens, G.; Bosman, W.
P.; Gelder, R de; Israe¨l, R.; Smits, J . M. M. The DIRDIF-94 program
system; Crystallography Laboratory, University of Nijmegen: Nijmegen,
The Netherlands, 1994. (b) Beurskens, P. T.; Admiraal, G.; Beurskens,
G.; Bosman, W. P.; Garc´ıa-Granda, S.; Gould, R. O.; Smits, J . M. M.;
Smykalla, C. The DIRDIF program system; Technical report of the
Crystallography Laboratory; University of Nijmegen: Nijmegen, The
Netherlands, 1992.
Isomerization can occur via two mechanisms. The
pseudo Berry rotation30 requires a square-pyramidal
(30) Berry, R. S. J . Chem. Phys. 1960, 32, 933.

Bulky Diphosphite-Modified Rh Catalysts
Organometallics, Vol. 15, No. 2, 1996 845
Sch em e 2. P r op osed Equ ilibr iu m for th e Ca ta lyst
P r ecu r sor s a t High CO/H2 P r essu r es
1-ene. These results are similar to those of the system
containing DPPE which can only coordinate in an ea
fashion.
Con clu sion s
In the hydroformylation of oct-1-ene the rhodium
catalysts modified by the rigid diphosphites 1-5, 8 and
9 give high rates and selectivities. The high regiose-
lectivity is caused by steric repulsion of the diphosphites
that coordinate bisequatorially to the rhodium center.
The kinetics of the catalyst derived from Rh(CO)2Acac
and 4 resemble those of the extensively studied PPh3
system.9 The flexible diphosphites 6 and 7 do not give
satisfying results in the hydroformylation of oct-1-ene.
Although rates are comparable with PPh3 (the latter not
added in large excess), the selectivity to normal alde-
hydes is moderate.
For styrene the selectivity can be directed to the
formation of the linear or the branched aldehyde
depending on the reaction conditions. The rigid diphos-
phites applied at high temperatures and low CO pres-
sures give rise to a catalyst that forms the normal
aldehydes in excess. When an ea-coordinating or a
flexible diphosphite is used at low reaction tempera-
tures, the catalyst favors formation of the branched
2-phenylpropanal.
phite and the present results we learned that the
equilibria in Scheme 2 exist in the catalyst system.
Although neither A nor B has been detected during the
formation of the active catalyst, J ongsma found that a
10-fold amount of bulky phosphite tris(2-tert-butylphe-
nyl) phosphite was required to reach a total conversion
of Rh(COD)Acac to the active complex HRh(CO)3P (P
) tris(2-tert-butylphenyl) phosphite) although only one
phosphorus ligand can coordinate to the rhodium center.
The results of our in-situ infrared study are comparable
to those of J ongsma,32 and therefore, we also added a
10-fold amount of diphosphite to be certain that the
same species is formed. Further addition of diphosphi-
tes does not increase the rate any further, and we
conclude that all rhodium is converted to HRh(CO)2-
(diphosphite) (results not shown).
We conclude from the X-ray structure of HRh(CO)2-
(4) that the metal center is very crowded. As a
consequence, there is little space for the incoming
substrate. Unsubstituted 1-alkenes are the least steri-
cally hindered and add to the metal most easily, followed
by respectively 2-substituted 1-alkenes and internal
alkenes. Upon migration of the hydride the steric
interactions can be most efficiently diminished by the
formation of a linear alkyl-rhodium intermediate.
From the present results we cannot distinguish whether
π-coordination or hydride migration or a combination
thereof determines the regioselectivity, but the overall
result remains the same. Viewed in that light we can
understand our observations in both the hydroformy-
lation of oct-1-ene and styrene. For oct-1-ene as the
substrate a high product linearity will be obtained. For
styrene, using the first type of ligands (1-5, 8, 9), ee
coordinating and rigid, a reversed product selectivity,
viz. an excess of linear phenylpropanal, can be obtained.
The formation of the branched (1-phenylethyl)rhodium
compound is now suppressed resulting in an increase
in the formation of the linear aldehydes. At mild
conditions the observed regioselectivity is low. This
enlarged preference for 3-phenylpropanal can be ex-
ploited by the variation of reaction conditions. At higher
temperatures and/or low CO pressures the rate of â-H
elimination increases and eventually dominates the
hydroformylation reaction. â-H elimination nearly ex-
clusively occurs for the secondary (1-phenylethyl)-
rhodium complex and prevents the formation of the
branched aldehyde. The linear 2-phenylethyl, however,
is almost irreversably formed and can complete the
reaction cycle. Thus a higher n/b ratio is obtained.
Exp er im en ta l Section
Gen er a l Meth od s. All operations were performed under
argon by standard Schlenk techniques. Solvents were distilled
from sodium/benzophenone before use. NMR spectra were
recorded on a Bruker AC-100 spectrometer (1H, 103Rh (mea-
sured via reverse 2D NMR) or a Bruker AMX-300 spectrometer
(1H, 13C, 31P, 31P[1H]), and chemical shifts were reported
referenced to tetramethylsilane and H3PO4, respectively.
Infrared spectra were recorded on a Nicolet 510 FT-IR spec-
trophotometer. Mass spectroscopic data were obtained with
the J EOL J MS SX/SX102A four-sector mass spectrometer,
coupled to a J EOL MS-MP7000 data system. Gas-liquid
chromatography analyses were done using a DB-1 column and
a Carlo Erba GC 6000 Vega Series 2 chromatograph. The
compounds 1-5 were prepared by a published method.5a Rh-
(CO)2(pentane-2,4-dionate) was purchased from J ohnson Mat-
they, triphenyl phosphite was purchased from Aldrich, and
both were used as received. Column chromatography was
performed with silica gel 60 230-400 mesh obtained from
Merck. Syngas 3.0, CO 3.0, and H2 5.0 were purchased from
Praxair. Hydroformylation experiments were performed using
Rh(CO)2(pentane-2,4-dionate) and diphosphite as catalyst
precursor in a stainless steel 316 autoclave, equipped with a
magnetic stirring bar, a pressure transducer, a thermocouple,
and a sampling device. Decane was used as the internal
standard.
The autoclave, charged with solvent and catalyst precursor,
was brought under pressure and heated to the reaction
temperature. Upon addition of the substrate and the internal
standard, the reaction started immediately. During the reac-
tion, several samples were taken, quenched with P(OPh3) to
deactivate the catalyst and analyzed by GLC.
The in-situ infrared experiment was performed in an SS 316
55 mL autoclave equipped with IRTRAN windows and a
mechanical stirrer. Rh(CO)2Acac (10 mg, 3.88 × 10-6 mol) and
4 (0.042 g, 3.88 × 10-6 mol) were dissolved in 15 mL of
cyclohexane. After pressurizing and heating of the mixture,
the autoclave was placed in the infrared spectrometer, and
while the samples were stirred, the infrared spectra were
recorded.
In the second type of systems, in which an ea-
coordinating (6) or a flexible (7) diphosphite is used,
sufficient space is available for the substrate to form
linear as well as branched alkylrhodium intermediates
which results in high rates but low selectivities for oct-

846 Organometallics, Vol. 15, No. 2, 1996
van Rooy et al.
Syn th esis of Dip h osp h ite Liga n d s. P r ep a r a tion of 1,2-
Bis(((4,4′,6,6′-tetr a-ter t-bu tyl-2,2′-bisph en oxy)ph osph in o)-
oxy)eth a n e (6). A solution of PCl3 (0.43 mL, 4.93 mmol),
4,4′,6,6′-tetra-tert-butyl-2,2′-bisphenol (2 g, 4.88 mmol) and
triethylamine (1.4 mL, 10.10 mmol) in 40 mL of toluene was
stirred for 2 h and additionally refluxed for 6 h. The mixture
was filtered, and to the filtrate were added triethylamine (0.7
mL, 5.05 mmol) and ethane-1,2-diol (0.14 mL, 2.51 mmol).
After being stirred for 12 h, the mixture was filtered and the
solvent and excess of triethylamine were evaporated. The
white compound was recrystallized from toluene/acetonitrile.
Yield: 1.77 g of a white powder (1.89 mmol, 78%). 31P[1H]
NMR: δ 134.7 ppm (s) (CDCl3). 1H NMR: δ (ppm) 7.41 (d,
4H arom, J ) 2.40 Hz), 7.16 (d, 4H, arom, J ) 2.46 Hz), 3.80
(d, 4H, C2H4, J ) 3.72 Hz), 1.46 (s, 36H, o-C4H9), 1.34 (s, 36H,
p-C4H9) (CDCl3). Mp: 195-198 °C. Anal. Calcd for
trile, recrystallized from toluene/acetonitrile, and dried in
vacuo. Yield: 1.04 g (1.24 mmol, 67.5%). 31P[1H] NMR: δ
(ppm) 145.8 (s) (CDCl3). 1H NMR: δ (ppm) 1.37 (s, 30H,
o-C4H9, p-C4H9), 6.69-7.59 (m, 20H, arom) (CDCl3). Mp: 148-
150 °C.
P r ep a r a t ion of 2,2′-b is(((4,4′,6,6′-t et r a -ter t-b u t yl-2,2′-
bisp h en oxy)p h osp h in o)oxy)-1,1′-bip h en yl (10). A solution
of PCl3 (0.6 mL, 7.40 mmol), 4,4′,6,6′-tetra-tert-butyl-2,2′-
bisphenol (3 g, 7.32 mmol) and triethylamine (4 mL, 28.86
mmol) in 50 mL of toluene was stirred for 1 day. After
filtration 2,2′-bisphenol (0.68 g, 3.66 mmol) and triethylamine
(2 mL, 14.43 mmol) were added and the reaction mixture was
stirred overnight. The resulting suspension was filtered, and
after evaporation of the toluene and excess of NEt3 a crude
product was obtained by crystallization from toluene/acetoni-
trile. The product was purified by column chromatography
(eluent: 5% EtOAc/toluene). Yield: 1.00 g (0.94 mmol, 26%).
31P[1H] NMR: δ (ppm) 137.6 (s) (CDCl3). 1H NMR: δ (ppm)
7.39 (d, 4H arom, J ) 2.43 Hz), 7.27 (m, 2H arom), 7.20 (d,
4H arom, J ) 2.43 Hz), 7.069 (m, 2H arom), 6.89 (m, 4H arom),
1.39 (s, 36H, o-C4H9), 1.29 (s, 36H, p-C4H9) (CDCl3). Anal.
Calcd for C68H88O6P2: C, 76.80; H, 8.35. Found: C, 76.33; H,
8.40.
Syn th esis of Rh od iu m Dip h osp h ite Com p lexes. In -
Situ P r ep a r a tion of Rh (CO)Aca c(1). Rh(CO)2Acac (5 mg,
1.94 × 10-5 mol) and 1 (0.021 g, 1.94 × 10-5 mol) were placed
in an NMR tube, and 0.3 mL of benzene-d6 was added. The
product was formed immediately as CO evolved. 1H-NMR: δ
(ppm) 1.19, 1.22, 1.25, 1.27, 1.72, 1.84 (s, 6 × 9H, C4H9), 1.29
(s, 2 × 9H, C4H9 + 3H Acac CH3) (C6D6). 31P[1H] NMR data
are listed in Table 4.
P r ep a r a tion of Rh Aca c(1). Rh(CO)2Acac (50 mg, 1.94 ×
10-4 mol) and 1 (0.21 g, 1.94 × 10-4 mol) were stirred in 4 mL
of toluene for 8 h at 40 °C. The product was precipitated by
addition of 5 mL of MeOH recrystallized from toluene/MeOH
and dried in vacuum. Yield: 0.11 g (8.23 × 10-5 mol, 42%) of
a light yellow powder. 31P[1H] NMR data (CDCl3) are listed
in Table 4. 1H NMR: δ (ppm) 0.89, 1.12, 1.32, 1.36, 1.40, 1.42,
1.65, 1.67 (s, 8 × 9H, C4H9), (s, 2 × 9H, C4H9 + 3H Acac CH3),
1.18 (s, 3H Acac CH3), 5.06 (s, 1H, Acac-H), 6.37-7.53 (m, 16H,
arom) (CDCl3). 13C NMR: δ (ppm) 26.93, 26.28 (2×, Acac CH3),
31.69, 31.42, 32.12, 32.28, 33.47, 34.63 (18×, t-Bu CH3), 31.99
(6×, t-Bu CH3), 34.25, 34.94, 35.02, 35.42, 35.86, 36.25, 36.51,
37.11 (8×, t-Bu C), 100.11 (Acac CH), 121.23-130.50 (16×,
CH arom), 128.24-151.59 (20×, C arom), 184.39, 185.97 (2×,
Acac C) (CDCl3). Anal. Calcd for C73H97O8P2Rh: C, 69.18;
H, 7.72. Found: C, 69.51; H, 7.96.
C
58H84O6P2: C, 74.17; H, 9.02. Found: C, 73.95; H, 9.09.
P r ep a r a tion of 1,3-Bis(((4,4′,6,6′-tetr a -ter t-bu tyl-2,2′-
bisp h en oxy)p h osp h in o)oxy)p r op a n e (7). A solution of
PCl3 (0.6 mL, 7.40 mmol), 4,4′,6,6′-tetra-tert-butyl-2,2′-bisphe-
nol (3 g, 7.32 mmol) and triethylamine (4 mL, 28.86 mmol) in
50 mL of toluene was stirred for 1 day. After filtration
propane-1,3-diol (0.27 mL, 3.74 mmol) and triethylamine (2
mL, 14.43 mmol) were added and the reaction mixture was
stirred overnight. The mixture was filtered and concentrated
in vacuum. A crude product was obtained by crystallization
from toluene/acetonitrile. The product was purified by column
chromatography (eluent: 5% EtOAc/petroleum ether 60-80).
Yield: 0.84 g (0.88 mmol, 24%). 31P[1H] NMR: δ (ppm) 136.2
(s) (CDCl3). 1H NMR: δ (ppm) 7.41 (d, 4H, arom, J ) 2.25
Hz), 7.16 (d, 4H, arom, J ) 2.25 Hz), 3.80 (m, 6H, C3H6), 1.46
(s, 36H, o-C4H9), 1.35 (s, 36H, p-C4H9) (CDCl3). Mp: 152 °C.
Anal. Calcd for C59H86O6P2: C, 74.33; H, 9.10. Found: C,
74.40; H, 9.09.
P r ep a r a tion of 2,2′-Bis(((2,2′-bisp h en oxy)p h osp h in o)-
oxy)-3,3′-di-ter t-bu tyl-5,5′-dim eth oxy-1,1′-biph en yl (8). 2,2′-
Bisphenoxyphosphorus chloride was prepared according to a
modified literature procedure.7a 2,2′-Bisphenol (20 g, 10.8
mmol) was dissolved in 50 mL of THF and added dropwise to
a solution of PCl3 (10 mL, 112 mmol) and NEt3 (30 mL, 216
mmol) in 100 mL of THF at 0 °C. The mixture was allowed
to cool to room temperature and subsequently refluxed for 6
h. After the mixture was cooled to room temperature, the Et3-
NHCl salts were removed by means of filtration and the
solvent was removed by evaporation. The product was purified
by destillation (bp 120 °C/0.2 mmHg). Yield: 13.45 g (54
mmol, 50%) of a white powder. 31P[1H] NMR: δ (ppm) 179.8
(s) (CDCl3). 1H NMR: δ 7.54-7.29 (m).
A solution of 4,4′-dimethoxy-6,6′-di-tert-butyl-2,2′-bisphenol
(0.66 g, 1.84 mmol) and pyridine (1 mL, 19.09 mmol) in 25
mL of THF was added dropwise to a solution of 2,2′-bisphe-
noxyphosphorus chloride (0.92 g, 3.67 mmol) in 40 mL of THF
at -50 °C. The solution was stirred overnight at room tem-
perature. After filtration the solvent and the excess of pyridine
were evaporated, and on addition of acetonitrile, a white
precipitate was formed. The product was washed 2 times with
cold acetonitrile (10 mL) and dried in vacuum. Yield: 1.14 g
(1.45 mmol, 79%). 31P[1H] NMR: δ 146.5 ppm (s) (CDCl3). 1H
NMR: δ (ppm) 1.34 (s, 18H, C4H9), 3.81 (s, 6H, OCH3), 6.78-
7.44 (m, 20H, arom) (CDCl3). Mp: 116-119 °C. Anal. Calcd
for C46H44O8P2: C, 70.22; H, 5.64. Found: C, 69.60; H, 5.57.
P r ep a r a tion of HRh (d ip h osp h ite)(CO)2. In a typical
experiment, Rh(CO)2Acac (10 mg, 3.878 × 10-6 mol) and
diphosphite 1 (0.041 g, 3.85 × 10-6 mol) were brought into a
small glass sample bottle containing a magnetic stirrer. A 2
mL volume of deuterated benzene was added. The bottle was
brought into an autoclave, and after closure, the autoclave was
filled with 12 bar of CO/H2 and heated to 40 °C. After 3 h,
the autoclave was cooled to room temperature and the pressure
was released. The yellow/orange solution was brought into
an NMR tube. Elemental analysis could not be performed
because of rapid decomposition of the complex in vacuum.
In -Situ P r ep a r a tion of Rh H(CO)P P h 3(4). An NMR tube
was charged with RhH(CO)(PPh3)3 (10 mg, 0.011 mmol) and
1 equiv of 4 (12.5 mg, 0.011 mmol). A 0.5 mL volume of
deuterated benzene was added. The tube was heated to 40
°C and regularly shaken. After 1 h NMR was measured. Data
after simulation: 31P NMR δ (ppm) 160.9 (P1), 156.3 (P2), 34.1
(P3 (PPh3)), J RhP1 ) 250 Hz, J RhP2 ) 256 Hz, J RhP3 ) 140 Hz,
P r ep a r a tion of 2,2′-Bis(((2,2′-bisp h en oxy)p h osp h in o)-
oxy)-3,3′,5,5′-tetr a -ter t-bu tyl-1,1′-bip h en yl (9). A solution
of 4,4′,6,6′-tetra-tert-butyl-2,2′-bisphenol (0.75 g, 1.83 mmol)
and pyridine (1 mL, 19.09 mmol) in 25 mL of THF was added
dropwise to a solution of 2,2′-bisphenoxyphosphorus chloride
(0.92 g, 3.67 mmol) at 50 °C. Subsequently, the reaction was
allowed to cool down to room temperature and stirred for 2
days. The pyridine salts were removed by means of filtration,
and the solvent and excess pyridine were evaporated. The
product was obtained by precipitation on addition of acetoni-
1
J P1P2 ) 280 Hz, J P1P3 ) 177 Hz, J P2P3 ) 161 Hz; H NMR δ
(ppm) -10.42, J HP1 ) -5 Hz, J HP2 ) 20 Hz, J HP3 ) 9 Hz, J HRh
) 4 Hz.
In -Situ P r ep a r a tion of Rh H(CO)P P h 3(6). The same
procedure as for RhH(CO)PPh3(4) was followed: 31P NMR δ
(ppm) 166.3 (P1), 162.4 (P2), 41.3 (P3 (PPh3)), J RhP1 ) 296 Hz,
J RhP2 ) 195 Hz, J RhP3 ) 154 Hz, J P1P2 ) 56, J P1P3 ) 432 Hz,

Bulky Diphosphite-Modified Rh Catalysts
Organometallics, Vol. 15, No. 2, 1996 847
1
J P2P3 ) 38 Hz; H NMR δ (ppm) -5.8, J HP1 ) 13 Hz, J HP2
)
Table 7. Final atomic coordinates and equivalent isotropic
thermal parameters are listed in Table 9.
-191 Hz, J HP3 ) 13 Hz, J HRh ) 4 Hz.
A transparent, colorless crystal was mounted on a Linde-
mann glass capillary and transferred into the cold nitrogen
stream on an Enraf-Nonius CAD4-T diffractometer on a
rotating anode. Accurate unit-cell parameters and an orienta-
tion matrix were determined from the setting angles of 25
reflections (SET4).37 Reduced-cell calculations did not indicate
higher lattice symmetry.38 Data were corrected for Lp effects
and showed no decay for the three periodically measured
reference reflections. An empirical absorption/extinction cor-
rection was applied (DIFABS34 as implemented in PLATON39).
The structure was solved by automated Patterson methods and
subsequent difference Fourier techniques (DIRDIF-92).33b
Refinement on F2 was carried out by full-matrix least-squares
techniques (SHELXL-93);40 no observance criterium was ap-
plied during refinement. A disorder model is refined for one
of the tert-butyl groups (C(18), C(19), C(20)). All non-hydrogen
atoms were refined with anisotropic thermal parameters
except for the disordered tert-butyl group. Hydrogen atoms
were refined with a fixed isotropic thermal parameter amount-
ing to 1.5 or 1.2 times the value of the equivalent isotropic
thermal parameter of their carrier atoms, for the methyl
hydrogen atoms and the other hydrogen atoms, respectively.
On an enhanced difference map a plausible electron density
was located for the hydrido atom. It was, however, not possible
to refine it freely. The hydrido was placed at the initial
position and not refined. It must be stressed that with this
crystal structure determination it is not possible to give any
certainty about either the presence or the position of the
hydrido. Weights were optimized in the final refinement
cycles. The structure contains a small void of 33 Å (at 0.099,
0.214, 0.544); however, no significant residual density was
found in that area (PLATON/SQUEEZE).41 Neutral atom
scattering factors and anomalous dispersion corrections were
taken from ref 42.
Cr ysta lliza tion of Rh (Aca c)(4). Rh(Acac)(4) was pre-
pared as Rh(Acac)(1) and precipitated from MeOH. Dark
yellow crystals were obtained from a pentane/EtOH solution.
X-r a y Str u ctu r e Deter m in a tion of Rh (Aca c)(4).
A
yellow crystal with approximate dimensions 0.15 × 0.40 × 0.45
mm was used for data collection of an Enraf-Nonius CAD-4
diffractometer with graphite-monochromated Cu KR radiation
in ω-2θ scan mode. A total of 14 136 unique reflections were
measured within the range 0 e h e 20, -22 e k e 22, -16 e
l e 15; of these, 9433 were above the significance level of 2.5σ-
(I). The maximum value of (sin θ)/λ was 0.59 Å-1
. Two
reference reflections (11h0, 011h) were measured hourly and
showed 8% decrease during the 172 h collecting time, which
was corrected for. Unit-cell parameters were refined by a
least-squares fitting procedure using 23 reflections with 80 <
2θ < 84°. Corrections for Lorentz and polarization effects were
applied. The structure was solved by the PATTY/ORIENT/
PHASEX option of the DIRDIF94 program system.32a After
isotropic refinement one of the tert-butyl groups (C(38)-C(41))
had rather high temperature factors. It was possible to
distribute the three methyl groups (C(39)-C(41)) into two half-
occupied positions which remained isotropic during the entire
refinement. A ∆F synthesis revealed 9 peaks which were
interpreted as being three molecules of ethanol, one of the
solvents used during crystallization. During refinement of
these solvent molecules it became apparent that their oc-
cupancy factors should be lower than 1.0; a value of 0.5 was
taken for all three solvent molecules. No attempts were made
to determine the hydrogen atoms of the solvent molecules. All
hydrogen atom positions were calculated. Full-matrix least-
squares refinement on F, anisotropic for the non-hydrogen
atoms and isotropic for the hydrogen atoms restraining the
latter in such a way that the distance to their carrier remained
constant at approximately 1.0 Å and keeping their tempera-
ture factors fixed at U ) 0.15 Å2 (the hydrogen atoms of the
disordered tert-butyl group were kept entirely fixed at their
calculated positions), converged to R ) 0.087, Rw ) 0.116, (∆/
σ)max ) 0.95, and S ) 1.04. A weighting scheme w ) [6.2 +
0.0129(σ(Fo))2 + 0.0002/(σ(Fo))]-1 was used. An empirical
absorption correction (DIFABS34) was applied, with corrections
in the range 0.62-1.67. A final difference Fourier map
revealed residual electron density between -1.4 and 1.3 e Å-3
in the vicinity of the heavy atoms. Scattering factors were
taken from Cromer and Mann.35a The anomalous scattering
of Rh, Cl, and P was taken into account.35b All calculations
were performed with XTAL36 unless stated otherwise. In
Table 5 these data are listed. Final atomic coordinates and
equivalent isotropic thermal parameters are part of the
supplementary data which have been deposited at the Cam-
bridge Crystallographic Data Centre.
Geometrical calculations and the ORTEP illustration were
performed with PLATON;39 all calculations were performed
on a DEC5000/125.
Ack n ow led gm en t. This work was supported in part
(A.L.S. and N.V.) by the Netherlands Foundation of
Chemical Research (SON) with financial aid from the
Netherlands Organization for Scientific Research (NWO).
We thank Warner van Leeuwen and Hillart Wagen-
makers for their experimental assistance. We thank
J an-Meine Ernsting for the 103Rh-NMR measurements.
Su p p or tin g In for m a tion Ava ila ble: Tables listing X-ray
parameters, fractional atomic coordinates and U values for the
non-hydrogen and hydrogen atoms, anisotropic thermal pa-
rameters, and bond distances and angles for 4 and 11 (31
pages). This material is contained in many libraries on
microfiche, immediately follows this article in the microfilm
version of the journal, can be ordered from the ACS, and can
be downloaded from the Internet; see any current masthead
page for ordering information and Internet access instructions.
Cr ysta lliza tion of 11. Colorless crystals were obtained by
dissolving 10 mg of Rh(CO)2Acac and 44 mg of 4 in 2 mL of
benzene, bringing the autoclave with its contents at 10 bar to
60 °C, releasing the pressure after 3 h, and, when most of the
solvent and the obtained Hacac were evaporated, cooling the
autoclave. Mp: dec >180 °C. FD MS: m/e 1236, [M+ - 2CO
- H].
OM950549K
(37) Boer, J . L. D.; Duisenberg, A. J . M. Acta Crystallogr. 1984, A40,
C410.
X-r a y Str u ctu r e Deter m in a tion of 11. Crystal data and
details on data collection and refinement are presented in
(38) Spek, A. L. J . Appl. Crystallogr. 1988, 21, 578.
(39) Spek, A. L. Acta Crystallogr. 1990, A46, C34.
(40) Sheldrick, G. M. SHELXL-93 Program for crystal structure
refinement. University of Go¨ttingen, Germany, 1993.
(41) Spek, A. L. ACA Abstracts 1994, 22, 66.
(42) Wilson, A. J . C., Ed. International Tables for Crystallography;
Kluwer Academic Publishers: Dordrecht, The Netherlands, 1992; Vol.
C.
(34) Walker, N.; Stuart, D. Acta Crystallogr. 1983, A39, 158.
(35) (a) Cromer, D. T.; Mann, J . B. Acta Crystallogr. 1968, A24, 321.
(b) Cromer, D. T.; Liberman, D. J . J . Chem. Phys. 1970, 53, 1891.
(36) Hall, S. R.; Flack, H. D.; Stewart, J . M., Eds. XTAL3.2 Reference
Manual. Universities of Western Australia, Geneva and Maryland,
1992.