Synthesis of pyrrolyl-, indolyl-, and carbazolylphosphanes and their catalytic application as ligands in the hydroformylation of 2-pentene

Article abstract of DOI:10.1002/1099-0690(200110)2001:20<3871::AID-EJOC3871>3.0.CO;2-V

The synthesis of π-acceptor ligands of the type PAr_xR_3-x (x = 0-2; R = pyrrolyl, indolyl, carbazolyl; Ar = aryl) (1-8, 10, 12, 13) and P(pyrrolyl)2(carbazolyl) (11) is described. These ligands can be prepared in good to excellent yields by treatment of the corresponding free heterocyclic amines with phosphorus chlorides in the presence of base. The utilization of pyrrolyl-, indolyl-, and carbazolylphosphanes in the rhodium-catalyzed hydroformylation of 2-pentene demonstrates the influence of the ligand π-acidity on regioselectivity and activity in the hydroformylation of internal olefins. In general, increasing π-acidity of the ligand results in an increased yield of the linear oxo product. The best n/iso ratios of about 60:40 are obtained at low synthesis gas pressure (10 bar) in the presence of the P(pyrrolyl)₃ (1) ligand.

Full text of DOI:10.1002/1099-0690(200110)2001:20<3871::AID-EJOC3871>3.0.CO;2-V

FULL PAPER
Synthesis of Pyrrolyl-, Indolyl-, and Carbazolylphosphanes and Their Catalytic
Application as Ligands in the Hydroformylation of 2-Pentene
Ralf Jackstell,[a] Holger Klein,[a] Matthias Beller,*[a] Klaus-Diether Wiese,[b] and
Dirk Röttger[b]
Keywords: Homogeneous catalysis / Hydroformylation / Phosphanes / Rhodium
The synthesis of π-acceptor ligands of the type PArxR3−x (x =
0−2; R = pyrrolyl, indolyl, carbazolyl; Ar = aryl) (1−8, 10, 12,
13) and P(pyrrolyl)2(carbazolyl) (11) is described. These li-
ium-catalyzed hydroformylation of 2-pentene demonstrates
the influence of the ligand π-acidity on regioselectivity and
activity in the hydroformylation of internal olefins. In gen-
gands can be prepared in good to excellent yields by treat- eral, increasing π-acidity of the ligand results in an increased
ment of the corresponding free heterocyclic amines with
phosphorus chlorides in the presence of base. The utilization
yield of the linear oxo product. The best n/iso ratios of about
60:40 are obtained at low synthesis gas pressure (10 bar) in
of pyrrolyl-, indolyl-, and carbazolylphosphanes in the rhod- the presence of the P(pyrrolyl)3 (1) ligand.
Introduction
In order to obtain the linear aldehyde as the major prod-
uct, isomerization of the internal olefin must be faster than
the hydroformylation reaction. In addition, there should be
a reasonable difference in the rates of hydroformylation of
the internal and terminal olefins and the catalyst should be
highly n-selective for the terminal olefin.[3] It is well known
that carbonylcobalt-based homogeneous catalysts display
similar activities both for terminal and for internal olefins.
Unfortunately, cobalt catalysts show low turnover frequen-
cies and need relatively high reaction temperatures and
pressures (up to 190 °C and 250 bar).[4] While water-soluble
cobalt catalysts work under milder conditions, the n/iso ra-
tios of cobalt-catalyzed hydroformylation reactions are
comparatively low.[5]
The hydroformylation of olefins (Scheme 1) was disco-
vered in 1938 by Otto Roelen and is one of the most im-
portant homogenous catalytic processes in industry.[1] Now-
adays more than 6.6 million tons of aldehydes and alcohols
are produced annually from olefins by this method. The
primary commercial application of the oxo products is as
plasticizer alcohols, especially for manufacturing poly(vinyl
chloride). For the last 50 years the major product made
by hydroformylation has been 2-ethylhexanol, produced by
hydroformylation of propene, followed by an aldol con-
densation of the resulting n-butyraldehyde and subsequent
reduction.
In general, phosphanerhodium catalysts display much
higher activities for terminal olefins than for internal ol-
efins, which are converted very slowly with little isomeriz-
ation.[6] Coordinatively unsaturated rhodium species, how-
ever, exhibit activity towards isomerization of the substrate.
Such unsaturated species are formed in the presence of
sterically demanding phosphites or phosphanes. To date,
the most selective catalysts for the hydroformylation of in-
ternal olefins to give linear aldehydes are based on rhodium
modified by sterically hindered chelating phosphites.[7] Be-
cause of the limited stability of these ligands, there is con-
siderable interest in new hydroformylation catalysts for the
conversion of internal olefins. Clearly, phosphanes would
offer advantages over phosphites in terms of stability. Van
Leeuwen and co-workers have recently described an import-
ant advance, demonstrating that dibenzophospholyl- and
phenoxaphosphanyl-substituted xanthene ligands preferen-
tially give linear aldehydes in the hydroformylation of 2-
and 4-octene, albeit at rates too low for industrial applica-
tion.[8]
Scheme 1. Hydroformylation of terminal olefins
As a consequence of the debate concerning the environ-
mental impact of current plasticizer products, there is grow-
ing interest in new plasticizer alcohols.[2] In order to be vi-
able, the respective olefin feedstock has to be economically
competitive compared to propene. This requirement is only
fulfilled by mixtures of internal and terminal olefins, such
as raffinate II, which consists of 1-butene, (E/Z)-2-butene,
and butane. Thus, the selective hydroformylation of internal
olefins to give linear aldehydes is an important goal in car-
bonylation chemistry.
[a]
Institut für Organische Katalyseforschung (IfOK) an der
Universität Rostock e.V.,
Buchbinderstr. 5Ϫ6, 18055 Rostock, Germany
Building on our interest in using carbon monoxide for
the synthesis of bulk and fine chemicals,[9] we were curious
[b]
Oxeno Olefinchemie GmbH,
Industriepark Marl, 45764 Marl, Germany
Eur. J. Org. Chem. 2001, 3871Ϫ3877
WILEY-VCH Verlag GmbH, 69451 Weinheim, 2001
1434Ϫ193X/01/1020Ϫ3871 $ 17.50ϩ.50/0
3871

R. Jackstell, H. Klein, M. Beller, K.-D. Wiese, D. Röttger
FULL PAPER
about how the electronic properties of the phosphane li- Moloy[11] et al. have shown that 1-pyrrolylphosphanes can
gands would influence the n/iso ratios in the hydroformyl- be obtained in good yields by direct treatment of chloro-
ation of internal olefins. In general, a higher regioselectivity phosphanes with pyrrole in the presence of NEt3. As shown
is obtained with ligands that possess strong π-acceptor and in Scheme 3, this procedure was also successful, giving
weak σ-donor properties.[10] Investigations based on the νco tris(1-indolyl)- and tris(9-carbazolyl)phosphane in 90 and
band positions and 31P NMR parameters in Rh(acac)(- 92% yield, respectively. Hence, addition of 3 equiv. vari-
CO)L complexes[11] show the high π-acceptor character of ously of pyrrole, indole, or carbazole to PCl3 in THF at 0
1-pyrrolylphosphanes compared to phosphanes and even °C and subsequent heating to reflux gave ligands 1؊3 as
phosphites (Scheme 2). Despite their interesting electronic colorless solids after crystallization from n-hexane. Sim-
properties, 1-pyrrolylphosphanes have only rarely been ap- ilarly, ligands 4؊8 were synthesized by treatment of chloro-
plied as ligands in homogeneous catalysis. While Breit used diphenylphosphane with indole, carbazole, and substituted
tris(1-pyrrolyl)phosphane for the regioselective hydrofor- pyrroles, with yields higher than 60% (Scheme 4).
mylation of styrene with limited success,[12] Ziolkowski and
co-workers demonstrated that the hydroformylation of hex-
1-ene proceeds in good yield in the presence of tris(1-pyrro-
lyl)phosphane and 1-(diphenylphosphanyl)pyrrole.[13] In
this manuscript we describe the synthesis of a number of
novel substituted pyrrolyl-type phosphanes, as well as their
first application in the hydroformylation of an internal ol-
efin.
Scheme 2. π-Acceptor and σ-donor properties of different P ligands
Scheme 4. Synthesis of ligands 4؊8
Results and Discussion
Monitoring of the ligand synthesis by 31P NMR spectro-
Ligand Synthesis
scopy showed that the substitution steps at the PϪCl bonds
occur sequentially, with all intermediates Ϫ RPCl2, R2PCl,
and R3P Ϫ being observed during the course of the reac-
tion. This observation enabled a simple synthesis of bis(1-
pyrrolyl)phosphanes containing a third, different substitu-
ent at the phosphorus atom to be accomplished. Treatment
of 2 equiv. of pyrrole with PCl3 in THF gave chlorobis(1-
pyrrolyl)phosphane (9) in excellent yield (96%). Subsequent
treatment of 9 either with Grignard reagents or with carba-
zole afforded ligands 10؊13 in good yield (76Ϫ84%)
(Scheme 5). The structure of all new ligands was unambigu-
There are two known routes for the preparation of 1-
pyrrolylphosphanes. Treatment of alkali metal pyrrolides
with the appropriate phosphorus chloride produces 1-pyr-
rolylphosphanes in only moderate yields.[14] More recently
1
ously proven by means of H, 13C, and 31P NMR spectro-
scopy, together with mass spectrometry and elemental ana-
lysis.
Hydroformylation Experiments
All of the synthesized ligands (1؊8 and 10؊13) were
tested in the rhodium-catalyzed hydroformylation of 2-pen-
tene (Scheme 6, Tables 1Ϫ4), which was used as a model for
the oxo reaction of 2-butene (a major component in raffin-
ate II). The active catalysts were prepared in situ by mixing
Rh(acac)(CO)2 with the appropriate quantity of phosphane
in anisole as the solvent, in the presence of isooctane (in-
ternal standard for gas chromatography). As in industrial
hydroformylation processes, a comparatively small amount
Scheme 3. Synthesis of ligands 1؊3
3872
Eur. J. Org. Chem. 2001, 3871Ϫ3877

Pyrrolyl-, Indolyl-, and Carbazolylphosphanes
FULL PAPER
improved compared to that of the Rh/PPh3 system. This
demonstrates the influence of electronic parameters in con-
trolling the regioselectivity, since the phenyl and pyrrolyl
groups have similar steric bulks. In the presence of all these
ligands a decrease in the synthesis gas pressure resulted not
only in a decrease in the degree of conversion but also in
an increase in the quantity of linear aldehyde. The higher n/
iso selectivity is explained by the fact that at lower pressures
HRh(CO)(PR3)3 is the main active catalyst species, while
HRh(CO)2(PR3)2 becomes the more important active cata-
lyst at higher pressures. The latter catalyst is sterically cle-
arly less crowded and thus enables the hydroformylation
giving the branched product to proceed more easily. Re-
gardless of the synthesis gas pressure used, a small amount
of 1-pentene was detected in all reaction mixtures.
The most significant n/iso ratio difference between 50 and
25 bar synthesis gas pressure was found in the case of 1-
(diphenylphosphanyl)pyrrole (4) as the ligand (Table 1, Ent-
ries 6Ϫ8). With this ligand, both a higher total yield of
aldehyde and a higher yield of the desired linear n-hexanal
(compared to the reference systems) were obtained at 25
bar. Electron-withdrawing groups such as 3,4-diethoxycar-
bonyl or 2-acetyl at the pyrrolyl ring afford better n/iso ra-
tios, but also lower total yields of aldehyde than obtained
with ligand 4 (Table 1, Entries 13Ϫ15). In the case of 2-
acetyl-1-(diphenylphosphanyl)pyrrole (8), there was virtu-
ally no conversion under the standard reaction conditions,
which might be explained by the lower stability of the li-
gand.
We next attempted to improve the quantity of linear alde-
hyde obtained, by using ligands with two pyrrolyl groups.
Phenyl (10), 2-biphenylyl (12), and bis(3,5-trifluoromethyl)-
phenyl (13) groups were used as additional aromatic sub-
stituents on the phosphorus atom. In addition, a ligand
with two pyrrolyl substituents and one carbazole moiety
(11) was tested. As with ligands 4؊8, a decrease in the pres-
sure from 50 to 25 or 10 bar resulted both in a smaller
degree of conversion and in an increase in n-hexanal, the
desired product. In agreement with this concept, ligands 10,
12, and 13 (Table 2) gave better results than (1-diphenyl-
phosphanyl)pyrrole (4).
Scheme 5. Synthesis of ligands 10؊13
of metal (0.01 mol %) with a large excess of ligand (P/Rh ϭ
50:1Ϫ100:1) was used in the test reactions. In general, the
hydroformylation experiments were performed at 120 °C
and 25 or 50 bar of synthesis gas (H2/CO ϭ 1:1). It should
be noted that the n/iso ratio is defined throughout this study
as the ratio of 1-hexanal to 2-methylpentanal and 2-ethyl-
butanal (Scheme 6).
Scheme 6. Hydroformylation of 2-pentene
At 50 bar, all ligands except 11 gave very good yields of
At the start of our investigations, RPPh2 (R ϭ pyrrolyl, C6 aldehydes (95Ϫ99%), with greater steric demand in the
indolyl, carbazolyl) ligands 4؊8 were compared with the aryl group resulting in an increased quantity of linear alde-
well-known Rh/PPh3 catalyst system and a phosphane-free hyde. Hence, ligand 10 gave an n/iso ratio of 20:80, while
Rh catalyst (so-called ‘‘unmodified Rh‘‘) (see Table 1). The ligands 12 and 13 gave 34:66 and 30:70, respectively. Such
phosphane-free catalyst afforded 96% conversion and an n/ behavior has also been observed with phosphite ligands.[15]
iso ratio of 39:61 at 50 bar, while at 25 bar the conversion Nevertheless, at high pressure (50 bar) none of the ligand-
was only 56% after 6 h (n/iso ϭ 41:59). The Rh/PPh3 cata- modified catalyst systems gave an n/iso ratio higher than
lyst gave complete conversion at 50 bar synthesis gas and that afforded by the ligand-free catalyst. Advantageously,
79% conversion at 25 bar. High selectivities (n/iso Ͼ 7:93) however, this was not true for hydroformylations performed
for the branched products were observed in both cases at lower pressures (25 and 10 bar). Here, ligands 10 and 13
(Table 1, Entries 1Ϫ5).
not only gave rise to higher selectivities for n-hexanal, but
Compared to the reference systems, the 1-(diphenylphos- also produced total yields of aldehyde significantly higher
phanyl)pyrrole ligands 4, 7, and 8, and also the 1-(di- than those afforded by the unmodified carbonylrhodium
phenylphosphanyl)indole ligands 5 and the (9-carbazole)di- catalyst (Table 2, Entries 3Ϫ4, 11Ϫ12).
phenylphosphanyl ligands 6, gave lower degrees of conver-
In order to evaluate the influence of π-acceptor pyrrole-
sion. On the other hand, the n/iso ratio was significantly like phosphanes systematically, a series of experiments was
Eur. J. Org. Chem. 2001, 3871Ϫ3877
3873

R. Jackstell, H. Klein, M. Beller, K.-D. Wiese, D. Röttger
FULL PAPER
Table 1. Hydroformylation of 2-pentene with PRPh2 ligands [general conditions: 23.7 mmol (1.67 g) of 2-pentene, 0.00237 mmol (0.61 mg)
of Rh(acac)(CO)2, 25 mL of anisole, 2 mL of isooctane, 120 °C, 6 h]
Entry
L
Rh [mol %]
Rh/L
p [bar]
Total yield of aldehydes [%]
n/iso
1
Ϫ
Ϫ
Ϫ
PPh3
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
Ϫ
Ϫ
Ϫ
50
25
10
50
25
50
25
17
50
25
50
25
50
25
50
50
25
96
56
10.5
99
79
78
66
48
83
34
74
43
59
33
3
39:61
41:59
33:67
7:93
8:92
9:91
33:67
36:64
20:80
27:73
27:73
37:63
24:76
37:63
19:81
29:71
38:62
2
3[a]
4
1:100
1:100
1:100
1:100
1:100
1:100
1:100
1:100
1:100
1:100
1:100
1:100
1:50
5
6
7
8
PPh3
4
4
4
5
5
6
6
7
7
8
8
8
9
10
11
12
13
14
15
16
17
34
14
1:50
[a]
16 h.
Table 2. Hydroformylation of 2-pentene with PR(pyrrolyl)2 as li-
Table 4. Hydroformylation of 2-pentene with P(pyrrolyl)3 ligand
gands [general conditions: 23.7 mmol (1.67 g) of 2-pentene, [general conditions: 23.7 mmol (1.67 g) of 2-pentene, 0.00237 mmol
0.00237 mmol (0.61 mg) of Rh(acac)(CO)2, 25 mL of anisole, 2 mL (0.61 mg) or 0.00711 mmol (1.83 mg) of Rh(acac)(CO)2, 25 mL of
of isooctane, 120 °C, 6 h]
anisole, 2 mL of isooctane, 120 °C, 6 h]
Entry
L
Rh [mol %] Rh/L p [bar] Total yield of n/iso
Entry
L
Rh [mol %] Rh/L p [bar] Total yield of n/iso
aldehydes [%]
aldehydes [%]
1
2
3
4
5
6
7
8
9
10
10
10
10
11
11
12
12
12
13
13
13
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
1:100 50
1:100 25
1:100 17
1:100 10
1:100 50
1:100 25
1:100 50
1:100 25
1:100 17
1:100 50
1:100 25
1:100 10
95
81
53
25
60
28
96
43
26
99
73
23
20:80
39:61
45:55
54:46
27:73
37:63
34:66
38:62
39:61
30:70
43:57
52:48
1
2
1
1
1
1
1
1
1
0.01
0.01
0.01
0.01
0.01
0.03
0.03
1:100 50
1:100 25
89
88
73
67
40
56
11
23:77
47:53
46:54
52:48
56:44
60:40
61:39
1:50
25
3
4
5
6
1:100 17
1:100 10
1:50
1:50
10
5
10
11
12
tained at 50 bar synthesis gas pressure in the presence of all
three ligands, with regioselectivities somewhat lower than
those obtained with the ligand-free catalyst system. When
2-pentene was treated in the presence of tris(1-pyrrolyl)pho-
sphane (1) at 25 bar, an 89% yield of C6 aldehydes was
obtained and n-hexanal was formed in 43% yield (Table 3,
Entry 2). Here, both the conversion and the product select-
ivity towards the linear aldehyde were significantly higher
than those obtained from reactions in the presence of the
reference catalyst systems.
Table 3. Hydroformylation of 2-pentene with PR3 (R ϭ pyrrolyl,
indolyl, carbazolyl) ligands [general conditions: 23.7 mmol (1.67 g)
of 2-pentene, 0.00237 mmol (0.61 mg) of Rh(acac)(CO)2, 25 mL of
anisole, 2 mL of isooctane, 120 °C, 6 h]
Entry
L
Rh [mol %] Rh/L p [bar] Total yield of n/iso
aldehydes [%]
Catalysts based on ligands 2 and 3, which possess less π-
acceptor and more basic character than 1, produced signi-
ficantly lower degrees of conversion/total yield of aldehydes
at lower pressure (Table 3, Entries 4, 6). The catalysis results
obtained clearly show that the consecutive substitution of
phenyl groups by π-acceptor pyrrole units in triphenylphos-
phane results in ligands that can provide an increased
amount of linear aldehyde in the hydroformylation of 2-
1
2
3
4
5
6
1
1
2
2
3
3
0.01
0.01
0.01
0.01
0.01
0.01
1:100 50
1:100 25
1:100 50
1:100 25
1:100 50
1:100 25
89
89
90
34
86
46
23:77
47:53
38:62
40:60
37:63
40:60
performed with tris(1-pyrrolyl)phosphane (1), tris(1-indo- pentene. Comparing the results obtained from the model
lyl)phosphane (2), and tris(9-carbazolyl)phosphane (3) (see reaction in the presence of the different ligands, it is obvious
Table 3). Again, excellent yields of C6 aldehydes were ob- that ligands 4؊7 give the lowest n/iso ratios, both at 50 and
3874
Eur. J. Org. Chem. 2001, 3871Ϫ3877

Pyrrolyl-, Indolyl-, and Carbazolylphosphanes
FULL PAPER
Hz). Ϫ Mass spectra were recorded with an AMD 402Ϫ3 (Intectra
GmbH) spectrometer. Ϫ Gas chromatograms were recorded with
a HewlettϪPackard HP 6890 series gas chromatograph (column:
HP1; 25 m).
at 25 bar. The total yield of C6 aldehydes is also comparat-
ively low with these ligands. At 50 bar, phosphanes with
more than one pyrrolyl, indolyl, or carbazolyl group give
results similar to those obtained using the unmodified Rh
catalyst. At 25 bar synthesis gas pressure, however, good
yields of C6 aldehydes are observed with ligands 1, 10,
and 13.
Electronic effects aside, the n/iso ratio can be further im-
proved by the introduction of steric hindrance, as shown for
the series of arylbis(1-pyrrolyl)phosphanes. Interestingly, of
all the ligands tested, tris(1-pyrrolyl)phosphane (1) gave the
highest n/iso ratio at 25 bar. On account of this we con-
ducted an additional optimization study of the hydrofor-
mylation of the model system with this ligand (see Table 4).
A further decrease in the synthesis gas pressure to 10 bar
produced an increase in the n/iso ratio, to 56% n-hexanal
and 44% of isopentanals, although at lower pressure the
total yields of aldehyde was reduced to 40%. We thus per-
formed two experiments, at 10 and 5 bar with higher cata-
lyst loadings (300 ppm of rhodium) (Table 4, Entries 5Ϫ6).
Thanks to the higher ligand concentration in solution (com-
pare Entries 4 and 5), the n/iso ratio was increased to 60:40
at 10 bar, while the total yield of C6 aldehydes increased
to 56%.
Hydroformylation Experiments: 2-Pentene (2.6 mL, 1.7 g,
23.7 mmol), the corresponding amount of ligand, and isooctane
(2 mL) as internal standard were added under argon to a solution
of Rh(acac)(CO)2 (0.61 mg, 2.37 ϫ 10Ϫ6 mol, 0.01 mol % for pen-
tene) in anisole (22 mL). This mixture was transferred to a cooled
(0 °C) 100-mL stainless steel autoclave equipped with a magnetic
stirrer, and pressurized to 5 bar synthesis gas (CO/H2 ϭ 1:1). The
autoclave was heated to 120 °C (ca. 15 min). The pressure at this
temperature was adjusted to the desired value, and this value was
maintained throughout the reaction (6 h). The autoclave was then
cooled to 0 °C and the pressure was released. The resulting reaction
mixture was analyzed by gas chromatography.
Ligand Synthesis. ؊ General Procedure for the Preparation of Li-
gands 1؊8 and 10 (Method 1): PCl3, PPh2Cl, or PPhCl2 (20 mmol)
was added to a cooled (0 °C) solution of NEt3 (60, 40, or 20 mmol)
in THF (100 mL) in a 250-mL three-necked, round-bottomed flask.
A solution of the corresponding stoichiometric amount of pyrrole,
indole or carbazole in THF (20 mL) was then added while the tem-
perature was kept at 0 °C. The reaction mixture was stirred for 1 h
at room temperature and then heated to reflux for a further 6 h.
After this had cooled to room temperature, the solvent was re-
moved in vacuo and the residue was dissolved in 100 mL of toluene.
The mixture was filtered and the solvent removed in vacuo. The
title compounds were crystallized from hot n-hexane. The products
were all colorless solids, except for (pyrrole)PPh2, which was a col-
orless oil.
Conclusion
In conclusion, we have shown that a number of 1-pyrro-
lyl-like phosphanes can easily be synthesized by treatment
of phosphorus chlorides and N-heteroaromatics or by treat-
ment of chlorobis(1-pyrrolyl)phosphane with Grignard re-
agents.
Tris(1-pyrrolyl)phosphane (1): Method 1, M ϭ 229.08 g/mol, yield
65%. Ϫ 31P NMR(166 MHz, CDCl3): δ ϭ 79.6. Ϫ 13C NMR
(166 MHz, CDCl3): δ ϭ 122.8 (d, JPC ϭ 14.3 Hz), 113 (d, JPC
ϭ
4.8 Hz). Ϫ 1H NMR (400 MHz, CDCl3): δ ϭ 6.77 (pquint, J ϭ
2.0 Hz, 6 H), 6.34 (pt, J ϭ 2.0 Hz, 6 H). - MS (70 eV): m/z (%) ϭ
230 (11) [Mϩ ϩ 1], 229 (86) [Mϩ], 164 (11), 163 (100), 136 (65),
135 (26), 118 (13), 97 (12), 96 (21), 70 (27), 69 (25). Ϫ HRMS:
calcd. for C12H12N3P [Mϩ] 229.07689; found 229.07599.
The application of these ligands to the rhodium-catalyzed
hydroformylation of internal olefins has been demonstrated
for the first time. At higher pressures, the Rh/PR3 systems
behave similarly to unmodified carbonylrhodium catalysts.
At lower synthesis gas pressures (Ͻ 25 bar), however, im-
proved results are obtained with ligands 1, 10 and 13. Al-
though the observed regioselectivities are far from satisfact-
ory, the results demonstrate that the hydroformylation of
internal olefins proceeds in the presence of certain pyrrolyl-
phosphanes in good to very good yields (total yield of all
aldehydes). Even with simple monophosphanes the n/iso ra-
tio can be modified to some extent, to afford the linear
aldehyde as the major product. Improvements in the re-
gioselectivity might be anticipated from use of chelating
and sterically more demanding ligands. We are at present
conducting further work in this area.
Tris(1-indolyl)phosphane (2): Method 1, M ϭ 379.12 g/mol, yield
90%. Ϫ 31P NMR (166 MHz, CDCl3): δ ϭ 67.2. Ϫ 13C NMR
(166 MHz, CDCl3): δ ϭ 139.0 (d, JPC ϭ 19.1 Hz), 131.0 (d, JPC
ϭ
2.9 Hz), 127.0, 123.5, 122.0, 121.3, 111.5 (d, JPC ϭ 14.3 Hz), 108.5.
1
Ϫ H NMR (400 MHz, CDCl3): δ ϭ 7.45Ϫ7.6 (m, 6 H), 7.05Ϫ7.2
(m, 6 H), 6.85Ϫ6.92 (m, 3 H), 6.55Ϫ6.65 (m, 3 H). Ϫ MS (70 eV):
m/z (%) ϭ 379 (36.8) [Mϩ], 263 (100), 147 (10), 116 (11). Ϫ HRMS:
calcd. for C24H18N3P [Mϩ] 379.12384; found 379.12387.
Tris(9-carbazolyl)phosphane (3): Method 1, M ϭ 529.17 g/mol,
yield 92%. - 31P NMR (166 MHz, CDCl3): δ ϭ 77.2. Ϫ 13C NMR
(166 MHz, CDCl3): δ ϭ 142.48 (d, JPC ϭ 9.1 Hz), 140.1, 139.8,
124.3, 120.9, 113.7 (d, JPC ϭ 13.1 Hz). Ϫ 1H NMR (400 MHz,
CDCl3): δ ϭ 7.8Ϫ8.0 (m, 6 H), 7.3Ϫ7.4 (m, 2 H), 6.80Ϫ7.20 (m,
16 H). Ϫ MS (70 eV): m/z (%) ϭ 529 (35) [Mϩ], 363 (100), 197
(20), 166 (30). Ϫ C36H24N3P: calcd. C 81.6, H 4.6, N 7.9; found C
81.6, H 4.55, N 8.3.
Experimental Section
General Information: All reactions were carried out under an inert
gas, using standard high-vacuum Schlenk line techniques. Diethyl
ether, tetrahydrofuran, toluene, and hexane were distilled under ar-
gon from NaK/benzophenone prior to use. Ϫ 1H, 13C, and 31P
NMR spectra were recorded with a Bruker ARX 400 NMR instru-
1-(Diphenylphosphanyl)pyrrole (4): Method 1, M ϭ 251.09 g/mol,
yield 85%. - 31P NMR (166 MHz, CDCl3): δ ϭ 48.2. Ϫ 13C NMR
(166 MHz, CDCl3): δ ϭ 136.9 (d, JPC ϭ 12.4 Hz), 131.9 (d, JPC
ϭ
21 Hz), 129.6, 128.5 (d, JPC ϭ 6.7 Hz), 125.4 (d, JPC ϭ 12.4), 111.5
(d, JPC ϭ 3.8 Hz). Ϫ 1H NMR (400 MHz, CDCl3): δ ϭ 7.15Ϫ7.35
ment, using the solvent as an internal standard (δ in ppm, J in (m, 10 H), 6.75 (pquint, J ϭ 2.0 Hz, 2 H), 6.24 (t, J ϭ 2.18 Hz, 4
Eur. J. Org. Chem. 2001, 3871Ϫ3877 3875

R. Jackstell, H. Klein, M. Beller, K.-D. Wiese, D. Röttger
FULL PAPER
H). Ϫ MS (70 eV): m/z (%) ϭ 251 (98) [Mϩ], 183 (100), 174 (14), (d, 3JPC ϭ 4.7 Hz). Ϫ 1H NMR (400 MHz, C6D6): δ ϭ 6.7 (pquint,
152 (13), 107 (16), 67 (27). Ϫ HRMS: calcd. for C16H14NP [Mϩ] J ϭ 2.2 Hz, 2 H), 6.18 (pt, J ϭ 2.2 Hz, 2 H).
251.08640; found 251.08713.
Preparation of (9-Carbazolyl)bis(1-pyrrolyl)phosphane (11): P(pyr-
1-(Diphenylphosphanyl)indole (5): Method 1, M ϭ 301.16 g/mol,
yield 67%. - 31P NMR (161 MHz, CDCl3): δ ϭ 35.8. Ϫ 13C NMR
(166 MHz, CDCl3): δ ϭ 141.1 (d, 2JPC ϭ 18.1 Hz), 136.2 (d, 2JPC ϭ
12.4 Hz), 131.9 (d, 1JPC ϭ 21.0 Hz), 130.4 (d, JPC ϭ 2.8 Hz), 130.1
(d, JPC ϭ 2.8 Hz), 129.6, 128.6 (d, JPC ϭ 6.7 Hz), 122.1, 112.2
(d, JPC ϭ 15.3 Hz), 106.5. Ϫ 1H NMR (400 MHz, CDCl3): δ ϭ
7.75Ϫ7.70 (m, 1 H), 7.40Ϫ7.20 (m, 11 H), 7.20Ϫ7.05 (m, 2 H),
6.93Ϫ6.86 (m, 1 H), 6.60Ϫ6.55 (m, 1 H). Ϫ MS (70 eV): m/z (%) ϭ
301 (100) [Mϩ], 222 (10), 183 (65), 152 (10), 107 (10), 77 (8). Ϫ
C20H16NP: calcd. C 79.7, H 5.4, N 4.65; found C 79.5, H 5.1, N 5.1.
rolyl)2Cl (9) (3.98 g, 20 mmol) was added to a cooled (0 °C) solu-
tion of NEt3 (2.5 g, 25 mmol) in THF (50 mL) in a 100-mL three-
necked, round-bottomed flask. A stoichiometric amount of carba-
zole in THF (20 mL) was then added, while the temperature was
maintained at 0 °C. The reaction mixture was stirred for 1 h at
room temperature and then heated to reflux for a further 6 h. After
this had cooled to room temperature, the solvent was removed in
vacuo and the residue was dissolved in 50 mL of toluene. The mix-
ture was filtered and the remaining solvent was removed in vacuo.
After crystallization from hot n-hexane, 11 was obtained (5.2 g,
79% yield). Ϫ M ϭ 329.11 g/mol. Ϫ 31P NMR (166 MHz, CDCl3):
9-(Diphenylphosphanyl)carbazole (6): Method 1, M ϭ 351.28 g/mol,
yield 71%. Ϫ 31P NMR (166 MHz, CDCl3): δ ϭ 32.2. Ϫ 13C NMR
(166 MHz, CDCl3): δ ϭ 144.1 (d, 2JPC ϭ 7.6 Hz), 134.8 (d, 2JPC ϭ
13.4 Hz), 131.8 (d, JPC ϭ 20.0 Hz), 129.7, 129.0 (d, JPC ϭ 5.7 Hz),
126.3, 126.0, 121.2, 120.4, 114.1 (d, JPC ϭ 12.4 Hz). Ϫ 1H NMR
(400 MHz, CDCl3): δ ϭ 8.07Ϫ8.04 (m, 2 H), 7.52Ϫ7.50 (m, 2 H),
7.46Ϫ7.41 (m, 4 H), 7.34Ϫ7.30 (m, 6 H), 7.28Ϫ7.22 (m, 4 H). Ϫ
MS (70 eV): m/z (%) ϭ 351 (100) [Mϩ], 274 (6), 185 (70), 166 (10),
107 (10), 77 (3). Ϫ C24H18NP: calcd. C 82.02, H 5.17, N 3.99;
found C 82.08, H 5.25, N 4.12.
2
δ ϭ 80.4. Ϫ 13C NMR (166 MHz, CDCl3): δ ϭ 141.8 (d, JPC
ϭ
2
8.6 Hz), 126.5, 126.3 (d, JPC ϭ 1.9 Hz), 122.9 (d, JPC ϭ 13.4 Hz),
122.0, 120.1, 113.2 (d, 3JPC ϭ 3.8 Hz), 112.8 (d, JPC ϭ 13.4 Hz). Ϫ
1H NMR (400 MHz, CDCl3): δ ϭ 8.00Ϫ7.90 (m, 2 H), 7.28Ϫ7.16
(m, 4 H), 7.12Ϫ7.04 (m, 2 H), 6.79 (pquint, J ϭ 2.0 Hz, 4 H),
6.25Ϫ6.3 (m, 4 H). Ϫ MS (70 eV): m/z (%) ϭ 329 (37) [Mϩ], 263
(60), 167 (100), 140 (13). Ϫ HRMS: calcd. for C20H16N3P
329.10818; found 329.10895.
General Procedure for the Preparation of Ligands 12؊13 (Method
2): A Grignard reagent was prepared from the appropriate aryl
bromide and magnesium in Et2O in a 50-mL three-necked, round-
bottomed flask. The Grignard solution was then added to a solu-
tion of (pyrrolyl)2PCl (3.98 g, 20 mmol) in THF (50 mL) in a 100-
mL three-necked, round-bottomed flask. The reaction mixture was
stirred for 1 h at room temperature and heated to reflux for a fur-
ther 6 h. After the reaction mixture had cooled to room temper-
ature, the solvent was removed in vacuo and the residue was dis-
solved in 50 mL of toluene. The mixture was filtered and the solv-
ent was removed in vacuo. The title compounds were crystallized
from hot n-hexane.
Diethyl
1-(Diphenylphosphanyl)pyrrole-3,4-dicarboxylate
(7):
Method 1, M ϭ 395.13 g/mol, yield 80%. Ϫ 31P NMR (166 MHz,
CDCl3): δ ϭ 56.4. Ϫ 13C NMR (166 MHz, CDCl3): δ ϭ 163.8,
2
134.3 (d, JPC ϭ 13.3 Hz), 132.0, 131.8 (d, JPC ϭ 12.4 Hz), 130.5,
128.4 (d, JPC ϭ 7 Hz), 118.5, 60.0, 14.2. Ϫ 1H NMR (400 MHz,
CDCl3): δ ϭ 7.3Ϫ7.4 (m, 2 H), 7.15Ϫ7.25 (m, 2 H), 4.20 (q, J ϭ
7.0 Hz, 4 H), 1.20 (t, J ϭ 7.0 Hz, 6 H). Ϫ MS (70 eV): m/z (%) ϭ
396 (18) [Mϩ ϩ 1], 395 (83) [Mϩ], 350 (19), 323 (8), 185 (100). Ϫ
HRMS: calcd. for C22H22NPO4 [Mϩ] 395.12863; found 395.12749.
2-Acetyl-1-(diphenylphosphanyl)pyrrole (8): Method 1, M ϭ
293.1 g/mol, yield 77%. Ϫ 31P NMR (166 MHz, CDCl3): δ ϭ 55.4.
(1,1Ј-Biphenyl-2-yl)bis(1-pyrrolyl)phosphane (12): Method 2, M ϭ
316.11 g/mol, yield 84%. Ϫ 31P NMR (166 MHz, CDCl3): δ ϭ
Ϫ
13C NMR (166 MHz, CDCl3): δ ϭ 188.4, 138, 137.8, 136.7 (d,
JPC ϭ 6.7 Hz), 133.1 (d, JPC ϭ 7.6 Hz), 133.0 (d, JPC ϭ 22.9 Hz),
129.0 (d, JPC ϭ 6.7 Hz), 122.0 (d, JPC ϭ 3.8 Hz), 111.7, 26.6. Ϫ
1H NMR (400 MHz, CDCl3): δ ϭ 7.26Ϫ7.32 (m, 6 H), 7.14Ϫ7.20
(m, 4 H), 7.03 (dt, J ϭ 1.4, J ϭ 3.8 Hz, 1 H), 6.31Ϫ6.33 (m, 1 H),
6.12 (pt, J ϭ 3.0 Hz, 1 H), 2.3 (s, 3 H).
68.4. Ϫ 13C NMR (166 MHz, CDCl3): δ ϭ 146.4 (d, JPC
ϭ
27.7 Hz), 139.7 (d, JPC ϭ 5.7 Hz), 135.1 (d, JPC ϭ 13.4 Hz), 130.5
(d, JPC ϭ 11.5 Hz), 130.4, 129.8 (d, JPC ϭ 4.8 Hz), 128.7 (d, JPC ϭ
3.8 Hz), 127.9, 127.6, 127.5, 124.3 (d, JPC ϭ 14.3 Hz), 112.0 (d,
J
PC ϭ 3.8 Hz). Ϫ 1H NMR (400 MHz, CDCl3): δ ϭ 6.25 (pt, 4 H),
Phenylbis(1-pyrrolyl)phosphane (10): Method 1, M ϭ 229.22 g/mol,
6.7 (pquint, 4 H), 6.74Ϫ6.77 (m, 1 H), 6.96Ϫ6.98 (m, 2 H),
yield 76%. - 31P NMR (166 MHz, CDCl3): δ ϭ 70.5. Ϫ 13C NMR 7.20Ϫ7.27 (m, 4 H), 7.33 (tt, J ϭ 7.0 , J ϭ 0.8 Hz, 1 H), 7.45 (td,
(166 MHz, CDCl3): δ ϭ 136.9 (d, JPC ϭ 4.8 Hz), 130.9, 130.5 (d, J ϭ 7.0, J ϭ 0.8 Hz, 1 H). Ϫ MS (70 eV): m/z (%) ϭ 316 (30)
JPC ϭ 10.5 Hz), 129.0, 125.2 (d, JPC ϭ 16.6 Hz), 112.8 (d, JPC
ϭ
[Mϩ], 250 (15), 183 (100), 154 (30), 67 (15), 77 (5). Ϫ C20H17N2P:
1
4.0 Hz). Ϫ H NMR (400 MHz, CDCl3): δ ϭ 7.25Ϫ7.33 (m, 3 H), calcd. C 75.9, H 5.4, N 8.9; found C 75.75, H 5.67, N 8.83.
6.98Ϫ7.00 (m, 2 H), 6.84 (pquint, J ϭ 2.0 Hz, 4 H), 6.2 (br. s, 4
H). Ϫ MS (70 eV): m/z (%) ϭ 240 (7) [Mϩ], 174 (100), 147 (19),
152 (13), 96 (12), 107 (16), 77 (11). Ϫ HRMS: calcd. for C14H13N2P
[Mϩ] 240.08279; found 240.08163.
[3,5-Bis(trifluoromethyl)phenyl]bis(1-pyrrolyl)phosphane
(13):
Method 2, M ϭ 376.06 g/mol, yield 80%. Ϫ 31P NMR (166 MHz,
CDCl3): δ ϭ 64.1. Ϫ 13C NMR (166 MHz, CDCl3): δ ϭ 113.5 (d,
JPC ϭ 4.8 Hz), 123.9, 124.5 (d, JPC ϭ 15.2 Hz), 130, 130.3 m, 132
(d, JPC ϭ 3.8 Hz), 141 (d, JPC ϭ 11.44). ؊ 1H NMR (400 MHz,
CDCl3): δ ϭ 6.30 (pt, J ϭ 2.18 Hz, 4 H), 6.85 (pquint, J ϭ 2.2, 4
H), 7.34 (s, 1 H), 7.35 (s, 1 H), 7.85 (s, 1 H). Ϫ MS (70 eV): m/z
(%) ϭ 376 (100) [Mϩ], 310 (87), 241 (17), 213 (4), 172 (4), 163
(38), 69 (5). Ϫ HRMS: calcd. for C16H11N2PF6 [Mϩ] 376.05640;
found 376.05590.
Preparation of Chlorobis(1-pyrrolyl)phosphane (9): PCl3 (7.0 mL,
80 mmol) was added to a cooled (0 °C) solution of NEt3 (25 mL,
18 mmol) in THF (200 mL) in a 500-mL three-necked, round-bot-
tomed flask. A solution of pyrrole (11.2 mL, 160 mmol) in THF
(20 mL) was then added, while the temperature was maintained at
0 °C. The reaction mixture was stirred for 6 h at room temperature.
The mixture was then filtered and the solvent removed in vacuo.
The residue was distilled in vacuo at 10Ϫ3 Torr to afford 9 (15.3 g,
96% yield) as a colorless liquid. Ϫ B.p. 60 °C (10Ϫ3 Torr); M ϭ
198.59 g/mol. Ϫ 31P NMR (166 MHz, C6D6): δ ϭ 104.7. Ϫ
[1]
For reviews on hydroformylation reactions see: [1a] A. M. Trzec-
iak, J. J. Ziolkowski, Coord. Chem. Rev. 1999, 190Ϫ192, 883.
2
13C NMR (166 MHz, C6D6): δ ϭ 122.6 (d, JPC ϭ 17.2 Hz), 113.9
Ϫ
[1b] C. D. Frohning, C.W. Kohlpaintner in Applied Homogen-
3876
Eur. J. Org. Chem. 2001, 3871Ϫ3877

Pyrrolyl-, Indolyl-, and Carbazolylphosphanes
FULL PAPER
eous Catalysis with Organometallic Compounds (Eds.: B.
M. Beller, M. Eckert, F. Vollmüller, S. Bogdanovic, H. Geissler,
Angew. Chem. 1997, 109, 1534; Angew. Chem. Int. Ed. Engl.
1997, 36, 1494.
Cornils, W. A. Herrmann), VCH Verlagsgesellschaft,
[1c]
Weinheim, 1996, vol. 1. Ϫ
M. Beller, B. Cornils, C. D.
[10] [10a]
Frohning, C. W. Kohlpaintner, J. Mol. Catal. 1995, 104, 17.
‘‘Europe votes on PVC toy ban’’, Chem. Ind. 1998, 18, 379.
B. Zimmermann, H. Geissler, M. Beller, Chem. Eur. J. 1999,
5, 1301.
B. Cornils, ‘‘Hydroformylation, Oxosynthesis, Roelen Reac-
tion’’ in New Syntheses with Carbon Monoxide (Ed.: J. Falbe),
Springer, Berlin, 1980, p. 1.
M. Beller, J. G. E. Krauter, J. Mol. Catal. 1999, 143, 31.
C. Botteghi, R. Ganzerla, M. Lenarda, G. Morretti, J. Mol.
Catal. 1987, 40, 129.
J. Grimblot, J. P. Bonnelle, A. Mortreux, F. Petit, Inorg.
[2]
[3]
[10b]
Chim. Acta 1979, 34. Ϫ
J. Grimblot, J. P. Bonnelle, C.
Vaccher, A. Mortreux, F. Petit, J. Mol. Catal. 1980, 9, 357. Ϫ
[10c]
C. Vaccher, A. Mortreux, F. Petit, J. Mol. Catal. 1981, 12,
329. Ϫ [10d] J. D. Unruh, J. R. Christenson, J. Mol. Catal. 1982,
[4]
[10e]
14, 19. Ϫ
H. Janecko, A. M. Trzeciak, J. J. Ziolkowski, J.
Mol. Catal. 1984, 26, 355. Ϫ [10f] E. Mieczynska, A. M. Trzec-
[5]
[6]
[10g]
iak, J. J. Ziolkowski; J. Mol. Cat.; 1992, 73, 1. Ϫ
A. S. C.
Chan, C.-C. Pai, T.-K. Yang, S. M. Chen, Chem. Commun.
1995, 2031.
[11]
[7]
[8]
E. Billig, A. G. Abatjoglou, D. R. Bryant, US 4769498, 1988;
EP 0213639, 1987; EP 0214622, 1987; US 4748261, 1988.
L. A. van der Veen, P. C. J. Kamer, P. W. N. M. van Leeuwen,
Angew. Chem. 1999, 111, 349; Angew. Chem. Int. Ed. 1999,
38, 336.
K. G. Moloy, J. L. Petersen, J. Am. Chem. Soc. 1995, 117, 7696.
[12] [12a]
[12b]
B. Breit, Chem. Commun. 1996, 2071. Ϫ
Mol. Catal. 1999, 143, 143.
B. Breit, J.
[13]
A. M. Trzeciak, T. Glowiak, R. Grzybek, J. J. Ziolkowski, J.
Chem. Soc., Dalton Trans. 1997, 1831.
[9a] W. Mägerlein, M. Beller, A. F. Indolese, J. Mol. Catal. 2000,
156, 213. Ϫ [9b] M. Beller, M. Eckert, W. Moradi, H. Neumann,
Angew. Chem. 1999, 111, 1562; Angew. Chem. Int. Ed. 1999, 38,
[14] [14a] K. Issleib, A. Brack, Z. Anorg. Allg. Chem. 1957, 292, 245.
[9]
[14b]
Ϫ
F. Marschner, H. Kessel, H. Goetz, Phosphorus 1976,
6, 135.
1454. Ϫ [9c] M. Beller, W. A. Moradi, M. Eckert, H. Neumann,
[15] [15a] P. W. N. M. Leeuwen, C. F. Roobeck, J. Organomet. Chem.
[9d]
[15b]
Tetrahedron Lett. 1999, 40, 4523. Ϫ
M. Beller, Med. Res.
1983, 258, 343. Ϫ
A. van Roy, E. N. Orij, P. C. J. Kamer,
[9e]
Rev. 1999, 19, 357. Ϫ
M. Beller, M. Eckert, Angew. Chem.
P. W. N. M. Leeuwen, Organometallics 1995, 14, 34.
Received February 7, 2001
[O01056]
[9f]
2000, 112, 1027; Angew. Chem. Int. Ed. 2000, 39, 1010. Ϫ
[9g]
M. Beller, M. Eckert, W. Moradi, Synlett 1999, 108. Ϫ
C.
W. Kohlpaintner, M. Beller, J. Mol. Catal. 1997, 116, 259. Ϫ [9h]
Eur. J. Org. Chem. 2001, 3871Ϫ3877
3877