Synthesis of electron-withdrawing butane- and arene-sulfonylamino phosphines and use in rhodium-catalyzed hydroformylation

Article abstract of DOI:10.1039/b208089c

Reaction of RSO₂N(H)CH₂CH₂N(H)SO₂R [R = Bu (1), 4-nitrobenzene (7), 1-naphthalene (9a), 2-naphthalene (9b)] with PhPCl₂ or EtPCl₂ gives monodentate phosphorus compounds 2 and 3 (R = Bu, PhP and EtP), and 8 (R = 4-nitrobenzene, PhP), and with Ph₂PCl gives the corresponding bidentate phosphine ligands Ph₂PN(SO₂R)CH₂CH₂N(SO₂R)P Ph₂ [R = Bu (10), 4-nitrobenzene (11), 1- and 2-naphthalene (12a,b)]; similar reactions of N,N′-(1-butanesulfonyl)-2,2′-diaminobiphenyl (4) give monodentate 5 (PhP) and 6 (EtP) and N,N′-bis(diphenylphosphino)-N,N′-(1-butanesulfonyl)-2,2′-diami nobiphenyl (16). A monodentate analogue of 10 was also prepared, Ph₂PN(Et)SO₂Bu (14). Diphosphorus compounds with two butanesulfonylamino groups on phosphorus were also prepared from 1 and Cl₂P(CH₂)_nPCl₂ (n = 2, 4) to give 19 and 20. Details of the ¹³C NMR false AA′X systems are reported for 19 and 20. Rhodium-catalyzed hydroformylation reactions were run at 60 and 80 °C, at CO/H₂ pressures from 4-11 atm, and in THF, toluene, CH₂Cl₂, and dioxane. Results show that the highest ratios of linear (n) to branched (iso) aldehydes were obtained with arenesulfonamides (n:iso > 10) while the bidentate alkanesulfonamide 10 gave a lower n:iso ratio of 7.2 but the highest rate [k₁ = 1.98 h^-1, turnover frequency = 1130 mol aldehyde (mol Rh)^-1 h^-1] in THF at 80 °C. Both the rate and n:iso ratio for 10 were found to increase with decreasing CO/H₂ pressure in THF and in toluene, although the rate change was small for toluene. Both the rate and n:iso ratio for 10 also increased in CH₂Cl₂, but this was found not to be due to lower CO/H₂ concentrations in solution, on the basis of solubility measurements in THF and CH₂Cl₂.

Full text of DOI:10.1039/b208089c

View Article Online / Journal Homepage / Table of Contents for this issue
Synthesis of electron-withdrawing butane- and arene-sulfonylamino
phosphines and use in rhodium-catalyzed hydroformylation†
Matthew P. Magee, Huan-Qiu Li, Oma Morgan and William H. Hersh*
Department of Chemistry and Biochemistry, Queens College and the Graduate Center of the
City University of New York, Flushing, NY 11367-1597, USA.
E-mail: william_hersh@qc.edu
Received 19th August 2002, Accepted 12th November 2002
First published as an Advance Article on the web 20th December 2002
Reaction of RSO N(H)CH CH N(H)SO R [R = Bu (1), 4-nitrobenzene (7), 1-naphthalene (9a), 2-naphthalene (9b)]
2
2
2
2
with PhPCl or EtPCl gives monodentate phosphorus compounds 2 and 3 (R = Bu, PhP and EtP), and 8
2
2
(
R = 4-nitrobenzene, PhP), and with Ph PCl gives the corresponding bidentate phosphine ligands
2
Ph PN(SO R)CH CH N(SO R)PPh [R = Bu (10), 4-nitrobenzene (11), 1- and 2-naphthalene (12a,b)]; similar
2
2
2
2
2
2
reactions of N,NЈ-(1-butanesulfonyl)-2,2Ј-diaminobiphenyl (4) give monodentate 5 (PhP) and 6 (EtP) and N,NЈ-
bis(diphenylphosphino)-N,NЈ-(1-butanesulfonyl)-2,2Ј-diaminobiphenyl (16). A monodentate analogue of 10 was also
prepared, Ph PN(Et)SO Bu (14). Diphosphorus compounds with two butanesulfonylamino groups on phosphorus
2
2
13
were also prepared from 1 and Cl P(CH ) PCl (n = 2, 4) to give 19 and 20. Details of the C NMR false AAЈX
2
2
n
2
systems are reported for 19 and 20. Rhodium-catalyzed hydroformylation reactions were run at 60 and 80 ЊC,
at CO/H pressures from 4–11 atm, and in THF, toluene, CH Cl , and dioxane. Results show that the highest ratios
2
2
2
of linear (n) to branched (iso) aldehydes were obtained with arenesulfonamides (n:iso > 10) while the bidentate
Ϫ1
alkanesulfonamide 10 gave a lower n:iso ratio of 7.2 but the highest rate [k = 1.98 h , turnover frequency = 1130 mol
1
Ϫ1 Ϫ1
aldehyde (mol Rh) h ] in THF at 80 ЊC. Both the rate and n:iso ratio for 10 were found to increase with decreasing
CO/H pressure in THF and in toluene, although the rate change was small for toluene. Both the rate and n:iso ratio
2
for 10 also increased in CH Cl , but this was found not to be due to lower CO/H concentrations in solution, on the
2
2
2
basis of solubility measurements in THF and CH Cl .
2
2
Introduction
solve both the steric bulk and solubility problems, we sought to
use the butanesulfonyl group in place of the p-toluenesulfonyl
group. We hoped that the twin four-carbon chains would render
the bis-sulfonamides sufficiently lipophilic to be soluble in sol-
vents such as THF and toluene, but without giving difficult-to-
purify oils. We also hoped that the flexible chains would impart
less steric hindrance than the toluene sulfonamides. In this
paper we report the synthesis of a variety of butanesulfonyl-
amino phosphine ligands, a short series of ethane-linked
bis(diphenylphosphino) ligands, details of the NMR spectra
of the new compounds, and the results of hydroformylation
reactions with these new compounds including kinetics and
solvent and pressure dependencies.
We recently reported a new series of trivalent phosphorus
compounds that contain one or more phosphorus-N-sulfonyl
linkages, including the monodentate 1,3,2-diazaphospholidine
phenylphosphino compound TosL and the bidentate1
tosylamino-linked bis(diphenylphosphino) compound diTosL.
These phosphorus compounds were found to be remarkably
electron-withdrawing. For instance, TosL is comparable to
(
CF ) P, and diTosL to (C F ) PCH CH P(C F ) , in donor and
3 3 6 5 2 2 2 6 5 2
acceptor ability towards tungsten carbonyl complexes. Since the
hydroformylation reaction is well known to be promoted more
efficiently by electron-deficient phosphines than by electron-
2–4
rich phosphines,
we examined rhodium-catalyzed hydro-5
formylation in the presence of these and related phosphines.
The chelating diTosL ligand was found to be an effective pro-
moter, while the non-chelating TosL was not. Attempts to
synthesize analogues of diTosL with different bite-angles, in
order to use Casey’s and van Leeuwen’s findings that increased
6,7
bite-angle would give rise to better yields of linear aldehyde,
failed due to the unexpectedly high degree of steric hindrance
of the arenesulfonylamino group. Attempts to synthesize chel-
ating analogues of TosL, in order to determine whether two
sulfonylamino moieties per phosphorus would be better than
one, were successful, but the resulting compounds were so polar
that they are insoluble in non-chlorinated hydroformylation
solvents. While we were able to use CH Cl successfully as a
hydroformylation solvent, catalyst decomposition was acceler-
ated under some conditions in this solvent, so discovery of a
more soluble ligand remained a desirable goal. In order to try to
Results
Synthesis of monophosphorus compounds
Synthesis of the simplest compound desired, N,NЈ-dibutane-
sulfonyl-1,2-diaminoethane (1) was accomplished by reaction
of BuSO Cl and 1,2-diaminoethane in pyridine as shown in
Scheme 1. This procedure allowed the product to be obtained as
a white crystalline solid in low but adequate yield, and without
chromatography. Initial attempts had been carried out by
analogy to a literature procedure for reaction of BuSO Cl with
trans-1,2-diaminocyclohexane in methylene chloride, but this
procedure gave difficult-to-purify dark solids.
Reaction of 1 with PhPCl or EtPCl (Scheme 1) gave the
2
2
2
†
Dedicated to the memory of Mohammad Salman Hamdani. Sal
worked in our labs at Queens College as an undergraduate student from
998–2001, and died as part of the rescue effort at the World Trade
2
8
1
Center in New York City, September 11, 2001.
2
2
Electronic supplementary information (ESI) available: details of the
syntheses, purification, and NMR and analytical data. See http://
www.rsc.org/suppdata/dt/b2/b208089c/
required 1,3,2-diazaphospholidines 2 and 3 in good yield. These
compounds were two of only three in this study that turned out
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 3
D a l t o n T r a n s . , 2 0 0 3 , 3 8 7 – 3 9 4
387

View Article Online
insoluble material filtered from hot ethanol; the soluble
impurity was most likely N,N,NЈ-tris(4-nitrobenzenesulfonyl)-
1
1
,2-diaminoethane on the basis of H NMR. Compound 8 is
unusual in that it was recrystallized from THF/hexane, and is
insoluble in chloroform.
Synthesis of bis(diphenylphosphino) compounds
The 1,2-diaminoethane derivatives were prepared as previously
described for diTosL, by reaction of the bis-sulfonamides with 2
equivalents of Ph PCl in the presence of Et N in THF (Scheme
2
3
2
). Compounds 11 and 12a,b were somewhat insoluble in THF,
Scheme 1
not to be crystalline solids, but the viscous oils were nevertheless
easily isolated in analytically pure form. The structure of the
ethyl-substituted heterocycle 3 was confirmed by 2D-COSY
and HETCOR, which allowed all of the alkyl peaks to be
identified unambiguously. Compound 3 was found not to be
thermally stable over a period of weeks, decomposing at room
temperature in the glovebox to a white solid that was found by
1
31
H and P NMR to contain a number of uncharacterized
products.
Reaction of 2,2Ј-diamino-1,1Ј-biphenyl with BuSO Cl in
2
pyridine gave the required bis-sulfonamide 4 in high yield
Scheme 2
(
Scheme 1). Reaction of 4 with PhPCl and EtPCl as before
2
2
gave the desired 7-membered ring 1,3,2-diazaphosphepine
heterocycles 5 and 6. The NMR spectra of 5 and 6 are much
more complex than those of 4, because while 4 has a C2
symmetry axis, and so the phenyl rings and butyl groups are
CH Cl , and CHCl , and were isolated by filtration of the THF
2
2
3
reaction mixtures and washing of the resultant solids with
ϩ
Ϫ
CH Cl2 or CHCl3 to remove the Et NH Cl by-product.
2
3
equivalent, the C symmetry of 5 and 6 is broken by the phenyl
or ethyl group on phosphorus, which lies on one side of the
Crystallization of these compounds was difficult, yields were
not optimized, and while they are spectroscopically pure, only
12b was eventually obtained analytically pure.
One monodentate analogue of the bis(diphenylphosphino)
compounds in Scheme 2 was also prepared in order to examine
the effect of chelation. As shown, the monodentate analogue of
2
7
-membered ring. Each of the methylene carbon atoms of the
13
two butyl groups exhibits distinct peaks in the C NMR
spectra as do the methyl groups of 6. No peak broadening was
observed in 5 and 6, and since all of the biphenyl carbon atoms
have different chemical shifts as do most of the two butanesul-
fonyl carbon atoms, there is no evidence of any rotation about
the biphenyl carbon–carbon bond or for inversion at phos-
phorus. On the basis of the separation of the two closest peaks
10 was prepared from Ph PCl and 13, and while the synthesis
2
and characterization were straightforward, the product 14, like
2 and 3, is an oil.
The reaction of 4 with Ph PCl in the presence of Et N, in a
2
3
1
3
(
∆ν = 4.7 Hz for the butanesulfonyl CH groups of 6 at 298 K)
manner analogous to the previous syntheses, gave only the
monophosphorus compound 15 (Scheme 2). Identification of
this compound was not straightforward, due to the presence
of atropisomers with respect to restricted rotation about the
aryl–nitrogen bonds and/or stereocenters at nitrogen due to
slow inversion at nitrogen, giving rise to a mixture of diastereo-
3
Ϫ1
and simulation of the NMR spectrum, giving kexch < 5 s , the
barrier to rotation about the biphenyl carbon–carbon bond
Ϫ1
must be >16.5 kcal mol at room temperature. In comparison,
a
1,3,2-dioxaphosphepine, lacking the branching SO Bu
2
moieties of 5 and 6, exhibited a barrier to rotation of ca. 10 kcal
Ϫ1 9
10,11
1
mol .
meric products.
Nevertheless, integration of the H NMR
1 31
One new bis(N-arenesulfonyl)-1,3,2-diazaphospholidine was
prepared for this study, using the 4-nitrobenzene group as an
electron-withdrawing analogue of the p-toluene group. Low but
adequate yields of the bis-sulfonamide 7 and the desired
heterocycle 8 were prepared using methods analogous to those
used previously (Scheme 1), although the low solubility made
the purifications difficult. For instance, 7 was isolated as the
spectrum and the fact that the same H and P NMR spectra
were generated by adding either one or two equivalents of
Ph PCl to 4 made it clear that only the monophosphorus
2
compound formed. This result was not unexpected on the basis
5
of previous results with the bis(toluenesulfonyl) analogue of 4.
Addition of 2 equiv. of BuLi to 4 gave what we presume to be
the dianion, and reaction at Ϫ35 ЊC with Ph PCl gave a new
2
3
88
D a l t o n T r a n s . , 2 0 0 3 , 3 8 7 – 3 9 4

View Article Online
1
13
product that could be isolated as a white solid. The H, C, and
31
P NMR spectra of this material were similar in complexity to
those of 15, again due to the presence of atropisomers/stereo-
centers at both nitrogen atoms. However, the integration of the
1
H NMR spectrum is consistent with it being the diphosphorus
compound 16, and the elemental analysis while not perfect is
consistent with this formulation as well. However, 16 appar-
ently is not very stable under many conditions and this may
account for why it was not obtained analytically pure; for
instance, 1 h of reaction between the dianion and Ph PCl
2
appears optimal, since at longer reaction times increasing
amounts of 15 were observed. In CD CN and CDCl , decom-
3
3
position of 16 to both 15 and 4 occurred, although 16 appeared
to be stable in C D and toluene-d . The variable temperature
6
6
8
NMR spectra of 15 and 16 will be described separately, since
the results are interesting but of limited relevance to the hydro-
formylation results.
Synthesis of chelating TosL analogues
One of the principal goals of this study was the synthesis of
chelating butanesulfonyl analogues of TosL. This was accom-
plished as shown in Scheme 3, by combining tetrachlorodiphos-
1
3
Fig. 1 Partial C NMR spectra of 19 and 20, exhibiting AAЈX
spectra for (a), (d), and (e), and ABX spectra for (b) and (c). 19: (a)
heterocycle ring CH , (b) bridging CH ; 20: (c) heterocycle ring CH , (d)
2
2
2
C of bridging carbon chain, (e) C of bridging carbon chain. Spectrum
1
2
(
c) is plotted at twice the horizontal expansion as the others, and the
vertical expansions are all different.
0
(
Hz, on the other hand [Fig. 1(d)], while the multiplet for C2
consisting of a doublet with a broad ‘triplet’ of slightly lesser
height at the center) is best fit for a small non-zero value of ∆ν
Fig. 1(e)].
[
Hydroformylation
Reactions were run according to the conditions in eqn. (1),
Scheme 3
phines 17 and 18 with bis-sulfonamide 1 in the presence of
Et N. The resultant compounds (19, 20) were quite soluble in
3
(
1)
THF and toluene, and were readily crystallized in analytically
pure form. For both 19 and 20, assignments of all alkyl peaks in
1
13
the H and C NMR spectra were confirmed by 2D-COSY and
HETCOR.
13
The C NMR spectra exhibit interesting effects due to
and data are collected in Table 1. For most of the reactions,
time points were taken by removing a reaction sample with a
gas-tight syringe, and the percentage yield of aldehyde vs. time
virtual coupling of the phosphorus atoms and the presence of
12
13
false AAЈX spin systems, in which the presence of a single
C
nucleus splits the symmetry of the molecule and gives rise to
31
was plotted. The turnover frequency (TOF) in units of mol
different P chemical shifts, and hence to an ABX system (with
Ϫ1 Ϫ1
13
aldehyde (mol Rh)
h
was calculated for the initial portion of
the C being the X nucleus). These effects have also been seen
with the p-toluenesulfonyl analogues of 19 and 20, 21 and 22
the reaction following any induction period, if present, as has
5,15
5
been described previously.
Most of the reactions exhibited
respectively, and for completeness we note there are other
13,14
12
13
first-order kinetics in 1-hexene consumption and aldehyde
formation, and for these the first-order rate constant was also
calculated; several examples of zero-order kinetics were also
seen and these are noted in Table 1. Representative zero-order
and first-order plots are shown in Fig. 2.
examples
not previously cited. Partial C NMR spectra
are shown for 19 and 20 in Fig. 1. Both 19 and its p-toluene-
sulfonyl analogue 21 exhibit six-line ABX multiplets for the
carbon atom adjacent to phosphorus; for 19, the spectra are
3
best fit with J = 17 Hz and a shift difference of the phos-
PPЈ
13
phorus atoms for the C -isotopomer of 5.3 Hz (0.033 ppm) on
1
Solubility of CO
1
a 400 MHz ( H) NMR spectrometer [Fig. 1(b)]. The ‘triplet’ for
the ring CH arises due to virtual coupling to the distant phos-
Because high n:iso ratios were seen at low CO/H pressure when
both THF and toluene were used as the reaction solvent, and in
2
2
5
3
phorus atom; that is, J = 0 Hz but the previously fit J = 17
PC
PPЈ
Hz gives rise to the central line in the multiplet [Fig. 1(a)]. The
CH Cl solvent at the same CO/H pressure as for THF and
2
2
2
spectrum is sufficiently broad that little change is seen for values
toluene, gas solubility in different solvents was investigated as a
of ∆νPPЈ between 0 and 2 Hz. In contrast, for 20 and 22 the
possible cause of this observation. We focussed on CO since the
5
4
much lower values of J give rise to a doublet for the ring
rate dependence on [H ] is considered to be small. Literature
PPЈ
2
CH of 22, but in 20 a small pair of peaks in the middle of the
doublet arises because in this case ∆ν ≠ 0 Hz [Fig. 1(c)]. The
data for CO solubility are available for toluene, diethyl ether,
2
16,17
and dioxane,
but not for THF and CH Cl . A method for
2 2
doublet for C of the 4-carbon bridge of 20 is best fit with ∆ν =
prediction of gas solubility using a functional group contribu-
1
D a l t o n T r a n s . , 2 0 0 3 , 3 8 7 – 3 9 4
389

View Article Online
a
Table 1 Results of the hydroformylation of 1-hexene
Aldehyde
yield (%)
2-Hexene
yield (%)
b
Ϫ1
c
Entry
L (equiv.)
Solvent
Time/h
n:iso
k1/h
TOF
1
2
3
4
5
6
7
8
9
9
2 (10)
THF
THF
THF
THF
THF
THF
THF
THF
THF
THF
THF
THF
THF
THF
CH Cl
6
4
6
6
6
4
6
6
4
4
5
6
6
2
4
5
2.6
1.7
3.5
3.0
2.8
4.3
4.4
4.4
5.1
5.0
4.3
6.6
3.3
7.2
11.1
5.0
43
89
1
19
10
1
10
27
8
7
5
6
7
5
15
3
11
8
6
0.130 ± 0.001
∼Zero-order
87
270
2
630
42
345
360
330
≥230
≥215
270
630
180
3 (10)
5 (10)
6 (10)
74
19
68
92
88
91
86
79
85
82
84
83
89
1.6 ± 0.2
8 (10)
0.110 ± 0.004
0.48 ± 0.03
0.47 ± 0.03
0.47 ± 0.01
Not determined
Not determined
0.39 ± 0.02
0.93 ± 0.03
0.30 ± 0.01
1.98 ± 0.07
0.83 ± 0.05
10 (1)
10 (2)
10 (10)
d
10 (10)
d
10 (10)
10
11
12
13
14
15
10 (50)
10 (10), 3.9 atm
10 (10), 11.9 atm
10 (10), 80 ЊC
10 (10)
1130
530
350
2
2
10 (10)
9:1
0.52 ± 0.01
THF:CH Cl
2
2
1
1
1
1
2
2
2
6
7
8
9
0
1
2
10 (10), 3.7 atm
10 (10)
10 (10), 10.5 atm
10 (10), 80 ЊC
10 (10)
11 (2)
12a (10)
Toluene
Toluene
Toluene
Toluene
Dioxane
THF
6
6
6
3.1
6.2
4
8.2
4.7
4.0
5.7
5.7
10.8
4.4
95
90
93
93
93
50
75
5
3
2
7
5
8
4
0.481 ± 0.007
0.435 ± 0.006
0.400 ± 0.007
0.96 ± 0.07
0.46 ± 0.03
∼Zero-order
420
300
290
850
450
140
180
THF
6
∼Zero-order
0.40 ± 0.03)
e
(
2
3
12b (10)
THF
6
12.1
83
8
∼Zero order
290
e
(
0.57 ± 0.05)
24
25
26
27
28
29
30
31
dppb
THF
THF
THF
THF
THF
CH Cl
6
3.2
5
4
6
6
5.5
4
3.6
2.6
3.4
—
1.9
2.0
10.1
17.3
18
89
39
0.8
78
65
85
82
0
8
9
0.040 ± 0.001
Zero-order
∼Zero-order
34
310
60
14 (10)
16 (2)
19 (2)
0
2
20 (2)
16
13
9
Not determined
Not determined
180
200
440
280
20 (2)
2
2
2
f
diTosL (10)
diTosL (10)
THF
0.86 ± 0.05f
CH Cl
11
0.54 ± 0.05
2
a
For all reactions except as noted, catalyst precursor [Rh(CO) (acac)] = 0.001 M, [alkene] = 1.0 M, alkene:Rh = 1000:1, temperature = 60 ЊC, and CO/
2
b
c
Ϫ1 Ϫ1
H pressure = 6.7 atm. n:iso = ratio of linear aldehyde (heptanal) to 2-methylhexanal. TOF = turnover frequency = mol aldehyde (mol Rh)
value reported is for the initial portion of the reaction following any induction period if present (see text). Successive aliquots of 1000 equiv. of
h ;
2
d
e
1
-hexene were used and the reactions were only monitored at 4 h each. A first-order rate constant can be approximated, but the kinetic order is
f
closer to zero; see text. Rate constant not previously reported for this reaction from ref. 5.
are much higher than that of toluene and have the effect of
increasing the error, we simply report our results without
correcting for the likely systematic error. Results are collected in
Table 2 for both literature data, calculations using the group
contribution method, and our experimental data, and molar
concentrations are tabulated to allow comparison with the
other reactants. Data for H are also included. The experi-
2
mental and calculated values for solubility of CO in THF were
found to be in reasonable agreement; that is, THF evidently
bears more similarity to diethyl ether than to dioxane as a
solvent for CO; however the calculated solubilities are poor for
H in both dioxane and THF. The CO solubilities for THF and
2
CH Cl are similar, and while the precise values are likely to be
2
2
accurate only to within 10%, they nevertheless make clear that
CO solubilities in these solvents are roughly twice that in
Fig. 2 First-order plots for representative reactions of 10 from Table
(left hand axis), and zero-order plot for reaction of 14 (right hand
axis). All runs plotted are 10 equiv. of ligand.
1
toluene. The H solubilities are all lower but in units of molarity
2
do not differ very much.
18
tion method has been described, but while it gives accurate
Discussion
solubilities for both CO and H in diethyl ether, it fails for
2
dioxane, which was presumably omitted from the diethyl ether
and alcohol parameterization that was carried out. A predic-
tion for THF using this method, therefore, was considered to be
unreliable. In addition, the method was not parameterized for
haloalkanes. We therefore measured CO solubilities in a man-
Four classes of ligand are evident, both in the results above and
those that we have recently described, namely: (1) inhibitors,
5
(2) non-chelating ligands that give low n:iso ratios and which
are poor promoters, (3) non-chelating ligands that give low
n:iso ratios but which are good promoters, and (4) chelating
ligands that give high n:iso ratios. The inhibitors are classified
with respect to the ‘blank’ results obtained in the absence of
any ligand, a 5% aldehyde yield and more importantly a 46%
19,20
ner based on that described in the literature,
using a simple
21
vacuum line/manometer set-up. For N and CO in toluene, we
2
reproducibly obtained values that were 7% higher than
16
5
reported. Because the vapor pressures of THF and CH Cl2
yield of isomerization. The other classes of ligands can be
2
3
90
D a l t o n T r a n s . , 2 0 0 3 , 3 8 7 – 3 9 4

View Article Online
Table 2 Gas solubility data
a
a, b
a
a, b
X (CO)
2
Calc. X (CO)
[CO]/M (3.35 atm)
X (H )
Calc. X (H )
[H ]/M (3.35 atm)
2
2
2
2
2
2
c
d
Toluene
THF
8.02
8.02
20.4
—
21.4
16.6
0.025
0.074
0.080
0.019
3.15d
2.70
4.06
3.17
4.89
—
4.15
6.33
0.0099
0.011
0.013
17.9
e
c
c
f
CH Cl2
15.4
4.9
2
d
Dioxane
Diethyl ether
1.84g
0.0072
17.0
6.24
a
4
b
c
d
e
Mol fraction × 10 , 298.15 K, 1 atm partial pressure. Calculated according to ref. 18. From ref. 17. From ref. 31 Measured at 5 ЊC to minimize
4
f
g
errors due to vapor pressure; at 20 ЊC, X × 10 = 35 ± 11. For 1,1,2,2-tetrachloroethane, from ref. 32. From ref. 32.
2
compared with PPh , which is used industrially, as a point of
phosphino)arenesulfonylamino ligands, 11 and 12b, also gave
3
reference, and gave under our conditions modest n:iso and TOF
values of 3.1 and 250, respectively, but a 99% aldehyde yield
with only 1% isomerization to 2-hexene.
high n:iso ratios, while the 1-naphthalenesulfonylamino ligand
12a gave a lower n:iso value that may be due to increased steric
hindrance. The bis(diphenylphosphino)butanesulfonylamino
ligand 10 gave somewhat different results, however. At 60 ЊC in
THF, n:iso ratios from 4.3 to 5.1 were observed, intermediate
between the chelating values of ca. 10 for the arene-
sulfonylamino ligands and ca. 2 for the non-chelating ligands.
The non-chelating analogue of 10, Ph PN(Et)SO Bu (14), gave
5
Two compounds in this study are inhibitors, namely, the
biphenyl heterocycle 5 and the 2-carbon bridged chelating
analogue of TosL, 19. Previously, we found that the EtP ana-
5
logue of TosL, 23, is also an inhibitor. All of these compounds
give virtually no aldehyde and no isomerization of 1-hexene.
Therefore, they are not simply innocent non-binding ligands,
but must bind to rhodium and prevent any further reaction.
While each of these inhibitors has the bis(aminosulfonyl) func-
tionality of TosL, several other bis(aminosulfonyl) compounds
are not inhibitors. Of these, TosL, its close relative 8 (with arene-
sulfonyl and PhP moieties), the arenesulfonylamino chelating
analogues 21 and 22, and the butanesulfonylamino PhP com-
pound 2 are poor hydroformylation promoters, giving TOF
values from 37 to 120, and more importantly comparable
amounts of hexene isomerization and aldehyde formation.
However, the bis(butanesulfonylamino) alkylphosphines 3 and
2
2
a lower n:iso ratio of 2.6, and so once again it is reasonable to
propose that the value seen for 10 is characteristic of chelation.
The exception to the above observations is biphenyl-linked
16, which as noted is difficult to synthesize and is presumably
quite sterically hindered; we propose that it does not chelate. All
of the other bis(diphenylphosphino) compounds are relatively
unhindered, and so are presumed to chelate. However, the high
n:iso ratios evidently require not just a four-atom bridge
between the phosphorus atoms, but also the N-sulfonyl group,
since dppb (entry 24, Table 1) is clearly a poor hydroformyl-
ation promoter, as are previously described analogues with
5,22
2
0 are active hydroformylation promoters. Interestingly, 6 is the
N-alkyl groups.
most active ligand we have found but the catalyst apparently
reproducibly dies within 2 h of reaction, as evidenced by the
absence of any further aldehyde formation or isomerization
that accompanies the formation of a dark orange color of the
reaction solution. Only 3, 6, and 20 have both an alkyl group on
phosphorus and the alkanesulfonyl group on nitrogen, while all
of the others except 19 have either an aromatic group on phos-
phorus and/or an arenesulfonyl group on nitrogen. At this
stage, we have no single proposal that reasonably accounts for
these facts. We can propose that 5 is so bulky that coordination
of a single ligand to rhodium kills the catalytic activity, and that
Ligand 10 gives a less active catalyst at 60 ЊC than diTosL –
for instance at a 10:1 ligand:Rh ratio, TOF and k values for 10
1
Ϫ1
and diTosL were 330 and 440, and 0.47 and 0.86 h (entries 8
5
and 30, Table 1), respectively. The only other PPh ligand to
2
which these can be compared are their non-chelating analogues
14 and Ph PN(Et)Ts. While 10 and diTosL give reactions that
2
are first-order in 1-hexene, 14 and Ph PN(Et)Ts give zero-order
2
reactions (Fig. 2), with TOF = 170–230 mol aldehyde (mol
Ϫ1 Ϫ1
5
Rh)
h
for Ph PN(Et)Ts (for L:Rh = 5 and 26) and TOF =
2
Ϫ1 Ϫ1
310 mol aldehyde (mol Rh)
h
for 14 (L:Rh = 10). Hence, the
butanesulfonyl analogue gives a higher rate when monodentate,
and a lower rate when chelating, than the toluenesulfonyl
analogue. One could propose that a second molecule of 10
competes with 1-hexene for coordination to rhodium, but since
the same rates were seen at three different concentrations of 10,
the better explanation is that 1-hexene competes with CO for
coordination to rhodium. The strong inverse dependence of
rate on [CO] in THF is consistent with this, although the much
weaker dependence in toluene is puzzling but could be related
to the greater than 2-fold lower concentration of CO in toluene
than in THF.
2
3 binds and 19 chelates too tightly to allow catalysis, but then
why is 3 a good catalyst rather than an inhibitor like 23, and
why are 21 and 22 just poor catalysts but 20 a good one and 19
an inhibitor? A correlation with high activity does seem to exist,
even if we cannot suggest a reason why it exists: four com-
pounds, namely, 3, 6, 19, and 20, are the only alkanesulfonyl-
amino/alkyl phosphines, and these comprise the three good
promoters and paradoxically one inhibitor (19). One might
suppose that all four compounds bind tightly via one phos-
phorus atom; for instance 3 and 20 gave nearly identically low
n:iso ratios (1.7 and 1.9, respectively), and we propose this
means that both are non-chelating and so the 4-carbon bridging
chain of 20 serves just to mimic the ethyl moiety of 3. Com-
pound 19, then, might simply bind tightly via one phosphorus
atom while the other serves as a bulky blocking group by virtue
of the short 2-carbon bridge. The EtP compounds 3 and 6 are
both more reactive than their PhP analogues 2 and 5, so further
testing of this simple correlation may be warranted; what is
somewhat mysterious is why 2 is just a poor promoter but 5 is
an inhibitor.
The difference in reaction order as well as the difference in
n:iso ratio suggests that the active catalyst from 14 or Ph P-
2
N(Et)Ts would contain the LRh(CO)3 unit [L = 14, Ph P-
2
N(Et)Ts], which readily dissociates CO giving zero-order
kinetics; L Rh might not form for steric reasons or is simply not
2
catalytic. Typically, electron donor ligands like PPh give first-
3
2
3
order kinetics, and electron-poor phosphite ligands give
23,24
zero-order kinetics,
so 14 and Ph PN(Et)Ts behave like
2
monodentate phosphites. On the other hand, the bulky chelat-
ing diphosphite 24 has not only been reported to give first-order
The bis(diphenylphosphino)sulfonylamino ligands generally
give significantly higher n:iso ratios than the above ligands, and
we have previously proposed that this is reasonably accounted
kinetics, but also gives a rate that is relatively insensitive to [H ]
and is inversely proportional to [CO]. Evidently, chelating 10
and diTosL behave like 24 and the active catalyst presumably
2
25
5
for by chelation of these ligands. That is, for instance, diTosL
contains the LRh(CO) unit (L = 10, diTosL).
2
gave an n:iso ratio of 10, while its non-chelating analogue
Like diTosL, 10 exhibits the same unusual CH Cl2 and
2
Ph PN(Et)Ts gave a ratio of 2.7. Two of the other bis(diphenyl-
thermal effects on the n:iso ratio. For instance, diTosL in
2
D a l t o n T r a n s . , 2 0 0 3 , 3 8 7 – 3 9 4
391

View Article Online
difficult to measure gas solubility due to its high vapor pressure.
Nonetheless, even though the CO solubility was measured at
lower temperature (20 and 5 ЊC), the actual solubility was com-
parable to that in THF at 25 ЊC. We also checked whether or
not the vapor pressure of CH Cl could be contributing to the
2
2
total measured pressure, and so bring down the actual partial
pressure of the CO/H , but under our reaction conditions
2
(
where the solvent is heated but not the total apparatus), the
vapor pressure of the CH Cl is negligible at 60 ЊC. Hence, the
2
2
higher rate and n:iso ratio must depend on some curious solvent
effect of the CH Cl on the reaction; it is not due to lowered gas
2
2
solubility.
Conclusion
This study completes an initial phase of screening of N-sulfonyl-
amino phosphine ligands for the hydroformylation reaction. We
have shown that these ligands can be active promoters, as well
as, paradoxically, active inhibitors of the reaction. Seemingly
minute changes in the structure of the ligands give rise to
a switch from active promoter to active inhibitor. The bis-
(sulfonylamino) phosphine ligands have not exhibited any
favorable qualities for rate or selectivity with the striking excep-
tion of the biphenyl 7-membered ring compound 6. Since this
ligand gives an exceptionally active catalyst before apparent
decomposition occurs, it is possible that the problem with this
class is rapid catalyst decomposition; future work will address
this issue. The sulfonylamino diphenylphosphine ligands
exhibit good rates for non-chelating ligands, and good rates
and selectivity for chelating ligands. Compound 10 is the
most active promoter we have discovered, giving at 80 ЊC a
CH Cl resulted in an increase of the n:iso ratio to 17.3 from
2
2
1
0.1 in THF at 60 ЊC (entry 31, Table 1), although the TOF
5
declined. For 10, the n:iso ratio increased from ca. 5 to 11, and
the TOF increased from 330 to 530. Since it would be reason-
able to suppose that the effect was due to a reaction of the
catalyst with CH Cl , a run was carried out in which the THF
2
2
solvent contained 10% CH Cl by volume. However, both the
2
2
n:iso and TOF values were the same as those observed in pure
THF. Finally, a run was carried out using the chelating TosL
analogue 20 in CH Cl (entry 29, Table 1), but it gave n:iso and
2
2
Ϫ1 Ϫ1
TOF values that were virtually the same as in THF.
turnover frequency of 1130 mol aldehyde (mol Rh) h , with
A somewhat smaller thermal effect on the n:iso ratio was
previously observed as well. At 80 ЊC diTosL was found to give
both higher n:iso and TOF values, increasing to n:iso = 15.8
and TOF = 760. For ligand 10, a similar effect was seen for
n:iso, increasing from ca. 5 to 7.2, but the TOF value increased
a linear:branched ratio of 7.2. A curious solvent effect has been
discovered for CH Cl which was found not to be due to solubil-
2
2
ity effects or direct reaction of the catalyst with the solvent, so
examination of a broader range of solvents for hydroformyl-
ation may be fruitful. Future work on these compounds will
combine a mechanistic approach involving the synthesis of
rhodium complexes of these ligands and examination of the
hydroformylation reactions by NMR, with the empirical
approach of this study which will continue with a search for
new linkers between the nitrogen atoms and a combinatorial
approach to varying the groups on nitrogen and phosphorus,
to elucidate features that contribute to both inhibition and
promotion of hydroformylation.
5
by more than a factor of 3, to 1130, and the rate by a factor of
Ϫ1
4
, to 1.98 h . In toluene, both effects are muted, with n:iso
increasing from 4.7 at 60 ЊC to 5.7 at 80 ЊC, and the TOF and
Ϫ1
rate constants increasing from 300 and 0.44 h at 60 ЊC to 850
Ϫ1
and 0.96 h at 80 ЊC. In comparison to other ligands, which we
5
have described in some detail, we note that there are no
previous reports of the CH Cl effect, and other ligands give
2
2
rise to a decrease in n:iso ratio as the temperature is
7,26
increased. The TOF value for 10 at 80 ЊC is still lower than
24
27
that for phosphite 25 and bis-amide 26 but the n:iso ratio for
Experimental
1
0 is much higher; bulky chelating phosphite 24 gave higher
25,28
TOF and n:iso ratios, but also more isomerization.
All manipulations of air-sensitive compounds were carried out
either in a Vacuum Atmospheres inert atmosphere glovebox
under recirculating nitrogen, or by using standard Schlenk
A further curious fact is that for both 10 and 24, which
exhibit similar kinetics, the n:iso ratio is inversely proportional
to CO pressure. We therefore wondered whether any of the
CH Cl and thermal effects could be explained by CO solubility
1
13
31
techniques. H, C, and P NMR spectra were recorded on a
Bruker DPX-400 spectrometer; chemical shifts are reported
relative to TMS or residual hydrogens in CDCl (δ 7.24), C D
2
2
in particular (little effect is seen for [H ] with 24). Under the
2
3
6
6
1
low pressure and temperature conditions used, gas solubility
is proportional to gas pressure (Henry’s Law), but obviously
the propotionality constant will change with temperature and
solvent. Gas solubility is expected to increase at higher temper-
(δ 7.15), CD Cl (δ 5.32), or DMSO-d (δ 2.49) for H NMR, to
2
2
6
C D at 128.0 ppm, CDCl at 77.0 ppm, CD Cl at 53.8 ppm, or
6
6
3
2
2
13
DMSO-d at 39.5 ppm for C NMR, and to external 85%
H PO at 0 ppm (positive values downfield) for P NMR.
3
6
31
4
16–18
ature,
however, so this would lead to a decrease in n:iso ratio
Elemental analyses were performed by Desert Analytics,
Tucson, AZ. NMR line-shape analyses were carried out using
gNMR (Cherwell Scientific Publishing, Inc.) on a Macintosh
computer.
All solvents were treated under nitrogen. Benzene, diethyl
ether, and tetrahydrofuran were distilled from sodium benzo-
phenone ketyl. Hexane was purified by washing successively
with 5% nitric acid in sulfuric acid, water, sodium bicarbonate
solution, and water, and then dried over calcium chloride and
distilled from n-butyllithium in hexane. Methylene chloride was
distilled from phosphorus pentoxide. Pyridine was dried over
potassium hydroxide pellets and distilled from BaO. Triethyl-
at high temperature (with constant pressure), in contrast to
what is observed. Since CO is more soluble in THF than in
toluene, one might again expect lower n:iso ratios in THF, but
again that did not occur. Since literature data were available for
dioxane, that was tested; diethyl ether could not be used due to
the insolubility of 10. Clearly here, gas solubility does not play
a role: CO is less soluble in dioxane than in toluene and in THF,
yet the rate constant is the same as in THF and toluene, while
the n:iso ratio is only somewhat higher. Since the largest rate
and n:iso ratio changes were seen for CH Cl , this provided the
2
2
best test case, but it was also the solvent in which it was most
3
92
D a l t o n T r a n s . , 2 0 0 3 , 3 8 7 – 3 9 4

View Article Online
amine and 1,2-diaminoethane were distilled under N from
in 3 mL of THF was then added dropwise with stirring, with
2
CaH . NMR solvents were treated as follows: CDCl3 and
immediate formation of a cloudy white mixture. The mixture
2
CD Cl were vacuum-transferred from phosphorus pentoxide,
was then stirred at room temperature for 1.5 h and then filtered
2
2
and C D was vacuum-transferred from sodium benzophenone
to remove Et NHCl. Solvent removal in vacuo gave a viscous
6
6
3
ketyl. The 1-hexene used for hydroformylation was passed
through a column of basic alumina, stirred over sodium,
yellow oil which was dissolved in 6 mL of CH Cl and passed
2
2
through a ca. 7 mL pad of silica gel packed in CH Cl on a 15
2
2
vacuum-transferred, and stored under N in the glovebox.
mL sintered glass frit. The product was eluted with a further 40
mL CH Cl , and the solvent removed under vacuum to give 0.6
2
The following chemicals were used as received: Cl PCH -
2
2
2
2
1
CH PCl2 (Strem Chemicals), Ph PCl and PhPCl2 (Aldrich),
g (60% yield) of product as a clear oil. H NMR (CD Cl ): δ
2
2
2 2
and butanesulfonyl chloride, 1-naphthalenesulfonyl chloride,
-naphthalenesulfonyl chloride, 4-nitrobenzenesulfonyl
chloride, and EtPCl (ACROS). The following compounds were
7.61 (m, 2H, PPh), 7.46 (m, 3H, PPh), 3.79 [m, 2H, C(H )H -
a
b
2
C(H )H ], 3.59 [m, 2H, C(H )H C(H )H ], 3.17 (AAЈBBЈ, J
≈
a
b
a
b
a
b
HH
8.0 Hz, 4H, SO CH ), 1.82 (∼quintet, J ≈ 7.8 Hz, 4H,
2
2
2
HH
5,29
30
prepared as previously described: 18
and Rh(CO) (acac).
SO CH CH ), 1.46 (sextet, J
= 7.4 Hz, 4H, SO CH -
HH 2 2
31
2
2
2
2
Hydroformylation reactions were carried out as previously
described in a 90 mL Fisher–Porter vessel (Andrews Glass Co.)
attached to an Andrews Glass Co. multi-ported stirring
CH CH ), 0.94 (t, J = 7.3 Hz, 6H, SO CH CH CH CH ). P
2 2 HH 2 2 2 2 3
13
NMR (CD Cl ): δ 91.27 ppm. C NMR (CD Cl ): δ 138.82
2
2
2
2
1
2
(d, J = 30.6 Hz), 131.03 (s), 129.60 (d, J = 21.1 Hz), 129.24
PC
PC
5
3
2
assembly.
(d, J = 5.6 Hz), 52.96 (SO CH ), 49.09 (d, J = 5.9 Hz, ring
PC 2 2 PC
CH ), 25.75 (SO CH CH ), 21.90 (SO CH CH CH ), 13.68
2
2
2
2
2
2
2
2
(
4
SO CH CH CH CH ). Anal. calc. for C H N O S P: C,
7.28; H, 6.70; N, 6.89. Found: C, 47.36; H, 6.65; N, 6.81%.
Syntheses
2 2 2 2 3 16 27 2 4 2
Sulfonamides. The butanesulfonyl compounds 1, 4, and 13,
were prepared by addition of n-butanesulfonyl chloride to a
pyridine solution of the amine. The 4-nitrobenzenesulfonyl
compound 7 was prepared by addition of 4-nitrobenzene-
sulfonyl chloride to the amine in aqueous NaOH solution, and
the naphthalenesulfonyl compounds 9a,b were prepared in
Diphenylphosphino compounds. The diphosphine compounds
0, 11, and 12a,b, and the monophosphine compounds 14 and
1
1
5 were prepared by reaction of Ph PCl with the appropriate
2
sulfonamide and Et N in THF. Details of the syntheses, purifi-
3
cation, and NMR and analytical data may be found as ESI†. A
representative procedure follows.
N,NЈ-Bis(diphenylphosphino)-N,NЈ-(1-butanesulfonyl)-1,2-
THF with Et N as the base. Details of the syntheses, purifi-
3
cation, and NMR and analytical data may be found as ESI†. A
representative procedure follows.
diaminoethane (10): in the glovebox, a solution of Et N
3
N,NЈ-bis(1-butanesulfonyl)-1,2-diaminoethane (1): to
a
(
(
632 mg, 6.25 mmol) in 2 mL THF was added to a solution of 1
751 mg, 2.50 mmol) in 20 mL THF, and then a solution of
solution of 1,2-diaminoethane (3.3 mL, 50 mmol) in 25 mL of
pyridine cooled to 0 ЊC, n-butanesulfonyl chloride (13.0 mL,
Ph PCl (1.10 g, 5.00 mmol) in 3 mL THF was added dropwise
2
1
00 mmol) was added dropwise via syringe and then the
with magnetic stirring, immediately giving a white precipitate.
The mixture was allowed to stir overnight at room temperature.
reaction mixture was allowed to stir at room temperature over-
night. The resultant dark brown mixture was poured into a
flask containing 30 mL of concentrated HCl and 55 g of ice.
The precipitate was filtered and washed with water to give 10 g
of a brown solid. This material was dissolved in 60 mL of
ethanol and boiled with 2.5 g Norit for 5 min and then filtered.
Water was added to the yellow solution until it became cloudy
After filtering off the Et NHCl, solvent removal in vacuo gave
3
1
.64 g of product as a light yellow solid. This was taken up in 5
mL of CH Cl and 10 mL of diethyl ether was layered on.
2
2
Cooling to Ϫ35 ЊC overnight gave 1.32 g of white crystals, and a
second crystallization using 4.5 mL CH Cl and 9 mL of diethyl
2
2
ether in the same way gave 1.16 g (70% yield) of product as
(
ca. 60 mL), and then it was warmed to redissolve the product
1
analytically pure white crystals. H NMR (CD Cl ): δ 7.40 (m,
2
2
2
and allowed to cool slowly to Ϫ20 ЊC. After filtration and
washing with water, 4.7 g of a light yellow solid was obtained. A
second treatment failed to remove the color. Heating 4 g of this
material with 1 g of Norit in 30 mL of boiling CH Cl was
3
0H, Ph), 3.19 (br ∼t, A AЈ X m, J ≈ 3.2 Hz, 4H, CH CH ),
2 2 2 PH 2 2
2
.86 (AAЈBBЈ, JHH ≈ 7.9 Hz, 4H, SO CH ), 1.57 (m, 4H,
2 2
SO CH CH ), 1.33 (sextet, J
= 7.4 Hz, 4H, SO CH -
2 2
31
2
2
2
HH
2
2
CH CH ), 0.88 (t, J = 7.3 Hz, 6H, SO CH CH CH CH ). P
2 2 HH 2 2 2 2 3
13
followed by the addition of hexane until the solution turned
NMR (CD Cl ): δ 59.21 ppm. C NMR (CD Cl ): δ 135.11
2
2
2
2
cloudy. Cooling to Ϫ20 ЊC gave 3.5 g of white crystals (25%
1
2
(
d, J = 16.7 Hz), 132.95 (d, J = 21.6 Hz), 130.26 (s, C ),
PC PC 4
3 2
1
yield). H NMR (CDCl ): δ 4.90 (s, br, 2H, NH), 3.30 [ca.
3
128.96 (d, J = 6.2 Hz), 53.74 (d, J = 3.1 Hz, CH ), 49.24
PC PC 2
3
4
1
:0.5:1 A AЈ X multiplet, J
JHHЈ ≈ 6 Hz, 4H, NCH CH N], 3.05 (AAЈBBЈ, J ≈ 8.0 Hz,
= 6 Hz, J
= 0 Hz,
2
2
(CH)(NH)
(CHЈ)(NH)
(SO CH ), 25.58 (SO CH CH ), 21.81 (SO CH CH CH ),
2 2 2 2 2 2 2 2 2
3
2
2
HH
13.67 (SO CH CH CH CH ). Anal. calc. for C H N O -
2 2 2 2 3 34 42 2 4
4
1
7
H, SO CH ), 1.80 (∼quintet, J ≈ 7.8 Hz, 4H, SO CH CH ),
2 2 HH 2 2 2
S P : C, 61.06; H, 6.33; N, 4.19. Found: C, 60.69; H, 6.39; N,
2 2
4.11%.
N,NЈ-bis(diphenylphosphino)-N,NЈ-(1-butanesulfonyl)-2,2Ј-
diaminobiphenyl (16): in the glovebox 2.75 mL of n-BuLi (1.6 M
in hexane, 4.4 mmol) was added dropwise to a solution of 4
(850 mg, 2.0 mmol) in 20 mL of THF that had been pre-cooled
at Ϫ35 ЊC. The mixture was stored in the glovebox freezer at
.46 (sextet, JHH = 7.7 Hz, 4H, SO CH CH CH ), 0.96 (t, J
=
2
2
2
2
HH
13
.4 Hz, 6H, SO CH CH CH CH ). C NMR (CDCl ; assign-
2
2
2
2
3
3
ments from HETCOR): δ 52.72 (SO CH ), 43.76 (NCH ),
2
2
2
2
5.59 (SO CH CH ), 21.52 (SO CH CH CH ), 13.56
2 2 2 2 2 2 2
(
SO CH CH CH CH ). Anal. calc. for C H N O S : C, 39.98;
2 2 2 2 3 10 24 2 4 2
H, 8.05; N, 9.32. Found: C, 40.13; H, 8.34; N, 9.14%.
Ϫ35 ЊC for 0.5 h, and then a solution of Ph PCl (883 mg, 4.0
2
1
,3,2-Diazaphospholidines and 1,3,2-diazaphosphepines. The
mmol) in 4 mL of THF was added dropwise. The mixture was
allowed to warm to rt with stirring for 1 h, and then the THF
was removed in vacuo. Methylene chloride (30 mL) was added
to precipitate out LiCl, and the mixture was filtered through
Celite and the solvent removed to give a sticky yellow solid.
This material was resuspended in 8 mL of benzene, filtered
through Celite, and the solvent was removed to give a yellow
powder. Recrystallization from CH Cl –hexane (1:3) at Ϫ35 ЊC
phosphorus heterocycles 2, 3, 5, 6, 8, 19, and 20 were prepared
by combining either PhPCl , EtPCl , Cl PCH CH PCl , or
2
2
2
2
2
2
Cl PCH CH CH CH PCl with the appropriate sulfonamide
2
2
2
2
2
2
and Et N in THF. Details of the syntheses, purification, and
NMR and analytical data may be found as ESI†. A represent-
ative procedure follows.
3
2
-Phenyl-1,3-bis(1-butanesulfonyl)-1,3,2-diazaphospholidine
2
2
1
(
2): in the glovebox a solution of Et N (632 mg, 6.25 mmol) in 3
gave 600 mg (38% yield) of fine white crystals. H NMR (C D ):
3
6 6
mL THF was added to a solution of 1 (752 mg, 2.50 mmol) in
0 mL THF, and the resultant solution was cooled for 10 min
in a Ϫ35 ЊC freezer. A solution of PhPCl (447 mg, 2.50 mmol)
δ 8.92 (d, 7.6 Hz, 1H), 8.08 (∼t, 6.7 Hz, 5H), 7.76 (d, 8.0 Hz,
1H), 7.70 (br s, 2H), 7.45 (∼t, 6.7 Hz, 1H), 7.04–7.25 (m, 17 H),
6.90 (t, 7.3 Hz, 1H), 2.80 (m, 1H), 2.41 (m, 1H), 1.90 (br m, 2H),
1
2
D a l t o n T r a n s . , 2 0 0 3 , 3 8 7 – 3 9 4
393

View Article Online
1
0
6
1
.52 (m, 2H), 1.37 (m, 1H), 1.17 (m, 1H), 0.93–1.04 (m, 4H),
References
31
.54–0.91 (m, 6H). P NMR (C D , ca. 21 ЊC): δ 70.08, 69.83,
6
6
13
1 W. H. Hersh, P. Xu, B. Wang, J. W. Yom and C. K. Simpson,
7.99 (ca. 4:54:42). C NMR (C D ): δ 140.01, 136.78, 136.24,
6
6
Inorg. Chem., 1996, 35, 5453.
R. L. Pruett and J. A. Smith, J. Org. Chem., 1969, 34, 327.
34.89, 131.27, 131.09, 130.51, 129.25, 129.16, additional peaks
2
likely overlapping C D at 128–129, 56.12, 53.87, 32.12, 25.37,
6
6
3 G. W. Parshall and S. D. Ittel, Homogeneous Catalysis,
Wiley-Interscience, New York, 1992.
2
3.21, 21.89, 21.68, 14.52, 13.57. Anal. calc. for C H N -
44 46 2
O S P : C, 66.65; H, 5.85; N, 3.53. Found: C, 65.16; H, 5.64; N,
4 P. W. N. M. van Leeuwen, C. P. Casey and G. T. Whiteker, in
Rhodium Catalyzed Hydroformylation, P. W. N. M. van Leeuwen
and C. Claver, eds. Kluwer Academic Publishers, Dordrecht, 2000.
4
2
2
3
.36%.
5
M. P. Magee, W. Luo and W. H. Hersh, Organometallics, 2002, 21,
3
62.
Measurement of gas solubility
6
C. P. Casey, G. T. Whiteker, M. G. Melville, L. M. Petrovich,
Dry distilled solvent was placed in a 1 L flask attached to a
vacuum stopcock, which was connected to a known volume
attached to a vacuum line and mercury manometer. The solvent
was degassed by opening it to the evacuated known volume
and manometer, and (with rapid stirring) allowing it to degas;
equilibration was rapid, and since the process was repeated
many times as described below, stirring speed is not important.
The solvent stopcock was closed and the system re-evacuated.
Since the vapor volume in the solvent flask and the known
volume were comparable, this had the effect of reducing the gas
concentration by around half with each cycle. By 8–11 cycles,
the vapor pressure was constant, and this value was used in the
final determination of gas solubility. An aliquot of gas was then
admitted to the known volume (with manometric pressure
determination), and then the stopcock was opened to the
rapidly stirred solvent. After equilibration, the pressure minus
the known vapor pressure allowed the amount of gas in
solution to be calculated by difference. The volume of solvent
was then determined by weight (since some is lost during the
degassing procedure). The final solubility was then extrapolated
to 1 atm partial pressure assuming Henry’s law (Pgas = KHenryX2)
J. A. Gavney and D. R. Powell, J. Am. Chem. Soc., 1992, 114, 5535.
7 M. Kranenburg, Y. E. M. van der Burgt, P. C. J. Kamer,
P. W. N. M. van Leeuwen, K. Goubitz and J. Fraanje, Organo-
metallics, 1995, 14, 3081.
8
S. E. Denmark, S. P. O’Connor and S. R. Wilson, Angew. Chem.,
Int. Ed., 1998, 37, 1149.
G. T. Whiteker, A. M. Harrison and A. G. Abatjoglou, J. Chem.
Soc., Chem. Commun., 1995, 1805.
9
10 J. Clayden, P. Johnson and M. Helliwell, J. Org. Chem., 2000, 65,
7
033.
1
1 T. Ohwada, I. Okamoto, K. Shudo and K. Yamaguchi, Tetrahedron
Lett., 1998, 39, 7877.
2 W. H. Hersh, J. Chem. Ed., 1997, 74, 1485.
3 G. Heckmann and E. Fluck, Magn. Reson. Chem., 1995, 33, 553.
14 R. S. Dickson, P. S. Elmes and W. R. Jackson, Organometallics,
1999, 18, 2912.
5 L. A. van der Veen, P. H. Keeven and A. L. Spek, Organometallics,
1
1
1
2
000, 19, 872.
1
1
6 E. Wilhelm and R. Battino, Chem. Rev., 1973, 73, 1.
7 R. W. Cargill, in IUPAC Solubility Data Series: Carbon Monoxide,
J. W. Lorimer, ed., Pergamon Press, Oxford, 1990, vol. 43.
18 M. Catté, C. Achard, C.-G. Dussap and J.-B. Gros, Ind. Eng. Chem.
Res., 1993, 32, 2193.
9 M. W. Cook and D. N. Hanson, Rev. Sci. Instrum., 1957, 28, 370.
0 R. Battino and H. L. Clever, Chem. Rev., 1966, 66, 395.
1 J. A. Waters, G. A. Mortimer and H. E. Clements, J. Chem. Eng.
Data, 1970, 15, 174.
22 J. Grimblot, J. P. Bonnelle, A. Mortreux and F. Petit, Inorg. Chim.
Acta, 1979, 34, 29.
23 A. van Rooy, J. N. H. Debruijn, K. F. Roobeek, P. C. J. Kamer and
P. W. N. M. van Leeuwen, J. Organomet. Chem., 1996, 507, 69.
4 A. van Rooy, E. N. Orij, P. C. J. Kamer and P. W. N. M. van
Leeuwen, Organometallics, 1995, 14, 34.
5 A. van Rooy, P. C. J. Kamer, P. W. N. M. van Leeuwen, K. Goubitz,
J. Fraanje, N. Veldman and A. L. Spek, Organometallics, 1996, 15,
835.
1
2
2
where X is the mole fraction of gas in solution (and by conven-
2
16,19,20
tion X1 is the mole fraction of solvent).
Duplicate
determinations were made for each solvent (toluene with N2
and CO, THF with CO, and CH Cl with CO at 20 and 5 ЊC),
2
2
using different initial gas pressures to reduce a possible source
of systematic error. Minor effects that were quantified were the
volume of the stir bar, and the variable known volume due to
the changes in the height of the mercury column. More major
effects that were not well controlled were the difference in gas
and solvent temperature, although successful use of such a
2
2
21
set-up has been reported; the solvent was stirred in a
thermostatted water bath, while the gas temperature was
approximated as that of the lab, rather than holding the entire
26 A. M. Trzeciak, T. Glowiak, R. Grzybek and J. J. Ziolkowski,
J. Chem. Soc., Dalton Trans., 1997, 1831.
7 S. C. van der Slot, P. C. J. Kamer, P. W. N. M. van Leeuwen,
2
J. Fraanje, K. Goubitz, M. Lutz and A. L. Spek, Organometallics,
19
apparatus in a thermostatted compartment.
2
000, 19, 2504.
2
2
3
8 A. G. Abatjoglou, E. Billig and D. R. Bryant (to Union Carbide),
US Pat., 4,769,498, 1988.
9 K. Diemert, W. Kuchen and J. Kutter, Phosphorus Sulfur, 1983, 15,
Acknowledgements
1
55.
0 Y. S. Varshavskii and T. G. Cherkasova, Russ. J. Inorg. Chem., 1967,
Financial support for this work from the National Institutes
of Health (GM59596) and the City University of New York
PSC-CUNY Research Award Program is gratefully
acknowledged.
1
2, 899.
3
3
1 E. Brunner, J. Chem. Eng. Data, 1985, 30, 269.
2 C. L. Young, in IUPAC Solubility Data Series: Hydrogen,
Deuterium, A. S. Kertes, ed., Pergamon Press, Oxford, 1981, vol. 5/6.
3
94
D a l t o n T r a n s . , 2 0 0 3 , 3 8 7 – 3 9 4