Syntheses and characterization of upper rim 1,2- And 1,3-diphosphinated calix[4]arenes and their corresponding 1,5-cyclooctadienylrhodium(I) complexes: Comparison of the catalytic hydroformylation properties of terminal alkenes

Article abstract of DOI:10.1021/om030095o

The 5,11- (1,2-isomer) and 5,17-bis(dialkylphosphino)-25,26,27,28-tetra-n- propoxycalix[4]arene (1,3-isomer) ligands (alkyl = Me, i-Pr) have been prepared and coordinated to Rh(COD) ⁺ fragments (COD = 1,5-cyclooctadiene). The ligands 5,17-bis(d

Full text of DOI:10.1021/om030095o

2862
Organometallics 2003, 22, 2862-2875
Syn th eses a n d Ch a r a cter iza tion of Up p er Rim 1,2- a n d
1,3-Dip h osp h in a ted Ca lix[4]a r en es a n d Th eir
Cor r esp on d in g 1,5-Cycloocta d ien ylr h od iu m (I)
Com p lexes: Com p a r ison of th e Ca ta lytic
Hyd r ofor m yla tion P r op er ties of Ter m in a l Alk en es
Franc¸ois Plourde, Karine Gilbert, J onathan Gagnon,† and Pierre D. Harvey*
De´partement de Chimie, Universite´ de Sherbrooke, Sherbrooke, P.Q., Canada, J 1K 2R1
Received February 11, 2003
The 5,11- (1,2-isomer) and 5,17-bis(dialkylphosphino)-25,26,27,28-tetra-n-propoxycalix[4]-
arene (1,3-isomer) ligands (alkyl ) Me, i-Pr) have been prepared and coordinated to Rh-
(COD)+ fragments (COD ) 1,5-cyclooctadiene). The ligands 5,17-bis(diphenylphosphino)-
11,23-dibromo-25,26,27,28-tetra-n-propoxycalix[4]arene and 5,11-bis(diphenylphosphino)-
25,26,27,28-tetra-n-propoxycalix[4]arene have been coordinated to M(COD)+ (M ) Rh, Ir)
and RhCl(CO) fragments, as well. On the basis of mass spectrometry and 31P NMR spin-
lattice relaxation time measurements (T1), all of the complexes are found to be dimers.
Molecular modeling provides evidence that ring stress favors the dimer over the monomer,
and the modeled structures for both 1,2- and 1,3-isomers have been corroborated by the
comparison of the photophysics of the Ir(COD)+ species at 77 K. The decrease in emission
lifetimes of the P2Ir(CdC)2+ lumophore in the presence of 1-hexene is more pronounced for
the 1,3-isomer, indicating reduced steric hindrance about the metallic center. The catalytic
hydroformylation of five terminal alkenes using all six Rh(COD)+ catalyst precursors has
been investigated under various conditions in order to extract the turnover frequencies (tof
as defined as the number of moles of products divided by the total number of moles of rhodium
used per hour) and the ratios of the product distribution (n/ i, normal vs internal). The
comparison of the data suggests that the basicity and the cone angle of the phosphines
influence the n/ i ratios. The tof’s are generally larger for the 1,2-series for 1-hexene, but
depend on the nature of the phosphine for styrene, vinyl acetate, vinyl benzoate, and vinyl
p-tert-butylbenzoate. All in all, these complexes are good catalysts with respect to literature
data, notably those containing the (i-Pr2P)- groups. During the course of this study, one
1
hydride carbonyl complex has been synthesized and characterized from H and 31P NMR,
IR, and T1 measurements.
In tr od u ction
cone angle and bulkiness,4 the diphosphine natural bite
angle,5 the chirality,6 and the diphosphine electronic
effects.7 With respect to diphosphine bite angle, angles
approaching 90° and 120° favor coordination at the
The syntheses and investigations of calixarenes and
related derivatives continue to be a very active field.1
Although applications of such molecules for cation
extractions, biomimetics, luminescence probes, sensors,
and nuclear waste treatment are well known,1b their
uses in homogeneous catalysis are more limited, con-
centrating mainly on lower rim functionalized calix[4]-
arenes.1a,2 Recently, numerous works in the area of
rhodium-catalyzed hydroformylation reactions3 have
focused on parameters influencing the mechanism and
the product distribution (n/i ratio) such as the phosphine
(2) See for examples: (a) Kunze, C.; Selent, D.; Neda, I.; Schmutzler,
R.; Spannenberg, A.; Bo¨rner, A. Heteroat. Chem. 2001, 12, 577. (b)
Parlevliet, F. J .; Kiener, C.; Fraanje, J .; Goubitz, K.; Lutz, M.; Spek,
A. L.; Kamer, P. C. J .; van Leeuwen, P. W. N. M. J . Chem. Soc., Dalton
Trans. 2000, 1113. (c) Cobley, C. J .; Ellis, D. D.; Orpen, A. G.; Pringle,
P. G. J . Chem. Soc., Dalton Trans. 2000, 1109. (d) Dieleman, C.; Steyer,
S.; J eunesse, C.; Matt, D. J . Chem. Soc., Dalton Trans. 2001, 2508. (e)
Steyer, S.; J eunesse, C.; Matt, D.; Welter, R.; Wesolek, M. J . Chem.
Soc., Dalton Trans. 2002, 4264.
(3) (a) van Leeuwen, P. W. N. M.; Claver, C. Rhodium Catalysed
Hydroformylation; Kluwer Academic Publishers: Dordrecht, 2000. (b)
Ungva`ry, F. Coord. Chem. Rev. 1999, 188, 263. (c) Ungva`ry, F. Coord.
Chem. Rev. 1998, 170, 245. (d) Ungva`ry, F. Coord. Chem. Rev. 1997,
167, 233. (e) Beller, M.; Cornils, B.; Frohning, C. D.; Kohlpaintner, C.
W. J . Mol. Catal. A 1995, 104, 17.
(4) See for instance: (a) Trzeciak, A. M.; Glowiak, T.; Grzybek, R.;
Ziolkowski, J . J . J . Chem. Soc., Dalton Trans. 1997, 1831. (b) 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. (c)
Agbossou, F.; Carpentier, J . F.; Mortreux, A. Chem. Rev. 1995, 95,
2485.
* To whom correspondence should be addressed. Fax: (819) 821-
8017. Tel: (819) 821-7092 or (819) 821-8000, ext. 2005. E-mail:
pharvey@USherbrooke.ca.
† Current address: De´partement de Chimie, Universite´ du Que´bec
a` Rimouski.
(1) (a) Asfari, Z.; Bo¨rner, V.; Harrowfield, J .; Vicens, J . Calixarenes
2001; Kluwer Academic Publishers: Dordrecht, 2001. (b) Gutsche, C.
D. Calixarene Revisited; Royal Society of Chemistry: Cambridge, U.K.,
1998.
10.1021/om030095o CCC: $25.00 © 2003 American Chemical Society
Publication on Web 06/10/2003

Catalytic Hydroformylation Properties
Organometallics, Vol. 22, No. 14, 2003 2863
trogen (Praxair), phenylsilane (Aldrich), silver(I) hexafluoro-
phosphate (Aldrich), styrene (Aldrich), thallium hexafluoro-
phosphate (Aldrich), tetrahydrofuran (Anachemia), toluene
(Anachemia), trichloroiridium(III) hydrate (Pressure Chem.
Co.), trichlororhodium(III) hydrate (Pressure Chem. Co.), vinyl
acetate (Aldrich), vinyl benzoate (Aldrich), and vinyl tert-
butylbenzoate (Aldrich). 5,17-Dibromo-25,26,27,28-tetra-n-
propoxycalix[4]arene (3),9 5,17-dibromo-11,23-diphenylphos-
phino-25,26,27,28-tetra-n-propoxycalix[4]arene (5c),9a 5,11-
dibromo-25,26,27,28-tetra-n-propoxycalix[4]arene9a (7), 5,11-
bis(diphenylphosphino)-25,26,27,28-tetra-n-propoxy-calix-
[4]arene9a (9c), and bis[chloro(1,5-cyclooctadienyl)rhodium-
(I)]10 were synthesized according to literature procedures. All
solvents were freshly distilled before use. All air- or moisture-
sensitive materials were handled using standard Schlenk
techniques, high-vacuum manifolds, and inert atmosphere in
a glovebox. Column chromatography was performed on 230-
400 mesh silica gel (60 Å).
apical-equatorial (ea) and diequatorial positions (ee),
respectively, of the triangular bipyramidal intermedi-
ates.
By examining the structural features of square-planar
and octahedral transition metals complexed onto the
upper rim of the 1,3-bis(diphenylphosphino)calix[4]-
arene reported by Matt and co-workers (1 and 2),8
a
rhodium intermediate exhibiting a diapical coordination
appears highly plausible. However steric hindrance may
slow the rates of reactions. To our knowledge, the
diapical case has not been reported and certainly
deserves exploration.
Ap p a r a tu s. All NMR spectra were acquired with a Bruker
AC-300 spectrometer (1H 300.15 MHz, 13C 75.48 MHz, 31P
121.50 MHz) using the solvent as chemical shift standard,
except in 31P NMR, where the chemical shifts are relative to
D3PO4 85% in D2O. All chemical shifts (δ) and coupling
constants (J ) are given in ppm and hertz, respectively. The
IR spectra were acquired on a Perkin-Elmer FT-IR 1600 series
spectrometer. The UV-vis spectra were acquired on a Hewlett-
Packard diode array spectrophotometer 8452A model. The
emission and excitation spectra were obtained using a PTI
LS100 spectrometer. The emission lifetimes were measured
with a nanosecond N2 laser system from PTI model GL-3300
pumping a dye laser model GL-302. The excitation wavelength
was 500 nm using the PLD500 C-500 dye. Electronic ionization
mass spectra were obtained on a VG Instruments ZAB-IF
spectrometer at a potential of 70 eV. Fast atom bombardment
mass spectra were acquired at the Universite´ de Montre´al.
MALDI-TOF mass spectra were acquired from a Bruker
Proflex III linear mode spectrometer with a nitrogen laser (337
nm) from the Universite´ de Bourgogne in Dijon (France) with
a dithranol matrix. Analytical gas chromatography was per-
formed on a Hewlett-Packard 5890 series II spectrometer
We now wish to report the syntheses of four new
diphosphinated calix[4]arene ligands, 1,3- and 1,2-bis-
(diisopropylphosphino)calix[4]arene, and -bis(dimethyl-
phosphino)calix[4]arene, and their corresponding Rh-
(COD)+ catalytic precursors (COD ) 1,5-cyclooctadiene)
as well as their 1,3- and 1,2-bis(diphenylphosphino)-
calix[4]arene analogues, which exist as dimers. Molec-
ular modeling suggests that the diapical coordination
is strongly favored. A comparative study of hydroformyl-
ation reactions of five terminal alkenes indicates dif-
ferences between the two types of isomers.
coupled with
a HP 5971 series selective mass detector.
Elementary analyses were performed at the Universite´ de
Montre´al.
5,17-Bis(d im e t h ylp h osp h in oyl)-25,26,27,28-t e t r a -n -
p r op oxyca lix[4]a r en e (4a ) (Con e). A solution of 100 mg
(0.13 mmol) of 3 in 5 mL of dry THF at -78 °C was treated
with a solution of 0.18 mL (0.30 mmol) of n-BuLi 1.6 M in
hexane. The solution was stirred for 1 h at -78 °C as 0.1 mL
(0.65 mmol) of Me2PCl was added. The solution was stirred
for an additional 20 h and was allowed to reach room
temperature. A 0.1 mL (0.88 mmol) portion of hydrogen
peroxide (30% in water) was then added, and the mixture
stirred for 1 h. The solvent was evaporated and the yellow
residue was purified using column chromatography under
inert atmosphere using a 10:90 ethanol/CHCl3 mixture (Rf )
Exp er im en ta l Section
Ma ter ia ls. The following products have been purchased:
acetone (ACP), argon (Praxair), n-butyllithium (Aldrich), tert-
butyllithium (Aldrich), chlorodiisopropylphosphine (Aldrich),
chlorodimethylphosphine (Aldrich), chloroform (Anachemia),
carbon monoxide (Praxair), COD (Aldrich), dichloromethane
(ACP), ethanol (Les alcools de commerce inc.), ethyl acetate
(ACP), hexanes (ACP), 1-hexene (Aldrich), hydrogen (Praxair),
hydrogen peroxide (Fisher), methylcyclohexane (BDH), dini-
0.25). A white solid was obtained. Yield: 20% (21 mg); Tm
)
280 °C; IR (neat solid) ν(PdO) ) 1206 cm-1; 1H NMR (CDCl3)
7.42 (d, 4H, J ) 11.60, Ph), 6.23 (t, 2H, J ) 7.54, Ph), 6.07 (d,
4H, J ) 7.55, Ph), 4.45 (d, 4H, J ) 13.42, ArCH2), 4.05 (t, 4H,
J ) 6.60, OCH2), 3.68 (t, 4H, J ) 6.81, OCH2), 3.20 (d, 4H, J
) 13.53, ArCH2), 1.90 (m, 8H, CH2CH3), 1.76 (d, 12H, J )
12.88, P(CH3)2), 1.08 (t, 6H, J ) 7.40, CH2CH3), 0.89 (t, 6H, J
) 7.43, CH2CH3); 13C {1H} NMR (CDCl3) 160.92, 155.29, 137.20
(d, J C-P ) 13.9, CAr-P), 132.78, 130.12, 129.99, 127.77, 122.48,
76.99, 76.81, 30.93, 23.20 (d, J C-P ) 21.1, CH3-P), 18.38, 17.42,
10.67, 9.85; 31P{1H} NMR (CDCl3) 38.37 (s); MS (FAB): 745
(5) (a) Casey, C. P.; Whiteker, G. T.; Melville, M. G.; Petrovich, L.
M.; Gavney, J . A.; Powell, D. R. J . Am. Chem. Soc. 1992, 114, 5535.
(b) Casey, C. P.; Petrovich, L. M. J . Am. Chem. Soc. 1995, 117, 6007.
(c) Kranenbrug, 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.
(6) See for instance: Horiuchi, T.; Shirakawa, E.; Nozaki, K.;
Takaya, H. Organometallics 1997, 16, 2981.
(7) (a) Casey, C. P.; Paulsen, E. L.; Beuttenmueller, E. W.; Proft, B.
R.; Petrovich, L. M.; Matter, B. A.; Powell, D. R. J . Am. Chem. Soc.
1997, 119, 11817. (b) Casey, C. P.; Paulsen, E. L.; Beuttenmueller, E.
W.; Proft, B. R.; Matter, B. A.; Powell, D. R. J . Am. Chem. Soc. 1999,
121, 63.
(8) (a) Wieser-J eunesse, C.; Matt, D.; DeCian, A. Angew. Chem., Int.
Ed. 1998, 2861. (b) Lejeune, M.; J eunesse, C.; Matt, D.; Kyritsakas,
N.; Wleter, R.; Kintzinger, J .-P. J . Chem. Soc., Dalton Trans. 2002,
1642.
(9) (a) Gagnon, J .; Ve´zina, M.; Harvey, P. D. Can. J . Chem. 2001,
79, 1439. (b) Larsen, M.; J orgensen, M. J . Org. Chem. 1996, 61, 6651.
(10) Giordano, G.; Crabtree, R. H. Inorg. Synth. 1989, 28, 88.

2864 Organometallics, Vol. 22, No. 14, 2003
Plourde et al.
(M+). Anal. Calcd for C44H58O6P2: C, 70.95; H, 7.79. Found:
C, 69.84; H, 8.09.
COD), 1.85 (br m, 8H, CH2CH3), 1.15-1.4 (br m, 12H, P(CH3)2),
1.11 (t, 6H, J ) 7.37, CH2CH3), 0.86 (t, 6H, J ) 7.51, CH2CH3);
31P{1H} NMR (CD2Cl2) -6.71 (d, Pcalix, J P-Rh ) 144), -141.13
(sept, PF6-, J P-F ) 710); MS (MALDI-TOF) 815 (100%,
(CalixP2)Rh), 923 (30%, (CalixP2)Rh(COD)), 1560 (5%, (CalixP-
(O)P)]2Rh), 1771 (10%, (CalixP2)Rh2PF6). Anal. Calcd for
C52H70O4F6P3Rh: C, 58.43; H, 6.60. Found: C, 58.35; H, 6.59.
5,17-Bis(d iisop r op ylp h osp h in oyl)-25,26,27,28-tetr a -n -
p r op oxyca lix[4]a r en e (4b) (Con e). This compound was
prepared according to the same procedure described for 4a
using i-Pr2PCl as phosphinating agent. A white solid was
obtained. Yield: 55%; Tm ) 222-225 °C; IR (neat solid): ν(Pd
O) ) 1206 cm-1; 1H NMR (CD2Cl2) 7.66 (d, 4H, J ) 12.00, Ph),
6.20 (t, 2H, J ) 7.62, Ph), 5.96 (d, 4H, J ) 7.64, Ph), 4.49 (d,
4H, J ) 13.60, ArCH2), 4.18 (t, 4H, J ) 8.10, OCH2), 3.65 (t,
4H, J ) 6.72, OCH2), 3.31 (d, 4H, 13.69, ArCH2), 3.05 (m, 4H,
PCH(CH3)2), 1.90 (m, 8H, CH2CH3), 1.53 (d, 6H, J ) 7.00, PCH-
(CH3)2), 1.47 (d, 6H, J ) 7.03, PCH(CH3)2), 1.40 (d, 6H, J )
7.12, PCH(CH3)2), 1.34 (d, 6H, J ) 7.12, PCH(CH3)2), 1.08 (t,
6H, J ) 7.20, CH2CH3), 0.89 (t, 6H, J ) 7.40, CH2CH3); 13C-
Bis[µ-bis(5,17-bis(d iisop r op ylp h osp h in o))-25,26,27,28-
tetr a -n -p r op oxyca lix[4]a r en e)(1,5-cycloocta d ien e)r h od -
iu m (I)] Hexa flu or op h osp h a te (6b). This compound was
prepared according to the same procedure as described for 6a
using 5b diphosphine as starting material. An orange solid
was obtained in 90% yield (51.4 mg): Tm > 160 °C (dec); IR
(neat solid) ν(PF6-) ) 838 cm-1; 1H NMR (acetone-d6) 7.39 (d,
4H, J ) 7.87, Ph), 6.09 (m, 6H, Ph), 5.25 (br m, 4H, dCH
COD), 4.48 (d, 4H, J ) 13.36, ArCH2), 4.12 (t, 4H, J ) 8.07,
OCH2), 3.67 (t, 4H, J ) 6.59, OCH2), 3.49 (br m, 4H, CH2 COD),
3.29 (d, 4H, J ) 13.41, ArCH2), 2.79 (sext, 4H, J ) 8.95,
PCH(CH3)2), 2.37 (br m, 4H, CH2 COD), 2.04-1.82 (m, 8H,
CH2CH3), 1.51 (d, 6H, J ) 7.06, PCH(CH3)2), 1.46 (d, 6H, J )
7.08, PCH(CH3)2), 1.31 (d, 6H, J ) 6.93, PCH(CH3)2), 1.26 (d,
6H, J ) 6.91, PCH(CH3)2), 1.13 (t, 6H, J ) 7.37, CH2CH3),
0.91 (t, 6H, J ) 7.42, CH2CH3); 31P{1H} NMR (acetone-d6)
43.98 (d, Pcalix, J P-Rh ) 149), -137.65 (sept, PF6-, J P-F ) 704);
MS (MALDI-TOF) 927 (100%, (CalixP2)Rh), 1069 (15%,
(CalixP2)Rh(COD)), 1175 (5%, (CalixP2)Rh2PF6, 1281 (15%,
(CalixP2)Rh2(COD)PF6). Anal. Calcd for C60H88O4F6P3Rh: C,
60.90; H, 7.50. Found: C, 60.49; H, 7.53.
{1H} NMR (CDCl3) 163.94, 155.08, 139.45 (d, J C-P ) 12.5, CAr
-
P), 134.81, 134.70, 131.77, 127.35, 122.46, 77.28, 77.00, 30.67,
23.29 (d, J C-P ) 16.1, (CH3)2CH-P), 20.97, 20.39, 17.32, 16.44,
10.75, 9.64; 31P{1H} NMR (CD2Cl2) 29.57 (s); MS (FAB) 857
(M+). Anal. Calcd for C52H74O6P2: C, 72.88; H, 8.64. Found:
C, 72.74; H, 8.73.
5,17-B is (d im e t h y lp h o s p h in o )-25,26,27,28-t e t r a -n -
p r op oxyca lix[4]a r en e (5a ) (Con e). A 40 mg (0.056 mmol)
sample of 4a was dissolved in 5 mL of toluene under an inert
atmosphere. The solution was refluxed, and 0.2 mL (1.62
mmol) of phenylsilane was added. The solution was stirred
and refluxed for 12 h. The solvent was then evaporated, and
the crude product was purified using column chromatography
under inert atmosphere using a 5:95 ethyl acetate/hexane
mixture (Rf ) 0.38). A white solid was obtained (41 mg).
Yield: 105%; Tm ) 204-206 °C; 1H NMR (CD2Cl2) 7.06 (d, 4H,
J ) 7.35, Ph), 6.26 (m, 6H, Ph), 4.43 (d, 4H, J ) 13.21, ArCH2),
3.97 (t, 4H, J ) 8.03, OCH2), 3.71 (t, 4H, J ) 7.02, OCH2),
3.15 (d, 4H, J ) 13.28, ArCH2), 1.93 (t of q (∼sext), 4H, J )
7.50, CH2CH3), 1.90 (t of q (∼sext), 4H, J ) 7.50, CH2CH3),
1.24 (d, 12H, J ) 2.65, P(CH3)2), 1.06 (t, 6H, J ) 7.43,
CH2CH3), 0.92 (t, 6H, J ) 7.47, CH2CH3); 13C{1H} NMR (CD2-
Cl2) 159.71, 157.49, 138.25 (d, J C-P ) 6.4, CAr-P), 135.49,
135.08, 132.94, 132.70, 129.43, 123.77, 78.87, 78.53, 32.78,
25.31, 25.00, 16.65 (d, J C-P ) 13.0, CH3-P), 12.38, 11.62, 2.64;
31P NMR (CD2Cl2) -44.50 (s); MS (EI) 712 (M+).
Bis[µ-bis(5,17-bis(d ip h en ylp h osp h in o))-11,23-d ibr om o-
25,26,27,28-tetr a -n -p r op oxyca lix[4]a r en e)(1,5-cycloocta -
d ien e)r h od iu m (I)] Hexa flu or op h osp h a te (6c). This com-
pound was prepared according to the same procedure as
described for 6a using the 5c diphosphine as starting material.
An orange solid was obtained in 95% yield: Tm > 160 °C (dec);
IR (neat solid) ν(PF6-) ) 837 cm-1; 1H NMR (acetone-d6) 7.4-
7.9 (m, 22H, Ph), 7.36 (d, 2H, J ) 12.0, Ph), 6.61 (s, 2H, PhBr),
4.48 (d, 4H, J ) 13.5, ArCH2), 4.0-4.25 (br m, 2H), 4.06 (t,
2H, J ) 7.5, OCH2), 3.83 (t, 2H, J ) 7.0, OCH2), 3.5-3.7 (br
m, 2H), 3.30 (d, 4H, J ) 13.5, ArCH2), 2.4-2.6 (br m, 2H),
1.8-2.2 (m, 8H, CH2CH3), 0.8-1.2 (m, 8H, CH3); 31P{1H} NMR
-
5,17-Bis(d iisop r op ylp h osp h in o)-25,26,27,28-t et r a -n -
p r op oxyca lix[4]a r en e (5b) (Con e). The reduction procedure
was the same as that used for 5a using 4b as starting material.
A white solid was obtained in quantitative yield (100%): Tm
(acetone-d6) 30.04 (d, Pcalix, J P-Rh ) 145), -137.62 (sept, PF6
,
J P-F ) 708); MS (MALDI-TOF) 1221 (100%, (CalixP2)Rh), 1345
(35%, (CalixP2)ORh(COD)), 2440 (1%, (CalixP2)2Rh2)), 2552
(1%, (CalixP2)2Rh2(COD)). Anal. Calcd for C72H76Br2O4F6P3-
Rh: C, 52.25; H, 4.63. Found: C, 52.19; H, 4.54.
1
) 135-136 °C; H NMR (CD2Cl2) 7.21 (d, 4H, J ) 6.68, Ph),
6.14 (t, 2H, J ) 7.49, Ph), 6.04 (d, 4H, J ) 7.35, Ph), 4.44 (d,
4H, J ) 13.28, ArCH2), 4.04 (t, 4H, J ) 8.25, OCH2), 3.65 (t,
4H, J ) 6.66, OCH2), 3.14 (d, 4H, 13.35, ArCH2), 2.12 (m, 4H,
PCH(CH3)2), 1.91 (m, 8H, CH2CH3), 1.11 (dd, 12H, J ) 6.96,
PCH(CH3)2), 1.10 (t, 6H, J ) 7.45, CH2CH3), 0.98 (dd, 12H, J
) 6.84, PCH(CH3)2), 0.89 (t, 6H, J ) 7.45, CH2CH3); 13C{1H}
NMR (CD2Cl2) 159.32, 155.57, 137.00 (d, J C-P ) 8.1, CAr-P),
135.72, 135.46, 127.62, 122.32, 77.41, 76.85, 31.28, 23.96, 23.67
(d, J C-P ) 14.8, (CH3)2CH-P), 20.48, 20.23, 19.24, 19.13, 11.10,
10.00; 31P{1H} NMR (CD2Cl2) 12.95 (s); MS (EI) 825 (M+).
Bis[µ-b is(5,17-b is(d im e t h ylp h osp h in o))-25,26,27,28-
tetr a -n -p r op oxyca lix[4]a r en e)(1,5-cycloocta d ien e)r h od -
iu m (I)] Hexa flu or op h osp h a te (6a ). A solution of TlPF6 (18.7
mg, 53.4 µmol) and [RhCl(COD)]2 (12.0 mg, 24.3 µmol) in 8
mL of dry THF was stirred for 1 h at room temperature. Then,
a solution of compound 5a (34.5 mg, 48.5 µmol) in THF was
slowly added dropwise (for 1 h) to the previous rhodium
solution. The solution was stirred for an additional 16 h. Then,
the solution was filtered, and the solvent was evaporated. An
orange solid was obtained in 90% yield (46.6 mg): Tm > 145
°C (dec); IR (neat solid) ν(PF6-) ) 837 cm-1; 1H NMR (CD2Cl2)
7.22 (d, 4H, J ) 7.39, Ph), 7.01 (t, 2H, J ) 7.36, Ph), 6.03 (br
m, 4H, Ph), 4.96 (br m, 4H, dCH COD), 4.45 (d, 4H, J ) 13.85,
ArCH2), 3.98 (t, 4H, J ) 6.11, OCH2), 3.66 (t, 4H, J ) 6.56,
OCH2), 3.17 (t, 4H, J ) 13.9, ArCH2), 2.37 (br m, 8H, CH2
5,11-Bis(d im e t h ylp h osp h in oyl)-25,26,27,28-t e t r a -n -
p r op oxyca lix[4]a r en e (8a ) (Con e). A solution of 100 mg
(0.13 mmol) of 7 in 5 mL of dry THF at -78 °C was treated
with a solution of 0.18 mL (0.30 mmol) of t-BuLi 1.7 M in
hexane. The solution was stirred for 1 h at -78 °C, and 0.1
mL (0.65 mmol) of Me2PCl was added. The solution was stirred
for an additional 20 h and was allowed to reach room
temperature. Then 0.1 mL (0.88 mmol) of hydrogen peroxide
(30% in water) was added, and the mixture was stirred for 1
h. The solvent was then evaporated, and the yellow residue
was purified using column chromatography under inert at-
mosphere using a 10:90 ethanol/CHCl3 mixture (Rf ) 0.21). A
white solid was obtained. Yield: 25%; Tm ) 110-112 °C; IR
(neat solid) ν(PdO) ) 1206 cm-1 1H NMR (CDCl3) 7.02 (d,
;
2H, J ) 11.77, Ph), 6.95 (d, 2H, J ) 11.76, Ph), 6.67-653 (m,
4H, Ph), 4.50 (d, 1H, J ) 13.48, ArCH2CH2), 4.46 (d, 2H, J )
13.08, CH2), 4.41 (d, 1H, J ) 12.83, CH2), 3.96-3.76 (m, 8H,
OCH2), 3.24 (d, 1H, J ) 10.35, CH2), 3.20 (d, 2H, J ) 13.68,
ArCH2), 3.15 (d, 1H, J ) 14.23, CH2), 1.92 (t of q, 8H, CH2-
CH3), 1.52 (d, 6H, J P-H ) 3.50, P(CH3)2), 1.47 (d, 6H, J P-H
)
3.48, P(CH3)2), 0.99 (m, 12H, CH2CH3); 13C{1H} NMR (CDCl3)
159.32, 156.42, 135.99 (d, J C-P ) 13.1, CAr-P), 135.11, 134.28,
129.98, 129.89, 129.25, 129.14, 128.31, 128.02, 122.48, 77.11,
77.00, 30.91, 23.16, 18.57, 17.47, 10.09; 31P{1H} NMR (CDCl3)

Catalytic Hydroformylation Properties
Organometallics, Vol. 22, No. 14, 2003 2865
35.40 (s); MS (FAB) 745 (M+). Anal. Calcd for C44H58O6P2: C,
70.95; H, 7.79. Found: C, 70.96; H, 7.84.
C52H70O4F6P3Rh: C, 58.43; H, 6.60. Found: C, 58.27; H,
6.70.
5,11-Bis(d iisop r op ylp h osp h in oyl)-25,26,27,28-tetr a -n -
p r op oxyca lix[4]a r en e (8b) (Con e). This compound was
prepared according to the same procedure described for 8a
using i-Pr2PCl as phosphinating agent. A white solid was
obtained. Yield: 45%; Tm ) 106-108 °C; IR (neat solid) ν(Pd
Bis[µ-bis(5,11-bis(d iisop r op ylp h osp h in o))-25,26,27,28-
tetr a -n -p r op oxyca lix[4]a r en e)(1,5-cycloocta d ien e)r h od -
iu m (I)] Hexa flu or op h osp h a te (10b). This compound was
prepared according to the same procedure described for 6a
using the 9b diphosphine as starting material. An orange solid
was obtained in 70% (32.1 mg) yield: Tm >150 °C (dec); IR
(neat solid) ν(PF6-) ) 838 cm-1; 1H NMR (CD2Cl2) 6.93 (d, 4H,
J ) 8.53, Ph), 6.72 (m, 4H, Ph), 6.56 (br t, 2H, Ph), 5.21 (br m,
4H, dCH COD), 4.47 (d, 3H, J ) 12.9, ArCH2), 4.45 (d, 1H, J
) 12.9, ArCH2), 3.91 (m, 8H, OCH2), 3.22 (m, 4H, ArCH2), 2.33
(m, 8H, PCH(CH3)2 + COD), 1.96 (m, 12H, CH2CH3 + COD),
1
O) ) 1206 cm-1; H NMR (CDCl3) 7.08 (d, 2H, J ) 9.59, Ph),
7.01 (d, 2H, J ) 5.42, Ph), 6.70 (br m, 4H, Ph), 6.55 (t, 2H, J
) 7.40, Ph), 4.50 (d, 1H, J ) 12.89, ArCH2), 4.48 (d, 2H, J )
12.83, ArCH2), 4.44 (d, 1H, J ) 12.72, ArCH2), 3.87 (m, 8H,
OCH2), 3.25 (d, 1H, J ) 13.07, ArCH2), 3.25 (d, 2H, J ) 12.98,
ArCH2), 3.16 (d, 2H, J ) 12.82, ArCH2), 2.11 (sept, 4H, J )
7.41, PCH(CH3)2), 1.98 (t of q (∼sext), 8H, J ) 7.50, CH2CH2
0.80-1.4 (br m, 36H, CH2CH3 + PCH(CH3)2); 31P{1H} NMR
-
CH3), 1.04-0.73 (m, 36H, PCH(CH3)2
+
CH2CH2CH3);
(CD2Cl2) 40.04 (d, Pcalix, J P-Rh ) 149), -137.35 (sept, PF6
,
13C{1H} NMR (CDCl3) 158.86, 155.93, 135.17, 134.70,
134.30 (d, J C-P)12.9, CAr-P), 132.30, 131.30, 128.31, 127.92,
122.67, 77.35, 76.50, 30.85, 30.73, 25.15, 24.25, 23.10 (d, J C-P
) 14.7, (CH3)2CH-P), 15.91, 15.73, 14.71, 10.12; 31P{1H}
NMR (CD2Cl2) 52.40 (s); MS (FAB) 857 (M+). Anal. Calcd
for C52H74O6P2: C, 72.88; H, 8.64. Found: C, 72.77; H,
8.71.
J P-F ) 709); MS (MALDI-TOF) 927 (90%, (CalixP2)Rh), 1069
(100%, (CalixP2)Rh(COD)), 1175 (50%, (CalixP2)Rh2PF6, 1281
(70%, (CalixP2)Rh2(COD)PF6). Anal. Calcd for C60H88O4F6P3-
Rh: C, 60.90; H, 7.50. Found: C, 60.89; H, 7.51.
Bis[µ-bis(5,11-bis(diph en ylph osph in o))-25,26,27,28-tetr a-
n -p r op oxyca lix[4]-a r en e)(1,5-cyclooct a d ien e)r h od iu m -
(I)] Hexa flu or op h osp h a te (10c). This compound was pre-
pared according to the same procedure described for 6a using
the 9c diphosphine as starting material. An orange solid was
obtained in 95% yield: Tm > 170 °C (dec); IR (neat solid)
ν(PF6-) ) 837 cm-1; 1H NMR (acetone-d6) 6.0-8.0 (br m, 32H,
Ph), 4.2-4.7 (br m, 4H, ArCH2), 4.2-4.0 (br m, 4H, OCH2),
3.8-4.0 (br m, 4H, OCH2), 2.9-3.5 (br m, 4H, ArCH2), 2.3-
2.5 (br m, 4H), 1.8-2.2 (br m, 8H, CH2CH3), 0.8-1.2 (br m,
5,11-B is (d im e t h y lp h o s p h in o )-25,26,27,28-t e t r a -n -
p r op oxyca lix[4]a r en e (9a ) (Con e). A 40 mg (0.056 mmol)
sample of 8a was dissolved in 5 mL of toluene under inert
atmosphere. The solution was refluxed, and 0.2 mL (1.62
mmol) of phenylsilane was added. The solution was stirred
and refluxed for 24 h. The solvent was then evaporated, and
the crude product was purified using column chromatography
under an inert atmosphere using a 5:95 ethyl acetate/hexane
mixture (Rf ) 0.37). A white, oily product was obtained.
Yield: 103%, 40 mg; Tm ) 133-134 °C; 1H NMR (CD2Cl2) 6.72
(d, 4H, J ) 7.42, Ph), 6.64-6.53 (m, 6H, Ph), 4.44 (d, 1H, J )
13.23, ArCH2), 4.43 (d, 3H, J ) 13.20, ArCH2), 3.85 (t, 4H, J
) 7.15, OCH2), 3.83 (t, 4H, J ) 6.87, OCH2), 3.17 (d, 1H, J )
13.23, ArCH2), 3.14 (d, 3H, J ) 13.25, ArCH2), 1.94 (t of q,
8H, J ) 7.51, CH2CH3), 1.07-0.97 (m, 24H); 13C{1H} NMR
8H, CH3); 31P{1H} NMR (acetone-d6) 33.25 (d, Pcalix, J P-Rh
)
146), -137.62 (sept, PF6-, J P-F ) 708); MS (MALDI-TOF) 1063
(100%, CalixP2Rh), 1171 (15%, CalixP2Rh(COD)), 1311 (40%,
CalixP2Rh2PF6), 1419 (90%, CalixP2Rh2(COD)PF6). Anal. Calcd
for C72H78O4F6P3Rh: C, 57.76; H, 5.25. Found: C, 57.05; H,
5.33.
tr a n s,tr a n s-Bis[µ-bis(5,17-bis(d ip h en ylp h osp h in o)-11,-
2 3 -d i b r o m o -2 5 ,2 6 ,2 7 ,2 8 -t e t r a -n -p r o p o x y c a l i x [4 ]-
a r en e)ch lor oca r bon ylr h od iu m (I)] (12). A solution contain-
ing [RhCl(COD)]2 (47 mg, 0.18 mmol) in dry THF (20 mL) was
bubbled with carbon monoxide for 2 min. A solution of
compound 5c (200 mg, 0.18 mmol) in THF (8 mL) was added
slowly (for 1 h) to the previous solution under agitation. The
solution was stirred for 16 h. The solution was then concen-
trated (0.5 mL) under vacuum and was precipitated with ethyl
ether. The solid obtained was filtered and was washed with
ethyl ether (2 mL). A beige solid was obtained in 95% yield
(CD2Cl2) 157.23, 156.84, 135.31, 135.40 (d, J C-P ) 6.4, CAr
-
P), 131.44, 131.15, 130.72, 130.51, 128.47, 128.40, 122.36,
77.21, 31.15, 23.67, 23.61, 10.48; 31P{1H} NMR (CD2Cl2) -44.51
(s); MS (EI) 712 (M+).
5,11-Bis(d iisop r op ylp h osp h in o)-25,26,27,28-t et r a -n -
p r op oxyca lix[4]a r en e (9b) (Con e). The reduction procedure
was the same as that used for 9a using 8b as starting material.
A white solid was obtained in 99% yield: 1H NMR (CD2Cl2)
6.84 (t, 2H, J ) 5.22, Ph), 6.69 (d, 4H, J ) 7.44, Ph), 6.56 (d,
4H, J ) 7.39, Ph), 4.46 (d, 4H, J ) 12.85, ArCH2), 3.88 (dt,
8H, J ) 8.03, OCH2), 3.18 (d, 3H, J ) 12.9, ArCH2), 3.16 ((d,
1H, J ) 12.9, ArCH2), 2.00 (m, 8H, CH2CH3), 1.97 (m, 4H,
PCH(CH3)2), 1.01 (dt, 6H, J ) 7.37, CH2CH3), 0.89 (dd, 12H,
J ) 7.06, PCH(CH3)2), 0.69 (dt, 6H, J ) 7.35, CH2CH3); 13C-
(0.23 g): Tm > 160 °C (dec); IR (neat solid) ν(CO) 1979 cm-1
;
1H NMR (benzene-d6) 8.18 (d, 2H, J ) 8.5, Ph), 8.15 (d, 2H, J
) 7.5, Ph), 7.93 (d, 2H, J ) 10.0, Ph), 7.32-7.38 (m, 12H, Ph),
7.12-7.20 (m, 8H, Ph), 6.61 (s, 4H, PhBr), 4.14 (d, 4H, J )
13.5, ArCH2), 3.86 (t, 4H, J ) 8.5, OCH2), 3.30 (t, 4H, J ) 7.0,
OCH2), 2.78 (d, 4H, J ) 13.5, ArCH2), 1.77-1.85 (m, 4H, CH2-
CH3), 1.54-1.61 (m, 4H, CH2CH3), 0.89 (t, 6H, J ) 7.5, CH3),
0.74 (t, 6H, J ) 7.5, CH3); 31P{1H} NMR (benzene-d6) 29.46
(d, Pcalix, J P-Rh ) 128); MS MALDI-TOF 1221 (100%, (CalixP2)-
Rh), 1255 (30%, (CalixP2)RhCl), 2339 (5%, (CalixP2)2Rh), 2408
(1%, (CalixP2)2Rh(CO)Cl), 2470 (5%, (CalixP2)2Rh2(CO)), 2534
(1.3% (CalixP2)2Rh2(CO)2-Cl), 2570 (<1%, (CalixP2)2Rh2(CO)2-
Cl2). Anal. Calcd for C65H64O5P2ClBr2Rh: C, 60.74; H, 5.02.
Found: C, 60.59; H, 5.05.
{1H} NMR (CD2Cl2) 157.32, 156.59, 135.22 (d, J C-P ) 8.4, CAr
-
P), 134.87, 134.63, 131.31, 128.57, 122.85, 77.48, 77.35, 31.23,
31.11, 23.68, 22.81 (d, J C-P ) 8.9, (CH3)2CH-P), 20.09, 19.84,
19.05, 18.82, 10.49; 31P{1H} NMR (CD2Cl2) 12.67 (s); MS (EI)
825 (M+).
Bis[µ-b is(5,11-b is(d im e t h ylp h osp h in o))-25,26,27,28-
tetr a -n -p r op oxyca lix[4]-a r en e)(1,5-cycloocta d ien e)r h od -
iu m (I)] Hexa flu or op h osp h a te (10a ). This compound was
prepared according to the same procedure described for 6a
using the 9a diphosphine as starting material. An orange solid
was obtained in 90% yield: Tm > 123 °C (dec); IR (neat solid)
ν(PF6-) ) 836 cm-1; 1H NMR (CD2Cl2) 6.20-7.40 (br m, 10H,
Ph), 4.7-5.0 (br m, 4H, dCH COD), 4.3-4.6 (br m, 4H, ArCH2),
3.6-4.1 (br m, 8H, OCH2), 3.00-3.40 (br m, 4H, ArCH2), 2.20-
2.50 (br m, 4H), 1.70-2.10 (br m, 12H), 0.70-1.40 (br m, 24H);
31P{1H} NMR (CD2Cl2) -1.33 (d, Pcalix, J P-Rh ) 144), -141.02
(septuplet, PF6-, J P-F ) 711); MS (MALDI-TOF) 815 (10%,
(CalixP2)Rh), 831 (70%, (CalixP(O)P)Rh) 939 (15%, (CalixP-
(O)P)Rh(COD)) 1560 (50%, (CalixP(O)P)2Rh). Anal. Calcd for
tr a n s,tr a n s-Bis[µ-b is(5,11-b is(d ip h en ylp h osp h in o))-
25,26,27,28-tetr a -n -p r op oxyca lix[4]a r en e)]ch lor o-ca r bo-
n ylr h od iu m (I)] (13). This compound was prepared according
to the same procedure described for 12 using the 9c diphos-
phine as starting material. A beige solid was obtained in 95%
yield: Tm > 170 °C (dec); IR (neat solid) ν(CO) 1980 cm-1; 1H
NMR (benzene-d6) 8.3-8.7 (br m, 4H, Ph), 6.0-8.0 (br m, 26
H, Ph), 4.59 (d, 4H, J ) 13.0, ArCH2), 4.29 (dd, 4H, J ) 13.0,
ArCH2), 3.26 (d, 4H, J ) 13.0, ArCH2), 3.4 (br m, 8H, OCH2),
1.5-2.0 (br m, 8H, CH2CH3), 0.6-1.0 (br m, 12H, CH2CH3);

2866 Organometallics, Vol. 22, No. 14, 2003
Plourde et al.
31P{1H} NMR (benzene-d6) 30.62 (d, Pcalix, J P-Rh ) 124), 29.11
(d, Pcalix, J P-Rh ) 129); MS MALDI-TOF 2156 (100%, (CalixP2)2-
Rh2(CO)), 2194 (95%, (CalixP2)2Rh2Cl2), 2289 (4%, (CalixP2)2-
Rh2(CO)Cl2. Anal. Calcd for C65H66O5P2ClRh: C, 69.24; H, 5.90.
Found: C, 69.04; H, 6.11.
0.003 mol of methylcyclohexane (as internal standard) in THF
was placed inside the reactor and allowed to reach the desired
temperature in 20 min. The reactor was purged three times
with CO gas. Then 35 atm of CO and 35 atm of H2 was
introduced into the reactor, while the stirring was activated.
All the runs were 72 h long. Samples were removed regularly
for GC/MS analysis through the sampling valve. The aldehyde
products were analyzed by temperature-programmed gas
chromatography (10 °C/min) from 60 to 250 °C with a DB-
5MS column. A 0.9 mL/min helium flow was used. Tof is
defined as the number of moles of products divided by the total
number of moles of rhodium used per hour. In this work, [Rh]
) 1 × 10-4 M. This latter value was used to calculate tof.
Com p u ter Mod elin g. The calculations were performed
using the commercially available PC-model from Serena
Software (version 7.0), which uses the MMX empirical model.
No constraints on bond distances or angles were applied to
ensure that deviations from normal geometry are observed.
Comparisons between literature X-ray structures and com-
puted models are provided in the text.
Cr ysta llogr a p h y. Single crystals of 5b were coated with
Paratone-N oil, mounted using a glass fiber, and frozen in the
cold nitrogen stream of the goniometer. A hemisphere of data
was collected on a Bruker AXS P4/SMART 1000 diffractometer
using ω and θ scans with a scan width of 0.3° and 30 s
exposure times. The detector distance was 6 cm. The data were
reduced (SAINT)11 and corrected for absorption (SADABS).12
The structure was solved by direct methods and refined by
full-matrix least-squares on F2 (SHELXTL).13 Three of the
propoxy groups were disordered, and the site occupancy was
determined using an isotropic model as 0.75 (C(31)), 0.25
(C(31A)), 0.65 (C(32)-C(33), C(38)-C(39)) and 0.35 (C(32A)-
C(33A), C(38A)-C(39A)) and fixed in subsequent refinement
cycles. All non-hydrogen atoms were refined anisotropically.
Hydrogen atoms were included in calculated positions and
refined using a rigid model. Hydrogen atoms at C(31A),
C(32A)-C(33A), and C(38A)-C(39A) were omitted. Thermal
ellipsoid plots are at the 30% probability level. In some plots,
hydrogen atoms have been omitted for clarity.
Bis[µ-bis(5,17-bis(d ip h en ylp h osp h in o))-11,23-d ibr om o-
25,26,27,28-tetr a -n -p r op oxyca lix[4]a r en e)(1,5-cycloocta -
d ien e)ir id iu m (I)] h exa flu or op h osp h a te (14). To a solution
of [IrCl(COD)]2 (9 mg, 0.0134 mmol) in CH2Cl2 (2 mL) was
added dropwise a solution of TlPF6 (7.5 mg, 0.0295 mmol) in
dry THF (1 mL). This orange solution was filtered over Celite
and was added directly to a solution of 5c (30 mg, 0.0268 mmol)
in CH2Cl2 (8 mL). The solution was stirred for 1 h. After 1 h,
the red solution was concentrated to ca. 3 mL and was
precipitated with ethyl ether. The solution was filtered, and a
pink solid was obtained in 65% yield (12.3 mg): Tm > 190 °C
(dec); IR (neat solid) ν(PF6-) ) 839 cm-1 1H NMR (CD2Cl2)
;
7.70-7.15 (m, 24H, Ph), 6.28 (s, 4H, PhBr), 4.58-4.49 (m, 4H,
dCH COD), 4.39 (d, 4H, J ) 14.3, ArCH2), 4.08 (t, 4H, CH2
COD), 3.59 (t, 8H, J ) 6.5, OCH2CH2CH3), 3.10 (d, 4H, J )
14.3, ArCH2), 2.56-2.77 (m, 2H, CH2 COD), 2.03-1.68 (m,
10H, OCH2CH2CH3 + CH2 COD), 1.04 (t, 6H, J ) 6.5, OCH2-
CH2CH3), 0.82 (t, 6H, J ) 6.5, OCH2CH2CH3); 31P{1H}
NMR (CD2Cl2) 21.1 (s, PPh2), -141.20 (sept, PF6-, J P-F
)
708). MS (FAB) 1311.2 (100%, (CalixP2)Ir), 2430 (30%,
(CalixP2)2Ir), 2538.2 (20%, (calixP2)2Ir(COD)). Anal. Calcd
for C72H76Br2O4F6P3Ir: C, 49.58; H, 4.39. Found: C, 49.70; H,
4.28.
Bis[µ-bis(5,11-bis(diph en ylph osph in o))-25,26,27,28-tetr a-
n -p r op oxyca lix[4]-a r en e)(1,5-cyclooct a d ien e)ir id iu m -
(I)] Hexa flu or op h osp h a te (15). This compound was pre-
pared according to the same procedure described for 14 using
the 9c diphosphine as starting material. The solution was
filtered, and a pink solid was obtained in 50% yield (12.3 mg):
Tm > 150 °C (dec); IR (neat solid) ν(PF6-) ) 839 cm-1; 1H NMR
(CD2Cl2) 7.61-5.79 (m, 30H, Ph), 4.66-4.31 (m, 8H, CH2+ d
CH COD), 4.35-3.61 (m, 12H, OCH2CH2CH3 +COD), 3.42-
2.82 (m, 4H, CH2+ dCH COD), 2.12-1.59 (m, 12H, OCH2CH2-
CH3 +COD), 1.19-0.72 (m, 12H, OCH2CH2CH3). 31P{1H} NMR
(CD2Cl2) 21.1 (br s, PPh2), -141 (sept, PF6-, J P-F ) 709); MS
(FAB) 1153.3 (100%, (CalixP2)Ir), 1453.7 (70%, (CalixP2)Ir2-
(COD)), 1598.7 (35%, (CalixP2)Ir2(COD)PF6), 2114.4 (67%,
(CalixP2)2Ir), 2222.9 (75%, (CalixP2)2Ir(COD)). Anal. Calcd
for C72H78O4F6P3Ir: C, 54.51; H, 4.95. Found: C, 54.41; H,
5.04.
Resu lts a n d Discu ssion
Syn th eses a n d Ch a r a cter iza tion . The 1,2- and 1,3-
diphosphinated ligands can be prepared in two steps
from the corresponding dibromides 3 and 7, or the
tetrabromide 3b (Schemes 1 and 2), using three differ-
ent dichlorophosphine-starting materials (R2PCl; R )
Me, i-Pr, Ph). For the highly oxygen-sensitive phos-
phines (R ) Me, i-Pr), hydrogen peroxide is added to
syntheses, allowing the fully oxidized species to be
formed (4a , 4b, 8a , 8b). These latter compounds lead
to better separation and purification, and this extra step
provides better yields. The subsequent reduction pro-
ceeds cleanly with phenylsilane.
tr a n s,tr a n s-Bis[µ-bis(5,17-bis(d iisop r op ylp h osp h in o))-
25,26,27,28-tetr a -n -p r op oxyca lix[4]a r en e)h yd r id oca r bo-
n ylr h od iu m (I)] (17). A solution containing 20 mg (8.5 µmol)
of 6b in acetone (15 mL) was inserted in a high-pressure
reactor heated at 50 °C. The reactor was purged three times
with CO, then 35 atm of CO was applied, followed by 35 atm
of H2. The solution was stirred for 24 h, then allowed to cool
to room temperature. The precipitate was filtered and washed
with acetone. A white solid was obtained in 90% yield (15.0
1
mg): Tm > 155 °C (dec); IR (neat solid) ν(CO) 1970 cm-1; H
During the course of this study, single crystals suit-
able for X-ray analysis were obtained for 5b (Figure 1
and Supporting Information).14 The data confirm the
cone conformation and indicate a flattened structure in
which the phosphine groups are placed away from each
NMR (CD2Cl2) 7.41 (m, 4H, Ph), 6.02 (m, 4H, Ph), 5.80 (t, 2H,
J ) 7.62, Ph), 4.43 (d, 4H, J ) 13.08, ArCH2), 4.09 (t, 4H, J )
8.07, OCH2), 3.62 (t, 4H, J ) 6.69, OCH2), 3.16 (d, 4H, J )
13.26, ArCH2), 2.94 (m, 4H, PCH(CH3)2), 2.05 (t of q (∼sext),
4H, J ) 7.69, OCH2CH2CH3), 1.87 (t of q (∼sext), 4H, J ) 7.17,
OCH2CH2CH3), 1.34 (m, 12H, PCH(CH3)2), 1.20 (m, 12H, PCH-
(CH3)2), 1.10 (t, 6H, J ) 7.35, OCH2CH2CH3), 0.93 (t, 6H, J )
(11) SAINT 6.02; Bruker AXS, Inc.: Madison, WI, 1997-1999.
(12) Sheldrick, G. SADABS; Bruker AXS, Inc.: Madison, WI, 1999.
(13) Sheldrick, G. SHELXTL 5.1; Bruker AXS, Inc.: Madison, WI,
1997.
(14) (a) X-ray data were obtained for 4a but were not of sufficient
quality for full analysis. The conclusions made for the structure of 5b
are the same for this compound. The data can be found in the
Supporting Information. (b) While this paper was being submitted,
another related structure, 5,17-bis(diphenylphosphinoyl)-25,26,27,28-
tetrapropoxycalix[4]arene, has been reported. The structure resembles
that of our own structures. See: J eunesse, C.; Matt, D.; J ones, P. G.;
Thoennessen, H. Acta Crystallogr. 2003, E59, 428.
7.35, OCH2CH2CH3); 31P{1H} NMR (CD2Cl2) 45.55 (d, J P-Rh
)
122).
Hyd r ofor m yla tion Exp er im en ts. Hydroformylation reac-
tions were performed in a 200 mL Parr Instrument autoclave
(4001 model) equipped with a gas inlet-outlet valve and a
liquid sampling valve. Mechanical stirring provided agitation.
The autoclave was operating at 50 ( 1 °C. The reactor system
was preheated to 50 °C before the experiment. A solution of
rhodium dimer (5 µmol), 0.01 mol of substrates (alkenes), and

Catalytic Hydroformylation Properties
Organometallics, Vol. 22, No. 14, 2003 2867
Sch em e 1
F igu r e 1. ORTEP drawing of the ligand 5b. The ellipsoids
are shown at 30% probability, and the H atoms are not
shown for clarity.
Ta ble 1. Com p a r ison of Selected Non bon d ed
Dista n ces for th e X-r a y a n d Mod eled Str u ctu r e
of 5b
distance
X-ray/Å
computed/Å
inward Cpara‚‚‚Cpara
outward Cpara‚‚‚Cpara
inward O‚‚‚O
outward O‚‚‚O
P‚‚‚P
4.080
10.253
5.943
3.380
13.360
4.725
10.277
5.974
3.789
13.360
distances are 4.342, 4.080, and 4.350 Å for the face-to-
face Cmeta‚‚‚Cmeta, Cpara‚‚‚Cpara, and Cmeta‚‚‚Cmeta separa-
tions, respectively, leading to an inward tilt of ∼15°. The
two face-to-face O‚‚‚O distances are 3.380 and 5.943 Å
for outward and inward tilts, respectively. The wide
Cpara‚‚‚Cpara and P‚‚‚P separations are 10.253 and 13.360
Å, respectively. The solution 1H, 13C, and 31P NMR
spectra are consistent with a 2-fold symmetry. The
comparison of this experimental structure with a mod-
eled one shows a good agreement (Table 1).
Sch em e 2
The coordination with the Rh(COD)+ fragment onto
the diphosphine ligands (6a ,b,c and 10a ,b,c) proceeds
in good yields (70-95%) with the use of the dimer bis-
[chloro(1,5-cyclooctadienyl)rhodium(I)] reactant. The
complexes did not provide single crystals suitable for
X-ray crystallographic analysis. The MALDI-TOF data
exhibit peaks that are attributable to fragments larger
than a monomer (i.e., two or more calix[4]arene ligands).
The analysis shows no evidence of chloride ion or
rhodium chloride fragment in the sample. This result
contrasts with the proposed related complex of the
species 5,17-bis(methylene(diphenylphosphino))-25,26,-
27,28-tetrapropoxycalix[4]arene,15 in which two coordi-
nated RhCl(COD) fragments are proposed to be dan-
gling from the calix[4]arene upper rim. In this latter
case, the preparation method differs from this one by
not using TlPF6 as a chloride-removing agent. The
authors proposed a structure of the type (PrO)4calix[4]-
arene[CH2(Ph2P)RhCl(COD)]2 (11a ). In addition, the
reaction of this same ligand with (COD)PdMeCl and
(COD)PtCl2 leads to polymers (or oligomers) of the type
(-calixP2-PdMeCl-)n (11b) and (-calixP2-PtCl2-)n
1
(11c), as shown below, based on H NMR spectra which
exhibit broad signals.15 In comparison with the 1,3-
1
complexes reported here, the H and 31P NMR spectra
other, avoiding intramolecular steric interactions. The
R2P groups oriented such that the molecule has C2-
symmetry. The cavity is somewhat restricted in this
solid state structure, where the nonbonding C‚‚‚C
(15) Fang, X.; Scott, B. L.; Watkin, J . G.; Carter, C. A. G.; Kubas,
G. J . Inorg. Chim. Acta 2001, 317, 276.

2868 Organometallics, Vol. 22, No. 14, 2003
Plourde et al.
do not exhibit large signals at room temperature.
Debye equation assuming a spherical shape if the
correlation time, τc, is known (τc ) Vηvisc/kT; V ) volume;
ηvisc ) solvent viscosity; k ) Boltzmann constant; T )
temperature). To measure τc, one has to know the spin-
lattice relaxation time related to dipole-dipole interac-
tions, T1DD, which is given by the sum over all dipole-
dipole interacting nuclei (31P and 1H in this work)
according to eq 1:
(p2γ γ 2/4π4rPH6)τc
(1)
DD
2
1/T1
)
∑
P
H
in the extreme narrowing limit,17 where p is the Planck’s
constant, γ is the magnetogyric ratio for the interacting
nuclei, and r is the distance between the probe nuclei
(here 31P) and the nearest interacting nuclei (here 1H).18
DD
1
The T1 data (here at 300 MHz H) can be extracted
from experimental T1 and NOE measurements (nuclear
Overhauser effect),19 according to eq 2:
DD
1/T1 ) η/(ηmaxT1)
(2)
with η ) the fractional NOE constant, ηmax ) the
maximum η value in the extreme narrowing limit (here
ηmax ) γH/2γP ) 1.235), and T1 is the experimental spin-
lattice relaxation time.20 The strategy is to compare the
hydrodynamic volume of a reliably known compound
(crystallographically characterized monomer, dimer, or
other) to that of the unknown ones. This standard
molecule must be closely related to the sample ones. In
this work, the probe nuclei is 31P because it provides a
better sensitivity than 13C (natural abundance is 1%)
and 103Rh, for instance, and is free from interference
In relation to the elucidation of the exact nature of
the newly synthesized Rh(COD)+ complexes, two RhCl-
(CO) complexes were also prepared (Scheme 2). Their
syntheses proceed similarly to that of 6c and 10c with
the difference that no Cl- abstractor (such as Tl+ or Ag+)
is added to the reaction under CO atmosphere. The
complexes exhibit mass spectra (MALDI-TOF) where
the molecular ion is observed for 12 and [M - (CO)]+
for 13, demonstrating their dimer nature in solution.
The 31P NMR spectra for 13 exhibit two very closely
located signals with very similar J constants (J PRh).
These are possibly due to the two relative orientations
of the Cl-Rh-CO axes (parallel and antiparallel) of the
face-to-face Rh-coordination planes (the proposed struc-
ture is presented below). The complexes 12 and 13 are
also neutral, and for T1 measurements described below,
this property is important. Their spectroscopic charac-
terizations provide comparative data for the correspond-
ing hydrides presented below.
1
(as in the H spectra) since the signal is unique in the
spectra. By combining eqs 1 and 2 to the Stokes-
Einstein-Debye equation, one obtains
T1(sam)
T1(sta)
η
PH(sam)
1/r 6(sta)
V(sta)
∑
∑
PH
)
(3)
1/r 6(sam)
η
PH(sta) V(sam)
PH
where sam and sta stand for sample and standard,
respectively. This equation requires the knowledge of
the interatomic distances between the 31P and 1H nuclei,
and ηPH. The X-ray structure presented for compound
5b above provides such distances (Scheme 3).21 By using
the corresponding ligand as a standard, the term [∑1/
Similarly to complexes 6c and 10c, the corresponding
iridium(I) compounds (14 and 15) have been prepared
in modest yields. The MALDI-TOF mass spectra also
exhibit fragment peaks indicating the presence of
dimers as well.
rPH6(sta)]/[∑1/r6 6(sam)] becomes unity. Because the
PH
chemical environment of the probe nucleus is the same
(17) T1 decreases with the field (H0), indicating the extreme nar-
rowing limit.
Sp in -La ttice Rela xa tion Tim e Mea su r em en ts.
Due to the limited mass data for the Rh(COD)+ species,
the exact nature of the complexes in solution cannot be
confidently addressed by this method. Moreover, due to
the relatively weak solubility, the determination of the
molecular weight using techniques such as the mea-
surements of the freezing point depression cannot be
applied. Recently, we demonstrated how spin-lattice
relaxation time measurements can overcome these
problems and provide information on the size of the
tumbling molecules in solution.16 This previous work
reported in detail the methodology, and only the key
features are presented here. The size (or volume) of a
molecule can be obtained from the Stokes-Einstein-
(18) The interactions between the 103Rh and 31P, despite a shorter
distance (2.27 Å according to literature X-ray data on related systems),
account only for 1% or less of the total contribution on T1DD because of
the small value for γRh (9.442 MHz at 7.0463T; i.e., γP ) 121.442 and
γH ) 300.000). See the cover table in Drago R. S. Physical Methods for
Chemists, 2nd ed.; Saunders College Publishers: New York, 1992.
These interactions were neglected in this work.
(19) Wehrli, F. W.; Wirthlin, T. Interpretation of Carbon-13 NMR
Spectra; Heyden: London, 1980.
(20) The contribution from chemical shift anisotropy, quadrupolar,
electron paramagnetic, chemical exchange, and spin-rotation pro-
cesses are negligible. Indeed, the complexes do not exhibit quadrupolar
nuclei nor are they paramagnetic. Although COD is considered labile,
no evidence for efficient chemical exchange was seen in these complexes
in the absence of other free ligands. The full-width at half-maximum
(fwhm) was not strongly temperature dependent between 20 and 40
°C. Because of the sixth power of r in eq 1, any longer P‚‚‚H distances
DD
generate terms that contribute less than 1% of T1
.
(21) The X-ray data for 4a , as well as the calculated structures, give
similar values.
(16) Turcotte, M.; Harvey, P. D. Inorg. Chem. 2002, 41, 2971.

Catalytic Hydroformylation Properties
Organometallics, Vol. 22, No. 14, 2003 2869
Ta ble 2. T1 Da ta for th e Sta n d a r d s 4a , 5b, a n d 5c a n d Com p lexes 6a , 6b, 6c, 10a , 10b, 10c, 12
b
compound
T1/(0.2 s
T1(sam)/T1(sta)
Va/Å3
V(sta)/V(sam)
free 1,3-(Ph2P)2Br2 (5c)
4.2
2.4
2.2
1.8
3.8
2.0
1.6
2.6
1.5
1.7
1279
2904
2746
2665
1121
2394
2411
978
[1,3-(Ph2P)2Br2Rh(COD)+]2 (6c)
[1,2-(Ph2P)2Rh(COD)+]2 (10c)
[1,3-(Ph2P)2Br2PRh(CO)Cl]2 (12)
free 1,3-(i-Pr2P)2 (5b)
0.57
0.53
0.43
0.44
0.46
0.48
[1,3-(i-Pr2P)2Rh(COD)+]2 (6b)
[1,2-(i-Pr2P)2Rh(COD)+]2 (10b)
free 1,3-(Me2PdO)2 (4a )c
0.53
0.42
0.47
0.46
[1,3-(Me2P)2Rh(COD)+]2 (6a )
[1,2-(Me2P)2Rh(COD)+]2 (10a )
0.57
0.65
2073
2140
0.47
0.46
a
b
Volume computed from modeling data using solvent-free models. The ratio V(sta)/V(sam) computed from modeling data. c 5a proved
too air-sensitive despite the precautions and oxidized too easily, even during measurements. Instead, 4a was used as standard.
Ta ble 3. Com p a r ison of th e X-r a y Da ta for 16 w ith
Com p u ted Str u ctu r ea
Sch em e 3
distance
X-ray data/Å
computation/Å
Pd-P
2.319(5), 2.313(5)
2.323(4), 2.289(4)
13.187(2)
13.326(6)
17.818(7)
2.37
2.19
13.35
13.39
17.63
9.17
Pd-Cl
Pd‚‚‚Pd′
P‚‚‚P′
Cl‚‚‚Cl′outer
Cl‚‚‚Cl′inner
8.667(8)
a
See ref 24.
average T1 values for the 1,2- and 1,3-isomers are 2.3
> 1.8 > 1.6 s for the Ph2P-, i-Pr2P-, and Me2P-
derivatives, respectively. The T1(sam)/T1(std) ratios indicate
the presence of dimers in solution, for all the species.
in the free and complexed ligands, ηPH are approxi-
mately the same, and the ratio T1(sam)/T1(sta) is
consequently approximately equal to that of V(sta)/
V(sam). In addition, the fact that the standard is neutral
and the investigated complexes are monocations per
unit of ligand, the hydrodynamic volume is anticipated
to be greater for the charged species. For this reason,
the related neutral complex 1,3-(PPh2)calix[4]arene-
[RhCl(CO)] (12) is also investigated to see the impor-
tance of this effect. This derivative, as well as the
corresponding 1,2-isomer 13, exhibits fragment peaks
close to the dimer, or exactly on the dimer mass as
stated above. The T1 data are summarized in Table 2.
The ratio T1(sam)/T1(sta) is 0.43 for the neutral com-
pound 12, which compares favorably to the computed
ratio V(sta)/V(sam) for a dimer species (i.e., 0.48).22
Other stoichiometries result in unacceptable ratios (1:
1, 0.93; 1:3, 0.31; 1:4, 0.23). Clearly, only a dimer species
can explain the T1 data. For the charged species 6c and
10c, this ratio increases, indicating an increase in
V(sam). This result is consistent with the increase in
hydrodynamic volume due to electrostatic interactions
with the solvent.23
Com p u ter Mod elin g. The purpose of molecular
modeling in this work is 2-fold. First, one may try to
understand why dimerization occurs, and second, in the
absence of X-ray data, what may be the steric hindrance
about the metal in these species.
Prior to addressing these questions, the computa-
tional methodology has been tested on a ligand (see data
for compound 5b in Table 1) and on a crystallographi-
cally characterized related dimer compound, [1,3-(Ph2P)2-
calixPdCl2]2, recently published.24 This dimer exhibits
a large macrocycle composed of two face-to-face calix-
[4]arenes in a flattened structure. The comparison
between the experimental and computed data (Table 3)
is acceptable, where a discrepancy of approximately 0.1
and 0.2-0.5 Å for the M-L bond lengths and nonbond-
ing separations, respectively, is noticed.
Computed monomer structures have been analyzed
as well for comparison purposes. These monomer models
exhibit computed P-Rh-P angles that are greater than
the corresponding computed dimers discussed below.
For instance, the 1,2-isomer computes a PRhP angle of
109° (instead of crystallographically found angles of 84°
for other related species).25 In addition, some of the
aromatic rings show significant deviations from planar-
ity, indicating clearly the presence of important ring
stress in the monomer models. However, the ring stress
in the 1,3-monomers is just only slightly greater than
the corresponding dimers, where a difference of about
1-2° only in P-Rh-P angles are computed. This very
small difference may explain why Matt and collabora-
tors8 have observed a monomer for the trans-PtCl2
derivatives (1) and Takenaka et al.24 have reported a
Comparison of the T1 data between the free ligands
5c, 5b, and 4a shows a decrease in this same order,
which is consistent with an increase in the number of
1
close H nuclei for 1H-31P dipole-dipole interactions.
This trend is also followed in the complexes as well. The
(22) The molecular volumes for 5b and 5c are 1121 and 1279 Å3,
respectively, according to the PC-Model. The X-ray data indicate a unit
cell volume of 2550 Å3 for 5b, giving an approximated value of 1275
Å3 for a single molecule, and 5753 Å3 for Z ) 4,9 giving an approximated
value of 1438 Å3. This comparison is good taking into account the
packing effect, which does not preclude an unavoidable void in the
lattice. No solvent molecule was found as crystallizing molecule. By
using the volume ratio, any systematic error caused by the method is
minimized.
(24) Takenaka, K.; Obora, Y.; J iang, L. H.; Tsuji, Y. Organometallics
2002, 21, 1158.
(25) Ball, R. G.; Payne, N. C. Inorg. Chem. 1977, 16, 1187.
(23) No charged diphosphinated calix[4]arene complex species that
are crystallographically characterized are available.

2870 Organometallics, Vol. 22, No. 14, 2003
Plourde et al.
F igu r e 2. Model structure for 6b.
Sch em e 4
F igu r e 3. One possible model structure for 10c, where
the Rh(COD)+ are away from each other.
dimer for the trans-PdCl2 species (16).
The 1,3-dimers can form only face-to-face species
where the calix[4]arene macrocycles slightly sit in each
other’s cavities, forming an S-shape structure, with the
C-H bonds penetrating the top of the cavity (Figure
2). Some degree of freedom is seen in the Rh(COD)+
fragments where the P2Rh(CdC)2 planes may tilt up or
down with respect to the bis(calix[4]arene) frame. Such
motions involve the cooperativity of the R2P groups. No
Rh‚‚‚Rh cooperativity can be anticipated as the Rh-Rh
separation is on the order of 15 Å in this model. The
calix[4]arene cavity is not available for interactions with
substrates.
The analysis of the 1,2-dimers is more complicated.
The ring flipping in the nonrigid calix[4]arene leads to
two enantiomeric conformers, leading to two nonsuper-
imposable molecules due to the flattened geometry of
the calix[4]arene (Scheme 4), which are called a and b.
These species coexist in 50% population in solution, and
the combinations of a + a (enantiomer), b + b (enan-
tiomer), and a + b dimers during the coordination
reactions are possible. However, the a + a and b + b
forms are unlikely because the two macrocycles are
placed side-by-side where both upper rims are directed
in the same direction. In this geometry, the steric
hindrance between the substituents is significant.
F igu r e 4. One possible model structure for 10c, where
the Rh(COD)+ are located above the calix[4]arene cavity.
In the a + b form, the two calix[4]arene diphosphine
ligands can be “turned around” with respect to each
other where the tetrapropyl groups of one calix[4]arene
are pointing in the opposite direction with respect to
those of the second macrocycle. Then, three symmetric
structures are still possible for this assemblage. First,
the Rh(COD)+ fragments can point away from each
other (Figure 3) or can be trapped over the cavity
(Figure 4). This latter structure is not without prece-
dent, as Matt and collaborators have recently reported
a structure of a bis(5,17-bis(diphenyl)diphosphino)calix-
[4]arene ruthenium(aryl) species where the aromatic
group is located inside the cavity.26 The third one places
the Rh(COD)+ groups toward each other and is physi-
cally impossible. In the two plausible models, the two
main features are (1) the Rh‚‚‚Rh separation is ∼8 Å
for both cases and (2) the cavity is greatly obstructed
either by the R2P groups (R ) Ph, i-Pr, Me) or by Rh-
(COD)+ itself.
(26) Lejeune, M.; J eunesse, C.; Matt, D.; Kyritsakas, N.; Welter, R.;
Kintzinger, J . P. J . Chem. Soc., Dalton Trans. 2002 1642.

Catalytic Hydroformylation Properties
Organometallics, Vol. 22, No. 14, 2003 2871
F igu r e 6. UV-vis (Ab), excitation (Ex), and emission (Em)
spectra of 14 in ethanol at 77 K.
F igu r e 5. Model structure for 13.
The side-by-side structures have been analyzed as
well, and ring stress has readily been detected where
calculated P-Rh-P angles > 96° are seen (instead of a
theoretical value of 90°) and long Rh-P bond distances
of 2.56 Å (instead of ∼2.28 Å) are detected. In addition,
important steric problems are noticed between the
substituents of the four -PR2 groups. Clearly, the side-
by-side is not a favored geometry.
ing chromophore (see compounds 18 and 19).
Compounds 12 and 13 have been modeled as well.
In the 1,2-series, multiple conformers exist also, but
fewer intramolecular steric interactions are noticed due
to the trans-positions of the phosphine ligands. One of
them (Figure 5) exhibits obstructed cavities by the R
groups of the R2P- fragments, leaving the ClRh(CO)
unit well outside the macrocycle. The metallic centers
appear completely accessible. The computed Rh‚‚‚‚Rh
separation is about 7.7 Å, which is the same as the P‚
‚‚P separations, and indicates the absence of direct
interactions. The CPK models show also structures in
which the ClRh(CO) groups can turn freely about the
P‚‚‚P′ axis when R ) Me and i-Pr and requires some
cooperation of the latter when R ) Ph. The models
investigated for the 1,3-isomers (see an example pro-
vided below for 17) draw similar conclusions. In this
case, the cavity is obstructed by the presence of the
neighboring calix[4]arene, and the Rh atom is not
obscured.
The major conclusion extracted from modeling is
that the 1,2-diphosphinated calix[4]arene ligands
clearly favor dimer structures and consequently the
apical-apical coordination in the catalysis intermedi-
ates. In addition, because the Rh atoms are far apart
in the dimer structure, their catalytic activities are
probably uncorrelated. As for the 1,3-species, both
monomer (chelate) and dimer (bridging) forms are
possible, as little or no stress is demonstrated by
modeling.
Figure 6 presents the room-temperature absorption
spectrum of 14 as an example in the 340-640 nm range
and exhibits the three characteristic absorptions known
+
for the P2Ir(CdC)2 chromophore. These bands are
located at 394, 496, and 580 nm, and similarly at 394,
496, and 578 nm for 15. The lowest energy absorption
band was previously assigned to a metal-to-ligand
charge transfer (MLCT) from the iridium center to the
olefins on the basis of EHMO and Heller’s time-
dependent theory.27 Both complexes are luminescent in
77 K glasses, and their maxima are observed at 609 and
604 for 14 and 15, respectively. The emission bands are
weakly structured and exhibit excitation spectra that
superimpose their corresponding absorption spectra.
The emission lifetimes, τe, are 12.7 ( 0.2 and 11.4 (
0.2 µs for 14 and 15, respectively. These values indicate
that the emissions arise from the 3MLCT state and
compare favorably to that found for the calix[4]arene
Ir(COD)+ species discussed above (13.5 ( 0.1 and 13.7
( 0.1 µs for 18 and 19).
In ter a ction s betw een 1-Hexen e a n d 14 a n d 15.
Photophysics is used to corroborate the structures for
6c and 10c. While the Rh(COD)+ complexes investi-
gated in this work are either nonluminescent or very
weakly emissive, their corresponding iridium complexes
(14 and 15) are phosphorescent at 77 K matrixes. A
recent work from this group27 has shown that it is
possible to tell from the change in emission lifetime that
interactions are present with the calix[4]arene-contain-
Addition of 1-hexene at room temperature and im-
mediately freezing the solution at 77 K induces two
important changes. The first is a rapid decrease of the
(27) Gagnon, J .; Loeber, C.; Matt, D.; Harvey, D. P. Inorg. Chim.
Acta 1996, 242, 137.

2872 Organometallics, Vol. 22, No. 14, 2003
Plourde et al.
have been examined at 70 atm of 1:1 H2/CO and 50 °C.
These experimental conditions do not meet industrial
standards, as the pressure is too high.3a The purpose of
this study is to compare the performance for six
catalysts over five substrates, which exhibit very dif-
ferent rates of reactivity. In addition, the larger tof’s
provide more sensitivity and a more accurate compari-
son. Moreover, at high pressure the risk of isomerization
of 1-hexene into other isomers, which leads to other
isomer products, is significantly smaller, as experimen-
tally observed.
The generally accepted mechanism explaining the
presence of n- (linear) and i- (branched) products is
shown in Scheme 6.5a,29 The P2Rh(COD)+ precursor is
transformed into a monohydride carbonyl complex,
which undergoes an addition with the free olefin. The
key step is the insertion of the olefin into the Rh-H
bond (migratory insertion). Depending on which carbon
atom the hydride migrates to, two intermediates are
possible, leading to the n- or i-products. Subsequently,
addition of CO in the Rh coordination sphere promotes
the carbonyl migration-insertion into the Rh-C bond.
The final oxidative addition of H2 on this intermediate
produces a Rh(III) compound, which undergoes a reduc-
tive elimination to release the aldehyde compounds. For
these dimers, the intramolecular steric factors deduced
in the modeling of the trans-RhCl(CO) and Rh(COD)+
(always cis-) species favor the apical-apical geometry for
the intermediates.
F igu r e 7. Comparison of the decay traces of the emission
intensity for 14 in the presence of 1:0 (a), 1:10 (b), 1:100
(c), and 1:1000 (d) equivalents of 1-hexene in solution.
Sch em e 5
emission intensity with [1-hexene].28 The second is a
decrease in τe. Indeed, in both cases τe is sensitive to
the amount of added 1-hexene, as exemplified in Figure
7. The data for 14 are 12.7, 7.8, 7.1, and <5 µs for ratios
of 1:0, 1:10, 1:100, and 1:1000, and for 15, 11.4, 11.0,
10.3, 7.4, and 6.4 µs for the above ratios in the same
order (plus 1:5000 last datum).
In the absence of excess of free diphosphine, the
reaction proceeds efficiently, as exemplified in Figure
8. The stability of the active catalysts was checked by
UV-vis spectroscopy during the catalyses of all hydro-
formylation reactions of 1-hexene. This substrate was
investigated because there is no optical interference in
the absorption region of the catalysts. In all cases, the
The fact that τe decreases faster in 14 than in 15 with
[1-hexene] indicates the greater relative steric hin-
drance in 15 about the chromophore center. These
results corroborate the modeled structures presented in
Figures 2-4.
Ca t a lysis P r op er t ies. The homogeneous catalytic
hydroformylations of five terminal olefins (Scheme 5)
Sch em e 6

Catalytic Hydroformylation Properties
Organometallics, Vol. 22, No. 14, 2003 2873
Ta ble 4. Rea ction Ra tes (tof) for th e Ca ta lyzed Hyd r ofor m yla tion Rea ction s Usin g 6a -ca
6a
6b
6c
reaction
time (h)
conversion
((0.1%)
tof
reaction
time (h)
conversion
((0.1%)
reaction
time (h)
conversion
((0.1%)
tof
substrate
1-hexene
styrene
vinyl acetate
vinyl benzoate
vinyl p-tert-butylbezoate
(h-1
)
tof (h-1
)
(h-1
)
72
72
72
72
72
>99
58.9
35.2
82.2
21.8
51
35
5
14
3
30
30
48
48
48
>99
>99
>99
>99
>99
97
82
49
65
67
24
48
72
72
48
>99
>99
95.0
>99
>99
141
80
65
66
94
a
Conditions: P ) 70 atm, CO/H2 ) 1, T ) 50 °C, [alkene] ) 0.1 M, [Rh] ) 10-4 M (i.e., [6a ,b,c] ) 5 × 10-5 M). The reaction times were
72 h maximum. Any value different from 72 h indicates when the reaction was completed.
Ta ble 5. Rea ction Ra tes (tof) for th e Ca ta lyzed Hyd r ofor m yla tion Rea ction s Usin g 10a -ca
10a
10b
10c
reaction
time (h)
conversion
((0.1%)
tof
reaction
time (h)
conversion
((0.1%)
reaction
time (h)
conversion
((0.1%)
tof
substrate
1-hexene
styrene
vinyl acetate
vinyl benzoate
vinyl p-tert-butylbezoate
(h-1
)
tof (h-1
)
(h-1
)
24
72
72
72
72
>99
37.9
34.0
53.8
72.5
106
11
7
8
14
24
48
48
48
24
>99
>99
>99
>99
>99
211
65
41
63
78
24
72
72
72
72
>99
84.3
55.8
>99
>99
250
11
8
33
57
a
Conditions: P ) 70 atm, CO/H2 ) 1, T ) 50 °C, [alkene] ) 0.1 M, [Rh] ) 10-4 M (i.e., [10a ,b,c] ) 5 × 10-5 M). The reaction times were
72 h maximum. Any value different from 72 h indicates when the reaction was completed.
Ta ble 6. Regioselectivity (n /i r a tio) for th e
Ca ta lyzed Hyd r ofor m yla tion Rea ction s for th e
Ca ta lysts 6a -c a n d 10a -ca
n/i ratio (( 0.001)
substrate
1-hexene
6a
10a
6b
10b
6c
10c
1.807 1.703 1.857 1.816 1.718 1.708
0.059 0.057 0.105 0.119 0.086 0.070
0.047 0.052 0.093 0.074 0.051 0.092
0.006 0.005 0.006 0.003 0.008 0.005
0.001 0.005 0.005 0.003 0.004 0.002
styrene
vinyl acetate
vinyl benzoate
vinyl
p-tert-butylbenzoate
a
Conditions: P ) 70 atm, CO/H2 ) 1, T ) 50 °C, [alkene] ) 0.1
M, [Rh] ) 10-4 M (i.e., [6a ,b,c] and [10a ,b,c]) 5 × 10-5 M).
and 1,2-precursors exhibit a trend in tof varying as
Ph2P- > i-Pr2P- > Me2P-. For the four other sub-
strates, the 1,3- and 1,2-catalysts behave differently. For
the 1,3-derivatives, the tof’s vary as Ph2P- g i-Pr2P-
> Me2P- as well, but the latter phosphine gives poorer
results. For the 1,2-series, the tof’s decrease as i-Pr2P-
> Ph2P- g Me2P-, where both latter phosphines
exhibit modest rates. Despite the relatively small size,
the Me2P- group does not lead to large tof, but exhibits
better regioselectivity with respect to styrene and the
O-substituted vinyl substrates.
F igu r e 8. Evolution of the hydroformylation products for
the catalysis of 1-hexene by 10b (+ ) 1-hexene, b )
1-heptanal, 2 ) 2-methylhexanal).
UV-vis spectra in the 200-400 nm region were un-
changed during the reactions for all the experimental
conditions used.
The data show that for all six catalysts (Tables 4 and
5) the 1-hexene is clearly more reactive than the
conjugated vinyls (greater tof), consistent with the
nonstabilized olefin. Despite the different structure, the
styrene substrate reacts with tof’s in the same order of
magnitude as the three other O-substituted vinyls.
The comparison of the tof data as a function of the
catalyst indicates two features. For 1-hexene, both 1,3-
The regioselectivities for the hydroformylation reac-
tions are presented in Table 6. The data can be
separated into two categories: first, 1-hexene, which
exhibits a modest n/i ratio of ∼1.7-1.9; second, the
conjugated vinyls for which 0.120 > n/i > 0.001. The
better results are obtained for vinyl benzoate and p-tert-
butylbenzoate, where the n/i ratios are <0.01 (i.e.,
i-selectivity greater than 99%), with the vinyl tert-
butylbenzoate giving the better results. These observa-
tions are consistent with the models presented in
Scheme 7.3a During the hydroformylation reaction, a
metal-alkyl bond is formed during the insertion of the
H atom onto the coordinated olefin (Scheme 6). In the
alkyl-rhodium intermediates, this bond is polarized with
a partial positive charge on the metal and a partial
negative charge on the carbon (Scheme 7). When R is
an aromatic group, the latter permits delocalization of
(28) The overall intensity decreases dramatically with [1-hexene]
(a change that is greater than the change in lifetimes), so that efficient
quenching occurs. Since energy transfer from the 3MLCT to 3π* state
of 1-hexene cannot operate (i.e., E(3MLCT) , E(3π*)) and electron
transfer from the same donor and acceptor cannot proceed (no sufficient
driving force as the E0/1-(H2CdCHR) is too large), only a ground state
nonemissive (or weakly emissive) complex can explain this quenching,
as no change in emission band shape is observed. One possible
explanation is that 1-hexene displaces COD to form a nonluminescent
species of the type P2Ir(1-hexene)2+. Additions of 1-hexene to Rh-
(COD)+-containing solutions in 1:1 ratio exhibit 1H NMR spectra where
no free COD and 1-hexene are observed. This result indicates that
these ligands undergo chemical exchange in solution. In 14 and 15,
such exchange also occurs, but at 77 K, the process is stopped and the
solutions are composed of a mixture of the starting material in
deceasing amount according to the increase in [1-hexene] and the
nonluminescent species.

2874 Organometallics, Vol. 22, No. 14, 2003
Plourde et al.
Sch em e 7
this partial negative charge at the i-position. Similarly,
for electron-withdrawing groups such as acyls, inductive
effects delocalize this partial charge at the i-position
more than at the n-position. As a result, the i-isomer is
strongly favored. When R is an electron donor group
such as an alkyl group, no delocalization occurs for
either isomer. These isomers appear as products in
similar amounts, with the exception that the i-isomer
is less favored due to an inductive effect. Despite the
fact that these diphosphine ligands exhibit different
basicity (the electronic parameters ø are 13.25, 10.60,
and 7.50 cm-1, for i-Pr2P-, Ph2P-, and Me2P-, respec-
tively)30 and cone angles (155°, 145°, and 125°, respec-
tively),31 the n/i ratios vary only modestly.
Addition of 4 equiv of triphenylphosphine to the
reaction mixture at 50 °C and 70 atm for the hydro-
formylation of 1-hexene using 6b as an example, leads
to a 2-fold decrease in tof. The reaction still goes to 97%
completion after 48 h. This result indicates that the free
phosphine competitively blocks the coordinating site
about the rhodium center in the intermediate species,
rather than stabilizing it. Similarly, the n/ i ratio has
been only modestly modified from 1.86 to 2.04 in these
experiments.
Under soft conditions (P ) 7 atm, T ) 35 °C),
compounds 6b and 10b also perform the catalytic
hydroformylation of 1-hexene. At lower pressure, the
isomerization of 1-hexene is possible, and other minor
products are observed (less than 5% of the total amount
of products, and were ignored in these present analyses).
Their performances (tof) compare with the best most
known rhodium catalysts (Supporting Information).4a,7a,32
The best catalyst known so far is the BISBI-3,5-CF3/
Rh(CO)2(acac) for both tof and n/i.7a
The comparison of the hydroformylation reaction
of styrene is somewhat difficult to make because of
different conditions used (Supporting Informa-
tion),2d,4c,5c,33-35 but it indicates that the species 6a ,b,c
and 10a ,b,c are good catalysts. Compounds 6a and 10a
are the best in this series, in particular in comparison
with the second best catalyst for this substrate (Py-
diphos/Rh(CO)2(acac)). This comparison shows the in-
verse relationship between n/i ratios and tof.
F igu r e 9. Comparison of the decoupled (a) and coupled
(b) {1H}31P NMR spectra of 17.
also shows the better conversion rates and tof for 6b
and 10b (as examples). However, n/i ratios are compa-
rable, one average, with other systems.
Syn th esis of Hyd r id e 17. One hydride carbonyl
complex (one suspected intermediate in the hydroformyl-
ation reactions) has been prepared from the correspond-
ing Rh(COD)+ precursors according to the same condi-
tions applied for the catalyses:
50 °C/70 atm CO/H2
6b
8
THF/24 h
[1,3-(i-Pr2P)2calixRh(H)(CO)]2 (17) (4)
1
Complex 17 is characterized from H NMR, 31P NMR,
and IR spectroscopy and T1 measurements. The chemi-
cal shift (δ 31P ) 45.6 ppm) is consistent with a i-Pr2P-
1
Rh environment, and the coupling constant J P-Rh of
122.2 Hz is in line with a trans geometry. Compounds
12 and 13 exhibit similar constants (128 and 124 Hz,
respectively). A strong peak at 1970 cm-1 evidences the
presence of CO, and the T1 measurements (T1 ) 1.4 s)
indicate that the complex is a dimer in solution. The
1
signals belonging to ligand COD are absent in the H
-
NMR spectra, and PF6 is absent as well according to
both 31P NMR and IR. However, the hydride signal has
not been found yet. It is unclear whether it is too wide
to be measured, or it is placed at an unusual shift and
1
is lost under other H signals.36 Its presence can clearly
be felt in two ways. First, the T1 datum of 1.4 s is shorter
than that of 6b, the corresponding COD complex. This
hydride adds an extra nuclei (S ) 1/2) for dipole-dipole
31P‚‚‚1H interactions (distance ∼2.93 Å according to
models, a distance similar to that of the aryl-1H‚‚‚31P
interactions; Scheme 3). Second, the coupled and de-
coupled {1H}31P NMR spectra exhibit an important
change in full-width at half-maximum (fwhm) (Figure
9). The fwhm are 53 and 14 Hz for the coupled and
decoupled spectra, respectively. The 31P couples with the
The comparison of the catalytic performances for vinyl
acetate with literature data (Supporting Information)4c,6,34
(29) Broussard, M. E.; J uma, B.; Train, S. G.; Peng, W.-J .; Laneman,
S. A.; Stanley, G. G. Science 1993, 260, 784.
(30) Bartik, T.; Himler, T.; Schulte, H. G.; Seevogel, K. J . Orga-
nomet. Chem. 1984, 272, 29.
(31) Dias, P. B.; Minas de Piedade, M. E.; Simo˜es, J . A. M. Coord.
Chem. Rev. 1994, 135/ 136, 737.
(32) Hughes, O. R.; Unruh, J . D. J . Mol. Catal. 1981, 12, 71.
(33) Doyle, M. P.; Shanklin, M. S.; Zlokazov, M. V. Synlett 1994,
615.
(34) Basoli, C.; Botteghi, C.; Cabras, M. A.; Chelucci, G.; Marchetti,
M. J . Organomet. Chem. 1995, C20-22, 488.
(35) Freixa, Z.; Pereira, M. M.; Pais, A. A. C. C.; Bayo´n, J . C. J .
Chem. Soc., Dalton Trans. 1999, 3245.
(36) In a recent work, the Pd-H signal was found at +5.1 ppm
because of the proximity to the deshielding region of the aromatic
phenyl rings. See: Meilleur, D.: Harvey, P. D. Can. J . Chem. 2001,
79, 552.

Catalytic Hydroformylation Properties
Organometallics, Vol. 22, No. 14, 2003 2875
center can occur only on the outside of the dimer
structure.
F in a l Rem a r k s
The reported precursors are good catalysts for the
hydroformylation reaction of terminal alkenes. The
apical-apical geometry is highly probable in the hydro-
formylation intermediates, but there is no evidence for
important differences in performance, better or worse,
when one compares data with other catalysts in the
literature. Within the list of precursors and substrates
presented in this work, the choice of catalysts strongly
depends on the desired performance and the specific
alkene targets. By examining the best catalyst for
1-hexene for example as listed Table 6, dissymmetry
within the ligand provides better selectivity. The syn-
theses of dissymmetric diphosphinated calix[4]arene
ligands are not such a hard task and also still provide
the opportunity to take advantage of a cavity if the
structures are favorable.
F igu r e 10. Model structure for 17.
2
CH(CH3)3 nuclei as well. The coupling constant J HP is
about 11 Hz, as seen in the doublet of septets for 4b
(the other compounds giving poorly resolved peaks).
Taking into account the natural fwhm of decoupled
spectra (coupling with 103Rh excluded), the predicted
Ack n ow led gm en t. This research was supported by
the Natural Science and Engineering Research Council
(NSERC). F.P. and J .G. thank the Universite´ de Sher-
brooke and NSERC, respectively, for scholarships.
2
fwhm is only 21 Hz. However, if a second J PH coupling
occurs for the hydride, with a constant of ∼30 Hz, then
the predicted value (∼51 Hz) would fit the experimental
value (53 Hz). A small H-P coupling constant is
consistent with a cis geometry.7a In addition, the fwhm
for the 31P NMR signal for 4b is about 12 Hz, which
compares well with the 14 Hz value extracted form the
decoupled spectrum of 17.
The computer modeling of 17 (Figure 10) also exhibits
the face-to-face geometry. In comparison with 6b (Fig-
ure 2), the calix[4]arene macrocycles are placed further
from each other, but still space is available for a
molecule to penetrate inside the large bis-calix[4]arene
frame. Interactions between a substrate and the metal
Su p p or tin g In for m a tion Ava ila ble: Tables listing de-
tailed crystallographic data and structure refinement, atomic
coordinates parameters and isotropic displacement param-
eters, bond lengths and angles, anisotropic displacement
parameters, hydrogen coordinates and isotropic displacement
parameters, torsion angles, and least-squares planes devia-
tions, ORTEP figures for compounds 4a and 5b. Tables
comparing tof and n/ i ratios with literature for 1-hexene,
styrene, and vinyl acetate. This material is available free of
charge via the Internet at http://pubs.acs.org.
OM030095O