Branched-Selective Hydroformylation of Nonactivated Olefins Using an N-Triphos/Rh Catalyst

Article abstract of DOI:10.1021/acscatal.8b00566

We report a catalytic system comprised of nitrogen-centered di- or triphosphine ligands in conjunction with rhodium that is capable of delivering branched aldehydes from terminal olefin substrates which commonly give more linear aldehydes than branched. The incorporation of an apical nitrogen atom into the ligand backbone dramatically improves the reaction rate. Mechanistic and labeling studies suggest the unusual selectivity is due to the irreversible trapping of the Rh-alkyl species along the branched pathway, in comparison to the more reversible linear pathway. A precatalytic equilibrium mixture of rhodium species was observed by high-pressure in situ NMR spectroscopy, suggesting this equilibrium is the catalytic resting state.

Full text of DOI:10.1021/acscatal.8b00566

Subscriber access provided by Kaohsiung Medical University
Article
Branched-Selective Hydroformylation of Non-
Activated Olefins Using a N-Triphos/Rh Catalyst
Andreas Phanopoulos, and Kyoko Nozaki
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.8b00566 • Publication Date (Web): 17 May 2018
Downloaded from http://pubs.acs.org on May 17, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 36
ACS Catalysis
1
2
3
4
5
6
7
8
9
BranchedꢀSelective Hydroformylation of Nonꢀ
Activated Olefins Using a NꢀTriphos/Rh Catalyst
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Andreas Phanopoulos and Kyoko Nozaki*
Department of Chemistry and Biotechnology, Graduate School of Engineering, The University
of Tokyo, 7ꢀ3ꢀ1 Hongo, Bunkyoꢀku, Tokyo 113ꢀ8656, Japan
Abstract
We report a catalytic system comprised of nitrogenꢀcentered diꢀ or triphosphine ligands in
conjunction with rhodium that is capable of delivering branched aldehydes from terminal olefin
substrates which commonly give more linear aldehydes than branched. The incorporation of an
apical nitrogen atom into the ligand backbone dramatically improves the reaction rate.
Mechanistic and labelling studies suggest the unusual selectivity is due to the irreversible
trapping of the Rh–alkyl species along the branched pathway, in comparison to the more
reversible linear pathway. A preꢀcatalytic equilibrium mixture of rhodium species was observed
by highꢀpressure in situ NMR spectroscopy, suggesting this equilibrium is the catalytic resting
state.
ACS Paragon Plus Environment
1

ACS Catalysis
Page 2 of 36
1
2
3
4
5
6
Key words: NꢀTriphos, rhodium, branchedꢀselective, nonꢀactivated olefin, irreversible branched
pathway
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Introduction
The production of aldehydes from olefins via hydroformylation annually amounts to a global
production of more than 10 million tons of soꢀcalled “oxo” products.[1] As such, it is one of the
most extensively studied homogeneous catalytic processes in both academia and industry.[2] Due
to the commercial interest in plasticizers derived from linear alcohols (e.g. bis(2ꢀ
ethylhexyl)phthalate derived from nꢀbutanal),[1a] the production of linear aldehydes have
traditionally been the most desired products, with many known catalysts now capable of very
high linearꢀselectivity.[3]
Branched aldehydes, which result from the branchedꢀselective hydroformylation of terminal
olefins, usually require special substrates, such as allenes, vinylarenes or vinylacetate to ensure
high regioselectivity.[1,4,5] In this context, the branchedꢀselective transformation of easily
available bulk terminal olefins represents a challenging but increasingly desirable reaction. These
products would have broad utility in the fragrance and flavor, and lifeꢀsciences industries, as well
as being important intermediates in organic synthesis.[6]
To date, few examples of catalyst systems capable of performing branchedꢀselective
hydroformylation of nonꢀactivated terminal olefins have been reported.[7] Two notable examples
of highly selective systems that represents the current stateꢀofꢀtheꢀart for this reaction are a
rhodium encapsulation complex reported by Reek and coꢀworkers[8] and a Rh/BOBPHOS system
reported by Clarke and coꢀworkers.[9,10] The former is constructed from a tripyridylphosphine
ligand that coordinates to a rhodium center through phosphorus, and three metallo–porphyrin
ACS Paragon Plus Environment
2

Page 3 of 36
ACS Catalysis
1
2
3
4
5
6
7
8
9
units via the three nitrogens within the pyridyl moieties of the ligand (Chart 1, bottom left).
Reek’s encapsulated rhodium catalyst is able to distinguish between not only the 1ꢀ and 2ꢀ
positions of 1ꢀoctene, but also between the 2ꢀ and 3ꢀpositions of transꢀoctꢀ2ꢀene.[8f] Similar
internal olefin differentiation has also been reported using a bulky trisꢀbinaphthyl
monophosphite ligand, but in this case the selectivity was proposed to stem from favorable
interactions between the metal center and terminal ester groups incorporated within the
substrate.[11]
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The Clarke system utilizes a phosphite–phospholane bidentate ligand (Chart 1, bottom right),
and owes its branchedꢀselectivity to subtle substrate–ligand interactions that render otherwise
low energy pathways to linear aldehydes unproductive.[10] This system shows not only high
branched selectivity for a range of substrates including alkyl olefins and allyl arenes, but also
delivers chiral aldehydes with a high degree of enantiomeric selectivity.[9a] Beyond an academic
setting, the Rh/BOBPHOS system has been used for the hydroformylation of allylglycine
derivatives, capable of affording a chiral intermediate that is a key compound for the synthesis of
new antibiotics.[9b]
ACS Paragon Plus Environment
3

ACS Catalysis
Page 4 of 36
1
2
3
4
5
6
C H
CH2Ph
CH2CN
CH2CH2OH
R =
cat.
4
9
CHO
linear aldehyde
+
R
R
CO/H2
R
CHO
branched aldehyde
7
8
9
Ligands and complexes discussed in this work
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
R2P
Ph2P
Ph2P
N
PR2
N
PPh2
PPh2
PR2
PPh2
2
3
R =
phenyl
1a
1b
1c
1d
1e
m-C6F2H3
p-C6H4OMe
p-C6H4NMe2
o-tolyl
N
P
P
H
Rh
P
P
N
P
P
P = PPh2
4
cf. Ligands previously reported for branch-selective Rh-catalyzed hydroformylation
N
N
Zn
N
tBu
Ph
N
N
N
N
P
N
N
O
O
Zn
P O
P
N
Ph
N
N
tBu
(Sax,S,S)-BOBPHOS
N
Zn
N
N
encapsulation ligand
Chart 1. Branchedꢀselective Rhꢀcatalyzed hydroformylation of olefins. The ligands and complex
discussed in this study (center) and ligands reported in previous studies (bottom).
ACS Paragon Plus Environment
4

Page 5 of 36
ACS Catalysis
1
2
3
4
5
6
7
8
9
Herein, we report a simple achiral catalyst system formed in situ from nitrogenꢀcentered diꢀ or
triphosphine ligands 1 and [Rh(acac)(CO)2] (acac = acetylacetonate) and their application to the
hydroformylation of terminal olefins described in Chart 1, top. While the substrates commonly
undergo linearꢀselective hydroformylation, rather unusual branchedꢀselectivities were detected in
this study. The triphosphine ligand 1a remained selective even at relatively low loadings and
temperatures, while bidentate ligand 2 afforded predominantly linear aldehydes under these
conditions. The results will be discussed in comparison to the conventional triphosphine ligand
3. Isolation of bisꢀligated complex 4 and its behavior as a catalyst precursor are also discussed.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Results and Discussion
1. Hydroformylation of 1-hexene catalyzed by Rh complexes of aza-triphosphine ligands
1a–e. The tridentate ligands 1a–e were synthesized in a modular fashion by adapted literature
procedures,[12] allowing facile generation of a series of tridentate phosphine ligands with various
substituents on the phosphorus atom (Scheme 1). With this family of ligands in hand, we
investigated their competency for the hydroformylation of 1ꢀhexene as a model substrate when
combined with [Rh(acac)(CO)2] in situ. In addition, we compared the nitrogenꢀcentered ligands
(1a–e and 2) with the carbonꢀcentered analogue, CH3C(CH2PPh2)3 (3).
ACS Paragon Plus Environment
5

ACS Catalysis
Page 6 of 36
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. Synthetic procedure for the facile generation of ligands 1a–e.
1.1. Catalytic Study with [Rh(acac)(CO)2]/1a.
As a starting point, tris(diphenylphosphinomethyl)amine 1a was used in conjunction with
[Rh(acac)(CO)2] as the catalyst for the hydroformylation of 1ꢀhexene. Representative results are
summarized in Table 1. In all cases, only 1ꢀheptanal and 2ꢀmethylhexanal products were
observed unless otherwise stated. When the reaction was performed with 1 mol % Rh loading, a
Rh:L = 1:2 at 100 °C under CO/H2 (1:1) = 2.0 MPa for 3.5 h ([Rh] = 5 mM, [L] = 10 mM), this
resulted in the production of aldehydes in high yields and a slight regioselective bias towards the
branched product (99.4%, b/l = 1.10, Table 1, entry 1). This was somewhat unexpected, as the
structurally related carbonꢀcentered ligand tris(diphenylphosphinomethyl)ethane (3) was
reported to be linearꢀselective under only slightly different conditions (b/l = 0.22–0.93).[13,14] To
establish whether the divergent regioselectivities are a direct result of the introduction of an
apical nitrogen atom to the ligand, a control experiment using 3 under identical conditions was
conducted (Table 1, entry 2). In this case, unlike the reported data, slightly improved branchedꢀ
selectivity was observed but at the expense of a greatly reduced rate, thus only 16.1% of 1ꢀ
hexene was consumed (see 2.4. Proposal for the active species and catalytic cycle for further
ACS Paragon Plus Environment
6

Page 7 of 36
ACS Catalysis
1
2
3
4
5
6
7
8
9
discussion). In the presence of ligand, no isomerization of 1ꢀhexene to 2ꢀhexene or 3ꢀhexene was
observed, however a control reaction without ligand afforded aldehydes derived from isomerized
hexenes as well as hydrogenated sideꢀproduct hexane (Table 1, entry 3). The bite angle of
multidentate ligands has previously been shown to influence the regioselectivity during
hydroformylation. The bite angle for 1a when coordinated to RhI cannot be assessed with great
precision due to the absence of published crystal structures, but lowꢀquality data obtained within
our laboratory has allowed us to estimate its bite angle to be 88°, while that of carbonꢀcentered 3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
is 89° based on crystal structure data. Consequently, bite angle does not appear to be a
differentiator between these two ligands.
Table 1. Hydroformylation of 1ꢀhexene with azatriphos (1a)/Rh under various reaction
conditionsa
[Rh]
pCO
pH2
CHO TOF
Entry Temp. (°C)
b/lb
Comment
(mM)
(MPa) (MPa)
(%)b
(h-1)
1c
2d
3c,e
4
100
100
100
100
90
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
5.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
99.4
16.1
85.6
71.4
33.9
15.9
6.30
2.80
0.84
0.00
28
4.6
24
1.10
1.27
Standard
Control with 3
1.11 Control without ligand
71
1.06
1.12
5
34
6
16
1.22
80
Temperature
1.41
7
6.3
2.8
0.8
0.0
70
(Figure 1)
1.92
8
60
9
5.46
ꢀꢀꢀ
50
10c
25
ACS Paragon Plus Environment
7

ACS Catalysis
Page 8 of 36
1
2
3
4
5
6
7
8
9
11c
12c
13c
14
15
16
17
18
19
100
100
100
100
100
100
100
100
100
1.0
1.0
1.0
0.5
1.5
3.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
0.5
1.5
3.0
90.3
72.4
24.4
83.9
53.8
27.7
56.3
70.6
75.0
181
290
488
84
1.09
1.05
1.10
1.09
1.11
1.13
1.08
1.07
1.07
2.5
1.25
0.25
5.0
[Rh]
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
pCO
5.0
54
(Figure S56)
5.0
28
5.0
56
pH2
5.0
71
(Figure S58)
5.0
75
aReaction conditions: toluene (3.75 mL), 1ꢀhexene (250 ꢁL, 500 mM), [Rh] = [Rh(acac)(CO)2],
L = 1a, Rh:L = 1:2, 1 h; branched and linear aldehydes were the only products in all cases.
bDetermined by GC, confirmed by NMR; b = 2ꢀmethylhexanal, l = 1ꢀheptanal. c3.5 h. d3.5 h, L =
3. eNo ligand, hexane formed (12.4%).
In order to investigate the reaction mechanism several parameters were examined, including
temperature, substrate and catalyst concentration, and partial pressures of CO and H2. Firstly,
temperature was varied from 25–100 °C at constant CO/H2 pressure (Table 1, entries 4–10). At
lower temperatures a significant increase in selectivity is observed (b/l up to 5.46 at 50 °C); but
at temperatures below 100 °C the overall activity is also significantly reduced (Figure 1), with no
reaction occurring at room temperature over 3.5 hours under 2.0 MPa CO/H2 (Table 1, entry 10).
Reducing the catalyst loading from 1.0–0.05 mol % ([Rh] = 5–0.025 mM) while keeping the
Rh:L ratio at 1:2 ([L] = 10–0.05 mM) did not affect the selectivity (Table 1, entries 11–13). This
suggests higher nuclearity species, for instance dimeric compounds with Rh–Rh bonds or
featuring 1a in a bridging mode between two Rh centers, are less likely to be responsible for the
ACS Paragon Plus Environment
8

Page 9 of 36
ACS Catalysis
1
2
3
4
5
6
7
8
observed selectivity (vide infra). Additionally, kinetic analysis of the rate compared to [Rh] and
[1ꢀhexene] revealed both to be first order (Figure S54 and S55), suggesting coordination of
olefin to a monometallic active species to be rateꢀdetermining.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
80
70
60
50
40
30
20
10
0
6.00
5.50
5.00
4.50
4.00
3.50
3.00
2.50
2.00
1.50
1.00
Aldehyde
Selectivity
40
50
60
70
80
90
100
110
Temperature (°C)
Figure 1. Temperature dependence of total yield of aldehydes and b/l selectivity for data in
entries 4–9 in Table 1.
Varying the partial pressure of CO had a dramatic effect on catalyst activity, with increased
pressures severely hindering the reaction (Table 1, entries 14–16). In parallel to decreasing the
overall reaction rate, a small increase in regioselectivity towards the branched product was
observed, with b/l increasing from 1.09 to 1.13 when CO pressure is increased from 0.5 to 3.0
MPa at 100 °C. To probe the dependence on CO pressure further, this parameter was also
evaluated at 50 °C, as this temperature was shown to promote the formation of branched
products (Figure S57). Again, a similar decrease in catalyst activity was observed with
increasing CO pressure, while overall the selectivity also increased from 1.60 to 2.09 as the CO
pressure is increased from 0.2 to 2.0 MPa. A slightly lower pressure range was used for the
ACS Paragon Plus Environment
9

ACS Catalysis
Page 10 of 36
1
2
3
4
5
6
7
8
9
analysis at 50 °C as at pressures above 2.0 MPa at this temperature only trace amounts of
products are formed, making accurate analysis impossible. In fact, even at 2.0 MPa, very small
amounts of aldehydes were formed (<2%), with values approaching the accurate detection limits
of the instruments. Consequently, although a general increase in branched selectivity was
observed with CO pressure, this relationship needs to be evaluated to a greater degree of
accuracy for conclusive results. Other mechanistic studies have also shown that CO pressure can
significantly affect the regioselectivity observed during hydroformylation, normally to a much
greater degree.[15] For instance, Rosales et al. reported that during the hydroformylation of 1ꢀ
hexene with ligand 3, increasing the CO pressure from 0.1 to 0.55 MPa resulted in an increase in
linear aldehydes from l/b = 1.5 to 5.8.[14] Under their catalytic conditions, olefin coordination is
regioꢀdetermining, and is consequently highly affected by CO pressure which competes for
coordination to the metal center. On the other hand, the regioꢀdetermining step in the present
system is likely to be hydride migration (see section 2.1. Deuterioformylation) and is
consequently less effected by CO pressure. The partial pressure of H2 had a smaller effect on the
activity, showing only a slight rate enhancement at high pressures, and also had no effect on the
selectivity (Table 1, entries 17–19). Preliminary kinetic analysis of this data at 100 °C shows an
order of ꢀ0.6 in pCO and +0.15 in pH2 (Figures S54 and S56, respectively).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1.2. Comparison of ligands 1a–1e: Electronic and steric effects
By modulating the substituents at phosphorus, the electronic and steric properties of the ligand
can be easily tuned. To this end, several triphosphine derivatives were examined that contain
various groups at phosphorus (Table 2). The introduction of electron withdrawing metaꢀ
difluorophenyl substituents (1b)[12c] did not retard the rate, but did alter the selectivity towards
ACS Paragon Plus Environment
10

Page 11 of 36
ACS Catalysis
1
2
3
4
5
6
7
8
9
more linearꢀselective products (Table 2, entry 1). On the other hand, the incorporation of
electronꢀdonating but sterically small substituents paraꢀmethoxyphenyl (1c) or paraꢀN,Nꢀ
dimethylaminophenyl (1d) improves the branchedꢀselectivity marginally (Table 2, entries 2 and
3) with a concomitant decrease in activity. This implies there is a slight electronic preference for
branched aldehyde formation with increasingly electronꢀdonating ligands.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The electronic nature of coordinating ligands has been extensively explored for monoꢀ and
bidentate systems during hydroformylation. In general, less basic phosphorus containing
moieties such as (di)phosphites (rather than phosphines) afford aldehydes with greater linear
selectivity.[16,17] This was reasoned to be due to faster βꢀhydride elimination from the branched
Rh–alkyl species but not the linear species.[16a] In the present system, however, formation of the
branched Rh–alkyl complex is suggested to be irreversible (see section 2.1.
Deuterioformylation), so the observed effect must be more subtle. In a separate line of thought,
the regioselectivity dependence on ligand electronic parameters was thought to originate from a
perturbation of the equatorial–equatorial:equatorial–axial (ee:ea) isomer ratio present during
catalytic turnover when bidentate ligands are used. Upon changing the electron donating ability
across a series of diphosphine ligands, the resultant complexes showed a higher proportion of the
ee isomer, which is known to give more linear aldehydes.[18–20] Subsequent investigations,
however, have revealed this relationship to be primarily steric in nature.[17,20] In the present
system, the more basic ligand could have caused the higher ‘hydricity’ of the Rh–H species[21] to
increase the proportion of branched Rh–alkyl species present since the terminal olefinic carbon is
more positive than the internal one.[22] As for the reaction rate, electron donating ligands
generally lower the hydroformylation rate due to a strengthening of the Rh–CO bonds, retarding
CO dissociation and the formation of the required unsaturated species for olefin coordination.[17]
ACS Paragon Plus Environment
11

ACS Catalysis
Page 12 of 36
1
2
3
4
5
6
7
8
As may be expected, the use of an electronꢀdonating but sterically encumbering substituent,
orthoꢀtolyl (1e), results in an increased linear bias (Table 2, entry 6), presumably due to localized
steric pressure around the metal center.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
As had been confirmed by the temperature dependence study, lowering the temperature from
100 to 50 °C gave improved branchedꢀselectivity using 1a as the ligand. Indeed, under otherwise
standard conditions (i.e. 3.5 h, CO/H2 (1:1) = 2.0 MPa, [Rh] = 5 mM, [L] = 10 mM), reducing
the temperature to 50 °C, increased b/l from 1.10 at 100 °C to 1.66 using 1a (Table 1, entry 1 and
Table 2, entry 7). A similar selectivity enhancement was observed for the most branchedꢀ
selective ligand, 1d (Table 2 entries 3 and 4), showing very high selectivity at 50 °C (b/l = 5.91),
albeit with low activity.
To increase the activity and maintain high branchedꢀselectivity, low CO pressures (0.2 MPa)
and low temperatures (50 °C) were implemented using 1a and 1d as ligands. Utilizing 1a,
conversion was relatively high, while selectivity was somewhat reduced compared to the
standard conditions (Table 2, entry 8). This further suggests higher CO pressures are indeed
beneficial for the formation of branched aldehydes. Still, this represents one of the best catalytic
systems capable of producing large quantities (>80% conversion) of branched aldehydes (b/l =
1.29) using relatively short reaction times (24 hours) reported to date. For comparison, the
BOBPHOS system gave 78% conversion after 46 hours (b/l = 3.0) for 1ꢀhexene, and additionally
showed high selectivity and activity across a series of unactivated olefins (b/l > 4 and TOF up to
417 hꢀ1). [9a] Using 1d under these low pressure and temperature conditions resulted in significant
amounts of hydrogenated products (68% hexane), and showed poor selectivity (Table 2, entry 5).
In this case, the migration of CO to form a Rh–acyl is likely severely retarded due to the low CO
ACS Paragon Plus Environment
12

Page 13 of 36
ACS Catalysis
1
2
3
4
5
6
pressure, and strong Rh–CO bond from the increased electron density at rhodium from the
donating ligand.
7
8
9
Table 2. Catalytic data for the hydroformylation of 1ꢀhexene with different azatriphos ligandsa
10
11
12
b/lb
Entry
Catalyst
[Rh]/1b
[Rh]/1c
Temp. (°C) pCO (MPa) pH2 (MPa) CHO (%)b TOF (h-1)
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1c
2c
3c
4c
5d,e
6
100
100
100
50
1.0
1.0
1.0
1.0
0.2
1.0
1.0
0.2
1.0
1.0
1.0
1.0
1.0
1.0
0.8
1.0
1.0
0.8
1.0
1.0
97.7
79.2
57.8
0.76
6.21
70.9
4.81
82.9
89.4
6.33
28
23
0.92
1.11
1.16
5.91
0.99
0.43
1.66
1.29
1.10
1.48
17
[Rh]/1d
0.2
0.3
20
50
[Rh]/1e
[Rh]/1a
100
50
7c
8e
9
1.4
3.5
26
50
100
50
4
10
1.8
aReaction conditions: toluene (3.75 mL), 1ꢀhexene (250 ꢁL, 500 mM), [Rh] = [Rh(acac)(CO)2]
(1.0 mol %, 5 mM), 1 (2.0 mol %, 10 mM), CO (1.0 MPa), H2 (1.0 MPa), 3.5 h; branched and
linear aldehydes were the only products in all cases unless otherwise stated. bDetermined by GC,
confirmed by NMR; b = 2ꢀmethylhexanal, l = 1ꢀheptanal. cAverage of two separate runs.
dSignificant hexane formed (67.9%). e24 h.
1.3. Comparison with aza-diphosphine ligand 2.
Mechanistic studies suggest that during catalysis the triphosphine ligands only coordinates to
rhodium through two arms, leaving one free phosphine ‘dangling’ (vide infra). This begs the
question as to whether the third arm of each ligand is necessary for the observed regioselectivity.
ACS Paragon Plus Environment
13

ACS Catalysis
Page 14 of 36
1
2
3
n
To this end, a diphosphine ligand, BuN(CH2PPh2)2 (2) that contains a nꢀbutyl carbon chain
instead of a third phosphine arm,[23] was evaluated and showed similar catalytic activity to 1a
under standard reaction conditions ([Rh] = 5 mM, [L] = 10 mM, 100 °C, CO/H2 (1:1) = 2.0 MPa:
Table 3, entries 1 and 5). However, at lower ligand loadings ([L] = 5 mM) under optimized
CO/H2 pressures (0.2/0.8 MPa) and low temperatures (25–50 °C), the tridentate 1a system not
only maintains, but shows greater branchedꢀselectivity (Table 3, entry 4), while the bidentate 2
system changes to give predominantly linear aldehydes (Table 3, entry 8). Additionally, at lower
ligand loadings using 2 afforded more hexane (2%) in contrast to 1a, which remains highly
chemoselective.
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
These observations show that under these conditions there is a switch in regioselectivity
depending on whether diꢀ or triphosphine ligands are used. This may be due to a change in regioꢀ
determining step, or the result in different speciation at lower temperatures. Although we are
uncertain what change is responsible, it is clear that the tridentate ligand engenders divergent
reactivity compared to the bidentate analogue.
Table 3. Catalytic data for the hydroformylation of 1ꢀhexene comparing biꢀ and tridentate
ligandsa
[Rh]
[L]
Temp.
(°C)
pCO
pH2
CHO TOF
Entry
Catalyst
b/lb
(mM) (mM)
(MPa) (MPa)
(%)b
(h-1)
1
2
5.0
5.0
5.0
5.0
5.0
10
5.0
5.0
5.0
10
100
100
50
1.0
1.0
0.2
0.2
1.0
1.0
1.0
0.8
0.8
1.0
99.4
90.1
66.3
3.73
93.1
28
26
1.10
0.99
1.30
2.06
1.10
[Rh]/1a
[Rh]/2
3c
4c
5
2.8
0.2
27
25
100
ACS Paragon Plus Environment
14

Page 15 of 36
ACS Catalysis
1
2
3
4
5
6
7
8
6
5.0
5.0
5.0
5.0
5.0
5.0
100
50
1.0
0.2
0.2
1.0
0.8
0.8
>99.9
99.4
29
4.1
1.5
1.07
1.26
0.54
7c
8c
25
35.4
9
aReaction conditions: toluene (3.75 mL), 1ꢀhexene (250 ꢁL, 500 mM), [Rh] = [Rh(acac)(CO)2]
(1.0 mol %), 3.5 h; branched and linear aldehydes were the only products in all cases unless
otherwise stated. bDetermined by GC, confirmed by NMR; b = 2ꢀmethylhexanal, l = 1ꢀheptanal.
c24 h.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1.4. Substrate scope.
Next the substrate scope for this catalytic system was evaluated with several terminal olefins
which have previously been targeted for branchedꢀselective hydroformylation (Table 4).[9a,10,24]
These substrates were subjected to the standard hydroformylation conditions using both 1a and 2
as ligands. Substrates with phenyl (Table 4, entries 1 and 4) and cyano groups were investigated
(Table 4, entries 2 and 5), as well as an alcoholic substrate, 3ꢀbutenol (Table 4, entries 3 and 6).
Allylbenzene was found to perform similar to 1ꢀhexene, producing only aldehydes and
predominantly branchedꢀselective products (b/l = 1.33) whether ligand 1a or 2 were used. Allyl
cyanide showed exceptionally high branchedꢀselectivity (up to b/l = 6.12), with 1a performing
better than 2, but with greatly reduced chemoselectivity compared to substrates without
heteroatoms (31.8–40.2% selectivity for aldehyde products). 3ꢀButanol, on the other hand, did
not afford aldehyde products at the end of the reaction, as both linear and branched products had
been cyclized to form lactols (Scheme 2). The amount of fiveꢀ versus sixꢀmembered ring lactols
allows us to determine how much branched and linear aldehydes had been formed, respectively,
ACS Paragon Plus Environment
15

ACS Catalysis
Page 16 of 36
1
2
3
4
giving a selectivity of b/l = 1.33–1.39 depending on which ligand was used. Again, despite
5
6
giving lactol products, the alcoholic moiety did not affect the regioselectivity of the system.
7
8
9
Table 4. Catalytic data for the hydroformylation of various substrates using ligands 1a and 2a
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Aldehydes/Cyclised Products
Entry
Catalyst
Substrate
Yield (%)b TOF (h-1)
b/lb
1
2
3
4
5
6
allylbenzene
allyl cyanide
3ꢀbutenol
99.8
40.2
99.2
99.7
31.8
97.4
29
11
28
29
9.1
28
1.33
6.12
1.39c
1.32
4.98
1.33c
[Rh]/1a
allylbenzene
allyl cyanide
3ꢀbutenol
[Rh]/2
aReaction conditions: toluene (total volume 4.0 mL), substrate (2 mmol, 500 mM), [Rh] =
[Rh(acac)(CO)2] (1.0 mol %, 5 mM), ligand (2.0 mol %, 10 mM), 100 °C, CO/H2 (1:1) = 2.0
MPa, 3.5 h. bDetermined by NMR, in each case starting material showed quantitative conversion.
cFiveꢀmembered lactol/sixꢀmembered lactol.
OH
O
O
CHO
HO
cat.
CO/H2
cat.
CO/H2
HO
CHO
OH
HO
Scheme 2. Hydroformylation of 3ꢀbutenol and subsequent cyclization to form lactols.
ACS Paragon Plus Environment
16

Page 17 of 36
ACS Catalysis
1
2
3
4
5
6
7
8
9
2. Mechanistic studies. Recently, Clarke and coꢀworkers reported a similar kinetic profile for
the asymmetric hydroformylation of allylbenzene using BOBPHOS as that observed in the
current system, both of which showed high branchedꢀselectivity.[10] Additionally, several kinetic
studies of diphosphine systems have also been reported, which revealed similar CO and H2
dependencies.[25] Several energetically similar steps were found to contribute to the overall rate
of the reaction, including the olefin insertion/hydride transfer and product forming
hydrogenolysis steps, suggesting a similar mechanism may be operational in this work. In
Clarke’s BOBPHOS system, the selectivity was determined during an irreversible olefin
insertion step to Rh–H, which was confirmed by a labeling study. In the present study with
Rh/1a, the olefin insertion step is also likely to be regioꢀdetermining as neither substrate
concentration nor H2 pressure affected the b/l ratio (Table 1, entries 17–19). To confirm whether
the olefin insertion is reversible or not, a deuterioformylation study was first conducted.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2.1. Deuterioformylation
By treatment of 1ꢀhexene with D2 and CO in the presence of Rh/1a, it was confirmed that
hydride migration is largely irreversible for branched Rh–alkyl formation, while it is reversible
for linear Rh–alkyl formation. The deuterioformylation was conducted using [Rh(acac)(CO)2] (5
mM) and 1a (10 mM) under a 1:1 CO/D2 mixture at 1.0 MPa for 1 hour. The products were
1
2
analyzed by H, H and GC/GCꢀMS (Figure S59) and compared to previously reported
deuterium incorporated products from the deuterioformylation of the same substrate.[26] The
results are summarized in Figure 2. The recovery of the unreacted 1ꢀhexene was 20.4%. The
branched aldehyde (branchedꢀCHO) and the linear aldehyde (linearꢀCHO) with deuterium
incorporated at the βꢀposition to the ꢀCDO group were obtained in 29.8% and 36.4% yield,
ACS Paragon Plus Environment
17

ACS Catalysis
Page 18 of 36
1
2
3
4
5
6
7
8
9
respectively. The other detected products were, 2ꢀdeuterioꢀ1ꢀhexene (hexaneꢀ2d, 5.6%) and
branched product 2ꢀmethylhexanal with deuterium labels at the αꢀ and aldehyde carbons
(branchedꢀCHO*, 2.7%), both of which are exclusive products from the reverse hydride
migration from an initial linear Rh–alkyl intermediate. Notably, no products were observed that
are the result of βꢀhydride elimination of a branched Rh–alkyl species such as hexaneꢀ1d or
linearꢀCHO*. This demonstrates that formation of linear Rh–alkyl intermediates is somewhat
reversible while formation of the analogous branched Rh–alkyl intermediate is not.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
= detected species
hexene
20.4% recovery
D
D
hexene-1d
0%
hexene-2d
5.6%
H
D
H
Rh
branched
insertion
linear
insertion
Rh
Rh
D
linear
insertion
branched
insertion
D
Rh
D
D
Rh
D
D
Rh
Rh
branched Rh–alkyl
linear Rh–alkyl
O
D
O
D
O
O
D
D
D
D
D
D
linear-CHO*
0%
branched-CHO
29.8%
linear-CHO
37.4%
branched-CHO*
2.7%
Figure 2. Graphical representation of the product distribution observed during
deuterioformylation. Percentages represent total amounts based on initial substrate loading.
2.2. NMR studies in the absence of olefin.
The following NMR studies provided some information towards determining the identity of
potentially active species. After subjecting the preꢀcatalytic 1:2 mixture of [Rh(acac)(CO)2] and
azatriphos 1a to the catalytic conditions ([Rh] = 5 mM, [1a] = 10 mM, CO/H2 (1:1) = 2.0 MPa,
ACS Paragon Plus Environment
18

Page 19 of 36
ACS Catalysis
1
2
3
4
5
6
7
8
9
100 °C, 0.5 h), the major species detected after removal of the solvent and any volatiles was the
bisꢀligated monohydride complex [RhH(κ2ꢀ1a)2] (4) by 31P{1H} NMR (δP: 15.4 (d, 1JPRh = 141.0
Hz, 4P), ꢀ30.9 (s, 2P)) (Scheme 3). This was confirmed to be the correct assignment by a
separate synthesis (see Supporting Information), and in solution is probably tetrahedral with the
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
hydride jumping from face to face, as evidenced from broadening of the hydride signal by H
NMR spectroscopy (Figure S46) as well as broadening of the coordinated phosphorus signals by
31P{1H} NMR spectroscopy at –50 °C (Figure S49).[27] When complex 4 was used as the catalyst
under otherwise identical conditions (Table 2, entries 9 and 10), the same reactivity and
selectivity was observed as the in situ generated catalyst, confirming that this complex is at least
a competent preꢀcatalytic species. Interestingly, this differs significantly from the carbonꢀ
centered triphos ligand 3, with which the monoꢀligated tridentate complex [RhH(CO)(κ3ꢀ3)] was
isolated and reported to be the catalytic resting state.[14] Although the corresponding tridentate
complex [RhH(CO)(κ3ꢀ1a)] (A) with ligand 1a was observed under pressures of CO/H2 (vide
infra), this complex is probably less stable compared to 4 (see section 2.5. Comparison of
azatriphos 1a and carbon-centered triphos 3 for further discussion). The propensity for bisꢀκ2ꢀ
ligation of 1a to Rh over κ3ꢀligation is evident from the reaction of 1a with [RhCl(CO)(PPh3)2] in
a 1:1 stoichiometry, which exclusively forms 0.5 equiv. [Rh(κ2ꢀ1a)2][Cl] and leaves 0.5 equiv.
[RhCl(CO)(PPh3)2] unreacted (Scheme 3).
ACS Paragon Plus Environment
19

ACS Catalysis
Page 20 of 36
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 3. Formation of bisꢀligated complex 4 with 1a and monoꢀligated complex with 3.
The catalyst precursor 4 was converted to Rh carbonyl complexes upon treatment with CO or
with a mixture of CO and H2 as shown in Figure 3. Exposure of complex 4 (7.5 mM) in C6D6 to
CO (5 bar) provided an equilibrium mixture of several species (Figure 3, top). The three major
species in equilibrium have been assigned as [RhH(CO)(κ3ꢀ1a)] (A), [Rh(CO)(ꢁꢀCO)(κ2ꢀ1a)]2
(B), and [RhH(CO)2(κ2ꢀ1a)] (C) based on 1H, 13C{1H} 31P{1H}, 1Hꢀ31P HMBC and DOSY NMR
spectroscopy (Figures S58–S63), as well as by comparison to similar reported compounds (see
Supporting Information for discussion on assignment).[14,28] These three species existed in the
ratio A:B:C = 1.0:9.0:6.4 at room temperature and free 1a and H2 (from the formation of dimeric
1
B, 76% accounted for by H NMR spectroscopy) were also observed. Two species (B and C)
were observed when 4 was treated with CO/H2 (1:1, 5 bar), in a ratio of B:C = 1.0:2.4 at 25 °C
1
(Figure 3, bottom). The considerably broader peaks detected by H NMR spectroscopy suggest
increased fluxionality in comparison to exposure to only CO. Heating the mixture of 4/CO/H2
from 25–80 °C resulted in the formation of species A in addition to B and C, while cooling back
to 25 °C reestablishes the original equilibrium mixture (Figure S66). When these experiments
were performed using a 1:2 mixture of [Rh(acac)(CO)2] and 1a instead of preꢀformed 4, the same
equilibrium mixture of A–C was observed.
ACS Paragon Plus Environment
20

Page 21 of 36
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Stacked 1H NMR spectra of complex 4 under 5 bar CO (left top) and CO/H2 (left
bottom); stacked 1H NMR spectra of hydride region of complex 4 under 5 bar CO (middle top)
and CO/H2 (middle bottom); and stacked 31P{1H} spectra of complex 4 under 5 bar CO (right
top) and CO/H2 (right bottom): showing equilibrium mixture of species A–C.
2.3. NMR studies in the presence of olefin.
In an attempt to observe Rh–acyl species, the behavior of complex 4 in the presence of 1ꢀhexene
and either CO or CO/H2 was investigated (Figure 4). In the absence of CO or H2, complex 4 is
unreactive towards 20 equiv. 1ꢀhexene (Figure 4b). Although, Rh–acyl species could not be
detected when the mixture of 4 and 1ꢀhexene were exposed to CO (5 bar), an almost identical
equilibrium mixture of species was observed as in the absence of olefin (A:B:C = 1.0:8.0:6.2,
Figure 4c). Upon changing the atmosphere to H2, aldehyde production was observed (Figure 4d);
ACS Paragon Plus Environment
21

ACS Catalysis
Page 22 of 36
1
2
3
4
5
6
suggesting at least one of the species generated from 4/CO is active. Furthermore, a branchedꢀ
selective bias in aldehyde production was observed (b/l = 1.05), supporting the involvement of
7
8
31
these species within the actual catalytic cycle. On the onset of aldehyde production, the P{1H}
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
NMR spectra are simplified, showing only two detectable species: 4 and A (Figure 4e–g).
31
Complexes 4 and A are also the only resolvable species by P{1H} NMR spectroscopy when a
mixture of 4 (12.6 mM) and 1ꢀhexene (20 equiv.) are directly exposed to CO/H2 (5 bar), with
concomitant production of aldehyde, even at 25 °C under these low CO pressures. This further
implicates olefin coordination after CO dissociation to be at least partially rate determining.
Figure 4. Stacked 31P{1H} NMR spectra showing sequential reaction of 4 with CO, then H2 in
the presence of 1ꢀhexene (left) and stacked 1H NMR spectra showing onset of aldehyde
production (middle) and Rh–H signals (right).
2.4. Proposal for the active species and catalytic cycle.
ACS Paragon Plus Environment
22

Page 23 of 36
ACS Catalysis
1
2
3
4
5
6
7
8
9
Although only bisꢀligated complex 4 and tridentate complex A are observed during catalytic
turnover, these species must lie offꢀcycle as they are coordinatively saturated 18ꢀelectron
compounds with no bound olefin, alkyl or acyl moiety. In order to account for the large negative
dependence on CO pressure with respect to rate, the active catalytic species is likely to be
present as a fast, unresolvable equilibrium on the NMR timeꢀscale. From the mechanistic results,
as well as the identification of B and C, the following catalytic cycle may be reasonably
proposed (Figure 5). In this cycle, an equilibrium exists as the resting state, with only 4 and A in
sufficient quantities or of sufficient lifetime to be detected by NMR spectroscopy during
catalytic turnover. The dissociation of CO and subsequent olefin coordination to form olefin
complex E is predominantly rate determining. Next, olefin insertion takes place accompanied by
additional CO coordination to form linear or branched Rh–alkyl species F. While formation of
the branched Rh–alkyl was irreversible, the formation of the linear Rh–alkyl species was proven
to be reversible to some extent. In conjunction with higher CO pressures increasing branched
selectivity, it seems the reversibility of the linear pathway acts to enhance an intrinsic propensity
for branched Rh–alkyl formation during olefin insertion, making this step likely to be regioꢀ
determining. Once F is formed, CO insertion followed by hydrogenolysis of the Rh–acyl species
completes the cycle.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
23

ACS Catalysis
Page 24 of 36
1
2
3
4
5
6
H
Rh
P
= present during/at end of catalysis
= present before H2
Olefin insertion:
Irreversible for branched pathway
likely regio-determining
CO
CO
H
P
R
P
P
Rh
CO
C
7
CO
R
R
– CO
8
E
H
9
P
CO
R
P
Rh
P
κ3–κ2
H
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
P
P
P
/
P
Rh
CO
D
P
Rh
P
Rh
CO
CO
– 1a
+ CO
A
CO
CO
0.5 H2
– 0.5 H2
F
0.5 H2
– 0.5 H2
O
O
H
P
Rh
R
– 1a
+ CO
– 0.5 H2
R
O
P
P
OC
P
P
H
H2
CO
P
P
CO
Rh
Rh Rh
0.5
G
P
P
P
4
O
B
O
CO
P
Rh
R
P
CO
H
Figure 5. Proposed catalytic cycle for the hydroformylation of 1ꢀhexene using
[Rh(acac)(CO)2]/1a as the catalyst. P represents PPh2 from a single arm of the potentially
tridentate 1a, in the cases where only two “P” groups are coordinated to the central Rh, the third
arm is uncoordinated.
2.5. Comparison of azatriphos 1a and carbon-centered triphos 3
As noted during the catalytic studies, the incorporation of a nitrogen atom into an otherwise
entirely hydrocarbon ligand backbone (i.e. comparing ligand 1a and 3) led to a dramatic rate
enhancement, which has also been reported for a related diphosphine systems.[29] Similar rate
enhancements are also observed for instance in CO2 hydrogenation when pendent amines are
tethered to diphosphine ligands.[30] Although the cause of this rate enhancement remains
debated,[3d,31] several explanations have been proposed such as the amines acting to assist in the
heterolytic cleavage of H2.[32]
ACS Paragon Plus Environment
24

Page 25 of 36
ACS Catalysis
1
2
3
4
5
6
7
8
9
Within the context of the present system, CO pressure was found to be the most important
parameter in altering the rate of reaction, not H2 pressure. Accordingly, formation of a species
with a vacant site (species D in Figure 5) to allow olefin coordination is vital to improve the rate.
With a bidentate coordinated 1a or 3, this vacant site could be occupied by olefin, CO or
phosphine during catalysis.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In the case of 3, we know the tridentate complex was the sole species observed under catalytic
conditions,[14] however ligand 1a prefers to bind in a bidentate fashion (e.g. species 4, B and C),
therefore increasing the amount of unsaturated species present (Scheme 3 and Figure 3). We
propose that the propensity of 1a to not coordinate through all three phosphine groups due to
electronic instability could be advantageous to form species D leading to the observed rate
enhancement. The electronic instability when 1a coordinates through all three phosphine
moieties is due to the lone pair on the apical nitrogen, which is donated into the three flanking
Cmethylene–P σ*ꢀorbitals, hindering backdonation from rhodium to phosphorus and resulting in a
disproportionately electron rich metal center (Chart 2).[12c,29] This electron donation does not
exist for carbonꢀcentered 3 which favors tridentate coordination due to chelation.[14]
Chart 2. Electronic instability of tridentate coordination of 1a.
Conclusions
ACS Paragon Plus Environment
25

ACS Catalysis
Page 26 of 36
1
2
3
4
5
6
7
8
9
In conclusion, a series of nitrogenꢀcentered diꢀ and triphosphine ligands have been synthesized
with various substituents at phosphorus and tested in conjugation with Rh for the
hydroformylation of terminal olefins. For 1ꢀhexene, allylbenzene, allyl cyanide, and 3ꢀbutenol,
branchedꢀselectivity was observed, ultimately delivering branched aldehydes, except when
phosphines with bulky substituents were used, forming predominantly linear products. Notable
divergent reactivity between the diꢀ and triphosphine analogues occurred at low loadings and low
pressures where the triphosphines maintain branched selectivity, while diphosphines did not.
Kinetic studies demonstrated that increasing H2 pressure results in a slight rate increase while
regioselectivity was negligible. On the other hand, increasing CO pressure dramatically reduces
the overall reactivity, and results in some selectivity enhancement towards the branched product.
A deuterioformylation study confirmed olefin insertion to be irreversible for the branched
pathway and likely regioꢀdetermining. The rare selectivity observed in this system likely stems
from a trapping of the branched Rh–alkyl species which inevitably forms branched aldehydes,
while the corresponding linear Rh–alkyl species shows some reversibility. Mechanistic studies
allowed the characterization of several rhodium species that likely exist in a dynamic equilibrium
as the catalytic resting state.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
AUTHOR INFORMATION
Corresponding Author
*Eꢀmail: nozaki@chembio.t.uꢀtokyo.ac.jp
Author Contributions
The manuscript was written through contributions of all authors. All authors have given approval
to the final version of the manuscript.
ACS Paragon Plus Environment
26

Page 27 of 36
ACS Catalysis
1
2
3
4
Funding Sources
5
6
7
8
Funds from the Japan Society for the Promotion of Science (JSPS) were used to support the
research in this manuscript.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Notes
The authors declare no competing financial interest.
ASSOCIATED CONTENT
Supporting Information.
The following file is available free of charge on the ACS Publications website at DOI: xxx.
Further experimental details; NMR, IR, and mass data; preliminary kinetic analysis;
deuterioformylation study results; additional in situ highꢀpressure NMR spectra including
variable temperature data; representative GC chromatograms and NMR spectra of aldehyde
product identification (PDF)
ACKNOWLEDGMENT
AP is grateful to the JSPS for provision of a JSPS Standard Fellowship (P16338). We are
grateful to Jingyuan Deng for proofreading and advice on final manuscript.
REFERENCES
(1) (a) Cornils, B.; Herrmann, W. A.; Beller, M.; Paciello, R. In Applied Homogeneous
Catalysis with Organometallic Compounds, 3rd ed.; WileyꢀVCH Verlag GmBH & Co.
KCaA: Weinheim, Germany, 2018; Vol. 1, p 25–89. (b) Franke, R.; Selent, D.; Börner,
A. Applied Hydroformylation. Chem. Rev. 2012, 112, 5675–5732.
ACS Paragon Plus Environment
27

ACS Catalysis
Page 28 of 36
1
2
3
4
5
6
7
8
9
(2) (a) van Leeuwen, P. W. N. M. V.; Claver, C. In Rhodium Catalyzed Hydroformylation, 1st
ed.; Kluwer Academic Publisher: New York, USA, 2002; Vol. 22, p 1–33. (b) Börner, A.;
Franke, R. In Hydroformylation. Fundamentals, Processes, and Applications in Organic
Synthesis, 1st ed.; WileyꢀVCH Verlag GmbH & Co. KCaA: Weinheim, Germany, 2016;
Vol. 1, p 24–35. (c) Tudor, R.; Ashley, M. Enhancement of Industrial Hydroformylation
Processes by the Adoption of RhodiumꢀBased Catalyst: Part I. Platinum Metals Rev.
2007, 51, 116–126. (d) Mahatthananchai, J.; Dumas, A. M.; Bode, J. W. Catalytic
Selective Synthesis. Angew. Chem., Int. Ed. 2012, 51, 10954–10990.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(3) (a) Casey, C. P.; Whiteker, G. T.; Melville, M. G.; Petrovich, L. M.; Gavney, Jr., J. A.;
Powell, D. R. Diphosphines with Natural Bite Angles near 120° Increase Selectivity for
nꢀAldehyde Formation in RhodiumꢀCatalyzed Hydroformylation. J. Am. Chem. Soc.
1992, 114, 5535–5543. (b) Kranenburg, M.; van der Burgt, Y. E. M.; Kamer, P. C. J.; van
Leeuwen, P. W. N. M. New Diphosphine Ligands Based on Heterocyclic Aromatics
Inducing Very High Regioselectivity in RhodiumꢀCatalyzed Hydroformylation: Effect of
the Bite Angle. Organometallics 1995, 14, 3081–3089. (c) Kalck, P.; Park, D. C.; Serein,
F.; Thorez, A. Influence of Various Parameters on the Selectivity of the Production of
Aldehydes Starting from Alkenes Issued from the Biomass and Using the Catalyst
Precursors Rh2(ꢀꢀSR)2(CO)2(PA3)2. J. Mol. Catal. 1986, 36, 349–357. (d) Jiao, Y.; Torne,
M. S.; Gracia, J.; Neimantsverdriet, J. W.; van Leeuwen, P. W. N. M. Ligand Effects in
RhodiumꢀCatalyzed Hydroformylation with Bisphosphines: Steric or Electronic? Catal.
Sci. Technol. 2017, 7, 1404–1414.
(4) (a) Gual, A.; Godard, C.; Castillón, S.; Claver, C. Highlights of the RhꢀCatalysed
Asymmetric Hydroformylation of Alkenes Using Phosphorus Donor Ligands.
ACS Paragon Plus Environment
28

Page 29 of 36
ACS Catalysis
1
2
3
4
5
6
Tetrahedron: Asymmetry 2010, 21, 1135–1146. (b) Agbossou, F.; Carpentier, J.ꢀF.;
Mortreux, A. Asymmetric Hydroformylation. Chem. Rev. 1995, 95, 2485–2506.
7
8
9
(5) (a) Nozaki, K.; Sakai, N.; Nanno, T.; Highashijima, T.; Mano, S.; Horiuchi, T.; Takaya,
H. Highly Enantioselective Hydroformylation of Olefins Catalyzed by Rhodium(I)
Complexes of New Chiral Phosphine–Phosphite Ligands. J. Am. Chem. Soc. 1997, 119,
4413–4423. (b) Yu, S.; Chie, Y.ꢀM.; Zhang, X. Highly Regioselective and Rapid
Hydroformylation of Alkyl Acrylates Catalyzed by a Rhodium Complex with a
Tetraphosphorus Ligand. Adv. Synth. Catal. 2009, 351, 537–540. (c) Wassenaar, J.; de
Bruin, B.; Reek, J. N. H. RhodiumꢀCatalyzed Asymmetric Hydroformylation with
TaddolꢀBased IndolPhos Ligands. Organometallics 2010, 29, 2767–2776. (d) Noonan, G.
M.; Cobley, C. J.; Mahoney, T.; Clarke, M. L. Rhodium/Phospholane–Phosphite
Catalysts Give Unusually High Regioselectivity in the Enantioselective
Hydroformylation of Vinyl Arenes. Chem. Commun. 2014, 50, 1475–1477.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(6) Li, H.; Dong, K.; Jiao, H.; Neumann, H.; Jackstell, F.; Beller, M. The Scope and
Mechanism of PalladiumꢀCatalysed Markovnikov Alkoxycarbonylation of Alkenes.
Nature Chem. 2016, 8, 1159–1166.
(7) (a) da Silva, A. C.; de Oliveira, K. C. B.; Gusevskaya, E. V.; dos Santos, E. N. Rhodiumꢀ
Catalyzed Hydroformylation of Allylbenzenes and Propenylbenzenes: Effect of
Phosphine and Diphosphine Ligands on Chemoꢀ and Regioselectivity. J. Mol. Catal. A:
Chem. 2002, 179, 133–141. (b) Baber, R. A.; Clarke, M. L.; Heslop, K. M.; Marr, A. C.;
Orpen, A. G.; Pringle, P. G.; Ward, A.; ZambranoꢀWilliams, D. E. Phenylphosphatrioxaꢀ
Adamantanes: Bulky, Robust, Electronꢀpoor Ligands That Give Very Efficient
ACS Paragon Plus Environment
29