Hydrolysis stability of bidentate phosphites utilized as modifying ligands in the Rh-catalyzed n-regioselective hydroformylation of olefins

Article abstract of DOI:10.1021/acscatal.6b02185

The stability of ligands and catalysts is an almost neglected issue in homogeneous catalysis, but it is crucial for successful application of this methodology in technical scale. We have studied the effect of water on phosphites, which are the most applied cocatalysts in the n-regioselective homogeneous Rh-catalyzed hydroformylation of olefins. The stability of the bidentate nonsymmetrical diphosphite L1, as well as its two monophosphite constituents L2 and L3, toward hydrolysis was investigated by means of in situ NMR spectroscopy under similar conditions as applied in industry. Hydrolysis pathways, intermediates, and kinetics were clarified. DFT calculations were used to support the experimentally found data. The acylphosphite unit L2, which reacts with water in an unselective manner, was proven to be much less stable than the phenolphosphite L3. The stability of the bidentate ligand L1 can be therefore mainly attributed to its phenolphosphite moiety. With an excess of water, the hydrolysis of L1 and L2 as well as their Rh-complexes is first-order with respect to the phosphite. Surprisingly, coordination to Rh significantly stabilizes the monodentate ligand L2, while in strong contrast, the bidentate ligand L1 decomposes faster in the Rh complex. NMR spectroscopy provided evidence for the existence of species from decomposition of phosphites, which can likewise coordinate as ligands to the metal. Electron-withdrawing groups in the periphery of the acylphosphite moiety decrease the stability of L1, whereas 3,5-disubstituted salicylic acid derivatives with bulky groups showed superior stability. These modifications of L1 also give rise to different catalytic performances in the n-regioselective hydroformylation of n-octenes and 2-pentene, from which the 3,5-di-t-butyl-substituted ligand offered a higher n-regioselectivity accompanied by a lowering of the reaction rate in comparison to the parent ligand L1.

Full text of DOI:10.1021/acscatal.6b02185

Subscriber access provided by SUNY DOWNSTATE
Article
Hydrolysis Stability of Bidentate Phosphites Utilized as Modifying
Ligands in the Rh-Catalyzed n-Regioselective Hydroformylation of Olefins
Baoxin Zhang, Haijun Jiao, Dirk Michalik, Svenja Kloss, Lisa Marie Deter,
Detlef Selent, Anke Spannenberg, Robert Franke, and Armin Boerner
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.6b02185 • Publication Date (Web): 12 Sep 2016
Downloaded from http://pubs.acs.org on September 13, 2016
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 free 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 accessible to all
readers and 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.
ACS Catalysis 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 56
ACS Catalysis
1
2
3
4
5
6
7
8
9
Hydrolysis Stability of Bidentate Phosphites Utilized
as Modifying Ligands in the RhꢀCatalyzed nꢀ
Regioselective Hydroformylation of Olefins
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
,
§
§
§,†
†
†
§
Baoxin Zhang,* Haijun Jiao, Dirk Michalik, Svenja Kloß, Lisa Marie Deter, Detlef Selent,
§
ǁ,‡
,§,†
Anke Spannenberg, Robert Franke, and Armin Börner*
§
LeibnizꢀInstitut für Katalyse an der Universität Rostock e.V., A.ꢀEinsteinꢀStr. 29a, Dꢀ18059
Rostock (Germany)
†
Institut für Chemie der Universität Rostock, A.ꢀEinsteinꢀStr. 3, Dꢀ18059 Rostock (Germany)
ǁ
Evonik Performance Materials GmbH, PaulꢀBaumannꢀStr. 1, Dꢀ45772 Marl (Germany)
‡
Lehrstuhl für Theoretische Chemie, RuhrꢀUniversität Bochum, Dꢀ44780 Bochum (Germany)
ABSTRACT The stability of ligands and catalysts is an almost neglected issue in homogeneous
catalysis, but crucial for successful application of this methodology in technical scale. We have
studied the effect of water on phosphites being the most applied coꢀcatalysts in the nꢀ
regioselective homogeneous Rhꢀcatalyzed hydroformylation of olefins. The stability of the
bidentate nonsymmetrical diphosphite L1, as well as its two monophosphite constituents L2 and
L3, towards hydrolysis was investigated by means of in situ NMR spectroscopy under similar
conditions as applied in industry. Hydrolysis pathways, intermediates and kinetics were clarified.
ACS Paragon Plus Environment
1

ACS Catalysis
Page 2 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
DFT calculations were used to support the experimentally found data. The acylphosphite unit
L2, which reacts with water in an unselective manner, was proven to be much less stable than the
phenolphosphite L3. The stability of the bidentate ligand L1 can be therefore mainly attributed
to its phenolphosphite moiety. With an excess of water the hydrolysis of L1 and L2 as well as
their Rhꢀcomplexes is first order with respect of the phosphite. Surprisingly, coordination to Rh
significantly stabilizes the monodentate ligand L2, while in strong contrast the bidentate ligand
L1 decomposes faster in the Rh complex. NMR spectroscopy provided evidence for the
existence of species from decomposition of phosphites, which can likewise coordinate as ligands
to the metal. Electronꢀwithdrawing groups in the periphery of the acylphosphite moiety decrease
the stability of L1, whereas 3,5ꢀdisubstituted salicylic acid derivatives with bulky groups showed
superior stability. These modifications of L1 also give rise to different catalytic performances in
the nꢀregioselective hydroformylation of nꢀoctenes and 2ꢀpentene, from which the 3,5ꢀdiꢀtꢀbutyl
substituted ligand offered a higher nꢀregioselectivity accompanied by a lowering of the reaction
rate in comparison to the parent ligand L1.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
KEYWORDS homogeneous catalysis, phosphite, hydroformylation, rhodium, hydrolysis
stability
1. Introduction
A catalyst accelerates the rate of chemical reactions by lowering of the energies of transition
states by forming associates with the substrate. In the final step of the catalytic cycle the catalyst
is released from the product. In most textbooks, it is conceptually assumed that the catalyst is
fully regenerated and it therefore remains unchanged over the whole catalytic process. However
in practice, the highly reactive catalyst is frequently altered due to side reactions with other
components in the reaction mixture, e.g. intermediates, products, solvents as well as impurities,
ACS Paragon Plus Environment
2

Page 3 of 56
ACS Catalysis
1
2
3
4
5
6
7
8
9
which may lead in the consequence to deviations from the expected catalytic activity and
1
selectivity.
Unlike in heterogeneously catalyzed processes, where the deactivated catalyst can be easily
isolated and separated, any separation or concentration processes in homogeneous catalysis may
lead to further degradations of the relevant catalytically active or nonꢀactive species causing
ambiguous analytic results. Moreover, the low concentration of the catalyst under application
conditions requires specific spectroscopic methods for analysis. Therefore, studies addressing 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
,2
stability of homogeneous catalysts are very challenging and up to now quite rare.
Of particular economic and academic importance is the nꢀregioselective hydroformylation of
3
olefins. It is worldwide one of the largest homogeneously catalyzed processes on an industrial
scale. Aldehydes produced are valuable final products in bulk and fine chemistry. They can be
easily transformed into numerous other functionalized compounds like alcohols, esters and
4
amines. In comparison to conventionally applied cobalt catalyst, the more recently developed
rhodium based systems modified with trivalent phosphorus ligands offer the advantage of
running the reaction at much lower syngas pressure (1.8ꢀ6.0 MPa) and moderate temperatures
5
(85ꢀ130 °C).
An issue of enormous importance is the stability of the precious rhodium catalyst. In
continuous approaches its integrity has to be assured over several days, weeks or even months.
Decomposition of the original phosphorus ligand and loss of its ligating properties lower the
concentration of the active catalyst. In the consequence an alteration of the desired catalytic
results takes place. Due to the loss of the organic ligand “naked” rhodium reacts with CO to give
nonꢀselective, less active and finally entirely inactive rhodium carbonyl clusters. Also in batch
reactions the reuse of the original catalyst is of great value due to its high cost.
ACS Paragon Plus Environment
3

ACS Catalysis
Page 4 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Scheme 1. Hydrolysis of the PꢀOꢀbond and competition between formation of a tetravalent oxide
V
phosphorus compound (we use in this article the classic form P=O double bond and P denoting
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
6
this kind of species only for the sake of expression convenience ) and coordination of the
trivalent species to a metal
Decomposition pathways depend on the nature of the phosphorus ligands, and reaction
conditions applied. Phosphines are prone to oxidation and PꢀC bond activation, whereas
7
phosphorus ligands with PꢀO bonds, especially phosphites suffer solvolysis with water. In
technical hydroformylation conditions water is continuously formed together with the soꢀcalled
highꢀboilers due to the aldol condensation of the product aldehydes. Since phosphites generate
more active catalysts for the nꢀregioselective hydroformylation of olefins than phosphines, they
constitute one of the most utilized classes of ligands in this regard. Especially bidentate
diphosphites are regularly applied. With the assistance of such diphosphites annually hundred
thousands of tons aldehydes are produced. Therefore their behaviors towards water are of crucial
importance especially for the design of more resistant ligands and for the proper choice of the
8
reaction conditions (temperature, removal of water etc.). Besides, to mitigate the degradation of
phosphites, usually proper stabilizers containing tertiary amines or epoxides, or the ionꢀexchange
9
methods are applied in industry.
ACS Paragon Plus Environment
4

Page 5 of 56
ACS Catalysis
1
2
3
4
5
6
7
8
9
The effect of water on phosphites as well as phosphinites and phosphonites, is due to the
nucleophilic attack of PꢀOꢀbonds. In the result the alkoxy or aryloxy rest is substituted. Mostly
4
the product with trivalent phosphorus is stabilized as tetravalent oxide species by tautomerism.
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
It was reported in the older literature that the hydrolysis of trialkylphosphites follows a first
1
0
order kinetics in phosphite and a second order in water (Scheme 1). Spectroscopic experiments
1
8
17
with Oꢀ and Oꢀlabeled water and measurement of a strong isotope effect with D O provided
2
evidence that a PꢀOꢀ and not a CꢀOꢀbond is cleaved and that the attack of water on phosphorus is
1
1
the rate determining step.
Hydrolysis products of phosphites do not compellingly lose their complexation properties. Due
the phosphinylidene tautomerization, the trivalent phosphorous acid diester is able to coordinate
on metals. This property of Hꢀphosphonates, also abbreviated as HASPOs (heteroatom
substituted phosphine oxides), stimulated several research groups to elucidate the relevant
1
2
13
tautomeric equilibrium, their coordination behavior to metals and catalytic properties of
1
4,15
formed metal complexes.
Remarkably, already several years ago van Leeuwen utilized a
16
simple HASPO (HP(O)Ph ) in the platinum catalyzed hydroformylation, which clearly suggests
2
that the catalytic impact of such degradation products should not be underestimated. With one
1
7
exception concerned to the asymmetric rhodium catalyzed hydrogenation, in all other
instances, HASPOs were prepared prior to the catalytic process. However, the fact that HASPOs
can be formed as a result of the decomposition of phosphites during the catalytic reaction has
been seldom in the focus. Only recently, we got evidence that a HASPO which has been formed
during a hydroformylation process from a phosphite due to hydrolysis is able to affect the Rh
1
8
catalyzed hydroformylation.
ACS Paragon Plus Environment
5

ACS Catalysis
Page 6 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
In order to get a complete picture about the decomposition pathways and the unusual
catalytically active intermediates, we have initiated a detailed study using in situ NMR
spectroscopic analysis and theoretical calculations. As an example of technical relevance we
have chosen the hybrid bidentate phosphite L1 containing a mixed anhydride moiety
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
9
(acylphosphite) and a biphenol phosphite moiety (phenolphosphite) (Figure 1).
Figure 1. The bidentate hybrid diphosphite L1, in which the phosphite unit on the left hand side
is denoted as “acylphosphite” (red), and the part on the right hand side is named herein
“phenolphosphite” (blue).
Ligands based on this structure were proven to produce up to 69% nꢀnonanal in the Rh
catalyzed isomerizing hydroformylation of isomeric nꢀoctenes containing only 3.3% terminal
ꢀ1
octene with turnover frequencies (TOF) up to 4700 h .
Due this excellent catalytic performance their application in largeꢀscale hydroformylations has
been recommended. In this regard the question of its hydrolysis stability is a crucial issue. With
1
13
31
in situ H, C, and P NMR spectroscopy and DFTꢀcalculations the hydrolysis reactions of L1
and each of its two phosphite components as well as their substructures were monitored
qualitatively and quantitatively under conditions close to those as applied in industry. We found
3
1
in our study, that in particular in situ P NMR spectroscopy is an ideal method for studying
ACS Paragon Plus Environment
6

Page 7 of 56
ACS Catalysis
1
2
3
4
5
6
7
8
9
degradation of such phosphites in a low concentration. Due to the 100% natural abundance of
3
1
P, it has a high sensitivity to Pꢀcontaining structures allowing a quantitative analysis.
Moreover, usual hydrolysis products of phosphite are tetravalent phosphorous oxides which have
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
31
a polar PꢀO bond. Relevant signals appear near zero ppm in the P NMR spectrum, whereas
trivalent phosphites resonate between 120~150 ppm. Thus, a satisfying peak separation between
product and reactant can be realized. This situation facilitates the interpretation of the spectra.
After these hydrolysis studies, the coordination modes of decomposition products to rhodium
were evaluated. Subsequently, the hydrolysis resilience of phosphites as “free” ligands and in the
Rh coordinated mode was compared. Based on these results, finally, a modification of the parent
diphosphite has been carried out in order to improve its hydrolysis stability. Simultaneously, the
effects of these modifications on the catalytic performance were evaluated.
2. Experimental Section
The hydrolysis experiment was performed in a 0.049 or 0.0245 M solution of the phosphite in
31
1
,4ꢀdioxane sealed in an NMR tube and quantitatively monitored by in situ P NMR
spectroscopy. As an external reference a solution of tris(nꢀoctyl)phosphine oxide (TOPO)
dissolved in pꢀxyleneꢀd with a concentration of 0.259 M was sealed in a capillary and placed
1
0
into the NMR tube together with the sample solution. The stability of TOPO towards hydrolysis
and oxidation guarantees its constant amount and chemical properties during the whole
experiment even at a high temperature. The sealed capillary makes sure that none of the
components in the sample solution is interfered with the reference. This setꢀup created a
separated reaction mixture, in which each species consumed and/or produced could be monitored
qualitatively as well as quantitatively. The concentration of the external reference was regulated
ACS Paragon Plus Environment
7

ACS Catalysis
Page 8 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
to give a comparable integral ratio between the phosphite and TOPO for the sake of accuracy.
Integrals of the detected species normalized by that of TOPO can now be regarded as their
concentrations in this isolated environment.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3. Results and Discussion
3.1. Hydrolysis of the bidentate hybrid phosphite (L1)
Ligand L1 bears two different phosphite units. This electronically and sterically unsymmetric
structure contains two alternative positions for the water attack. In order to assess the relative
stability of the “left arm” and the “right arm” phosphite units in L1 the diphosphite was treated
with 100 equivalents of water at 120 °C.
1
31
V
Figure 2. In situ H (left and middle, Ph and tBu regions) and P (right, P region) NMR spectra
of the hydrolysis of L1 with 100 eq. of water at 120 °C. Reaction time increased from below to
above. The last spectrum was taken after 6.5 h of the hydrolysis
.
ACS Paragon Plus Environment
8

Page 9 of 56
ACS Catalysis
1
2
3
4
5
6
7
8
9
In the beginning of the hydrolysis of L1 free salicylic acid and an intermediate phosphorous
1
acid monoester (L1DP1) were detected by H NMR spectroscopy (left and middle spectra in
Figure 2). L1DP1 exhibits signals of four nonꢀequivalent tꢀBu groups due to lack of Cs
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
31
symmetry. In the P NMR spectrum besides L1DP1 and the HASPO of L3 (structure discussed
in section 3.6), here denoted as L1DP4, another compound (herein abbreviated as L1DP2)
giving two signals in the tꢀBu region was found and characterized by a singlet at 0.03 ppm.
Further decomposition fragments of L1 such as free biphenol L1DP5 and phosphite phenol
1
L1DP3 became subsequently discernible in the H NMR spectrum. At this stage of hydrolysis
31
the signal of phosphorous acid could also be attributed in the P NMR spectrum. Corresponding
1
31
PꢀH splittings of all phosphorous acid derivatives agree well in both H and P NMR spectra.
1
3
Degradation to L1DP1 was additionally confirmed by in situ C NMR spectroscopy. After 5.5 h
of heating with water the only peaks of prominence could be attributed to salicylic acid and the
substituted biphenol L1DP5.
ACS Paragon Plus Environment
9

ACS Catalysis
Page 10 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
31
III
Figure 3. In situ P NMR spectra of hydrolysis of L1 (100 eq. of water, 120 °C): phosphite (P )
V
region (left), tetravalent oxide phosphorous (P ) region (middle) and concentration profiles of
hydrolysis products of L1 (100 eq. of water, 120 °C) (right).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
1
Noticeably, in the P NMR spectra no new signals appeared in the region of phosphites with
salicylic acid moiety. Also no HASPO of the acylphosphite and/or its derivatives could be
detected. An acylphosphite phenol structure has been proven to be unstable. Thus, the only
possible phosphorous acid derivative derived from L1 with two nonequivalent tꢀBu groups is a
phosphorous acid monoester with the biphenol (L1DP2 in Scheme 2). From these observations
the hydrolysis pathway of L1 can be established. L1 is selectively hydrolyzed to L1DP1 in the
first step. Hydrolysis of L1DP1 leads unselectively either to the phosphite phenol L1DP3 or to
the monoester L1DP2 and the HASPO L1DP4. The latter is decomposed to L1DP2, which is in
3
1
accord with the unsteadily low concentration profile of the HASPO found in the in situ P NMR
spectra. As described previously, the phosphite phenol (L1DP3) is quite resistant towards
18
hydrolysis. It is accumulated in the product mixture. L1DP2 degrades finally to the free
biphenol (L1DP5) and phosphorous acid. According to the proposed pathway, the amount of
phosphorous acid should be equal to the sum of L1DP3 and L1DP5, which was verified by the
31
quantitative analysis of the NMR spectra. The assignment of the signals of timeꢀdepending P
NMR spectra was achieved (Figure 3). Fitting concentration correlations of L1DP3/H PO and
3
3
III
V
P /P of L1DP1 consisting with the proposed pathway could be extracted from concentration
profiles of the products. The decay in concentration of L1 obeys first order kinetic under the
flooding condition, showing a halfꢀlife time of 24.49 h at 120 °C. A comparison of halfꢀlife times
of L1 under different conditions can be found in the supporting information.
ACS Paragon Plus Environment
10

Page 11 of 56
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
Scheme 2. Hydrolysis pathways of L1
3.2. Hydrolysis of monoꢀacylphosphite L2
As evidenced above the acylphosphite unit is more sensitive to the attack of water. In order to
simplify the analysis of the highly complicated NMR spectra of the emerging hydrolysis
intermediates, we investigated the hydrolysis of an isolated monoꢀphosphite structure derived
from the structure of L1. The monoꢀacylphosphite L2 as a model for the “left arm” of L1
consists of a sixꢀmembered cyclic mixed anhydride. It was prepared by the reaction of salicyl
phosphorchloridite and 2ꢀ(tertꢀbutyl)ꢀ4ꢀmethoxyphenol. Its molecular structure was solved with
Xꢀray single crystal diffraction (Figure 4). The structure, in which three PꢀO bonds are of similar
lengths, has a distorted trigonal pyramidal geometry about phosphorus with the lone electron pair
pointing towards the apical direction.
ACS Paragon Plus Environment
11

ACS Catalysis
Page 12 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
O
O
O
P
O
tBu
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
OMe
L2
Figure 4. Molecular structure of L2 (thermal ellipsoids at the 30 % probability level). Selected
bond lengths [Å] and angles [°]: P1ꢀO1 1.631(1), P1ꢀO2 1.638(1), P1ꢀO3 1.632(1), O1ꢀP1ꢀO3
1
02.09(5), O2ꢀP1ꢀO3 94.65(5), O1ꢀP1ꢀO2 99.93(5).
Figure 5. Concentration profile during the hydrolysis of L2 with 10 eq. of water at room
temperature.
We firstly studied the hydrolysis of L2 with a tenꢀfold amount of water (Figure 5). The
concentration profile of the starting phosphite under this flooding condition follows a firstꢀorder
kinetic with respect to L2, giving a halfꢀlife time of 16.5 h at room temperature. Two Pꢀ
containing products from the degradation process were observed and characterized by
1
1
resonances at δ = 2.6~3.6 ppm ( JP,H = 700 Hz) and δ = 4.5~6.3 ppm ( J
= 683 Hz),
P,H
respectively. Both resonances represent tetravalent phosphorous oxide species whose normalized
ACS Paragon Plus Environment
12

Page 13 of 56
ACS Catalysis
1
2
3
4
5
6
7
8
9
integrals along with the hydrolysis show that the latter is the predominant product. Both signals
2
0
shifted downfield when the reaction progressed.
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
31
Figure 6. Hydrolysis of L2 (with 1 eq. of water, at room temperature) monitored with in situ P
V
NMR spectroscopy (P region).
1
Figure 7. Left: selected in situ H NMR spectra for hydrolysis of L2 with 1 eq. of water at 90 °C
1
(
lowꢀfield and highꢀfield regions). Right: assignment of the H NMR spectrum of the reaction
solution after hydrolysis with 1 eq. of water at 90 °C for 6 h.
ACS Paragon Plus Environment
13

ACS Catalysis
Page 14 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
For a more detailed understanding of the hydrolysis mechanism of L2, in turn it was treated
with only one equimolar amount of water. The reaction was again monitored by in situ NMR
spectroscopy. Under the same conditions, four new sets of signals appeared in the timeꢀ
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
31
depending quantitative P NMR spectra (Figure 6). All resonances are located in the region
V
typical for P compounds. According to its concentration profile, the concentration of the
1
product at δ = 6.5~7.1 ppm ( JP,H = 689 Hz) increased at the beginning. This period is followed
by a decrease of the concentration towards the end. The other three species showed steadily
1
increasing concentration profiles. One signal (at δ = 3.7 ppm, JP,H = 700 Hz) even shows a high
formation rate. In comparison to the hydrolysis experiment with the tenꢀfold amount of water,
the trial with a phosphite/water ratio of 1:1 led to two new species with constantly low
concentration profile, besides the previously described main products.
31
Noteworthy, both hydrolysis experiments did not afford compounds characterized by P NMR
peaks in the region typical for trivalent phosphorus.
1
The assignment of each product was confirmed by H NMR spectroscopy. To accelerate the
1
reaction the hydrolysis was performed at 90 °C. The complete assignment of the signals in the H
NMR spectrum of the reaction mixture was based on the chemical shifts, coupling constants,
1
13
1
1
intensities of the H and C signals and the results of twoꢀdimensional NMR spectra ( H, H
1
1
1
13
1
13
COSY, H, H NOESY, H, C HSQC, H, C HMBC). Figure 7 and 8 show an example
describing the emerged components after a hydrolysis experiment at 90 °C for 6 h. We found
four PꢀH signals. Their coupling constants correspond well to those measured in the timeꢀ
3
1
depending P NMR spectra. The same holds for the relation of their integrals. This finding
3
1
allows an interpretation of the four P NMR peaks for products (Figure 6). Several overlapped
signals could be identified by integration, which remain true for every measured spectrum,
ACS Paragon Plus Environment
14

Page 15 of 56
ACS Catalysis
1
2
3
4
5
6
7
8
9
referenced by the MeOꢀ and tꢀBuꢀsignals of higher resolution. An interesting feature of the
tetravalent phosphorus species found in the products is that their PꢀH chemical shifts as well as
PꢀH coupling constants vary depending on the pH value and water amount of the solution. This
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
31
observation was made in both the H and P NMR spectra.
As illustrated in Figure 7, two new species bearing MeOꢀ and tꢀbutylꢀgroups became
observable after 0.5 h of reaction. One compound contains a proton directly bound to phosphorus
appearing at δ = 6.8 ppm with a large PꢀH coupling constant of 690 Hz. This PꢀH coupling
3
1
constant is consistent with that observed in the P NMR spectra observed at δ = 6.5~7.1 ppm. It
characterizes L2DP1. We also observed signals of the four aryl protons of salicylic acid which is
1
released from the starting phosphite. Another PꢀH signal, found at δ = 7.0 ppm ( JP,H = 703 Hz),
can be unambiguously assigned to phosphorous acid corresponding to the signal at δ = 3.7 ppm
1
31
(
JP,H = 700 Hz) in the P NMR spectra and belongs to phosphorous acid (L2DP2). The integral
1
31
ratios of L2DP1 and L2DP2 in the H and P NMR spectra were in nice agreement. After 6 h of
reaction time, signals of sets of three aryl protons became visible characterizing two new aryl
compounds with MeOꢀ and tꢀBuꢀgroups. These signals were overlapped by peaks of L2 and
salicylic acid. One set of the new signals belongs to 4ꢀmethoxyꢀ3ꢀtꢀbutylꢀphenol (L2DP3) that
arises from the hydrolysis of the exocyclic PꢀO bond of L2. The other product is a HASPO based
on the sixꢀmembered ring with salicylic acid backbone. The four aryl proton peaks of the
1
HASPO (L2DP4) were found as well although their intensity was low. The H chemical shifts of
L2DP4 were confirmed by comparison with the spectrum of the pure HASPO prepared in
31
1
parallel by us. It displays a P resonance at δ = ꢀ2.3 ppm ( J = 778 Hz) which agrees well
P,H
1
with the signal (δ = ꢀ2.3 ppm, JP,H = 772 Hz) detected in the hydrolysis experiment.
ACS Paragon Plus Environment
15

ACS Catalysis
Page 16 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
To summarize the in situ NMR investigations of the hydrolysis of L2 the following facts can
be concluded: a) at the beginning of hydrolysis signals of salicylic acid and phosphorous acid
were observed accompanied by two species bearing a phenol moiety; b) one of these new species
has been identified as 4ꢀmethoxyꢀ3ꢀtꢀbutylꢀphenol (L2DP3) arising from the cleavage of the
exocyclic PꢀO bond of L2; c) signals of a HASPO with a sixꢀmembered ring were likewise
detected at the beginning of the hydrolysis; d) by integral analysis it was proven that the amount
of salicylic acid was constantly equivalent to phosphorous acid and a new species formed. Also
the amount of the substituted phenol L2DP3 was always equal to the sum of phosphorous acid
and HASPO. Considering these observations with regard to the structure of L2, the only possible
structure left bearing an anisole unit resulting from hydrolysis of L2 is a phosphorous acid
monoester, which will be herein denoted as L2DP1. It is released due to cleavage of the two
endocyclic PꢀO bonds of the acylphosphite. The other product is salicylic acid. Taking in mind
that the phenol L2DP3 was already observed at the beginning of the hydrolysis together with the
HASPO and that both compounds remained in a low concentration profile, the full hydrolysis
pathways of L2 can be described as illustrated in Scheme 3. Most intermediates of the
decomposition are observable on the NMR time scale. In the presence of equimolar amounts of
water, the hydrolysis of the monoꢀacylphosphite proceeds in an unselective manner. Either both
endocyclic PꢀO bonds are cleaved simultaneously yielding the phosphorous acid monoester or
the exocyclic PꢀO bond is broken giving rise to the HASPO L2DP4. These phosphorous acid
monoꢀ and diesters are prone to further degradation. Especially L2DP4 immediately undergoes
hydrolysis via L2DP5 to give phosphorous acid. Therefore, two stoichiometric relations shown
in Scheme 3 can be deduced from these pathways, which are exactly in accordance with the
experimental findings.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment
16

Page 17 of 56
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
1
3
Figure 8. C NMR spectrum with assignment of reaction mixture of hydrolysis of L2 (90 °C, 1
eq. of water, after 6 h).
Scheme 3. Proposed hydrolysis pathways of L2 (“I” means the normalized integral area)
We were able to grow crystals from the hydrolysis product L2DP4. Its crystal structure is
depicted in Figure 9. The relatively short bond P1ꢀO2 (1.456 Å) suggests that the phosphorus
atom is in the tetravalent state. Thus, in solid state the molecule exists as HASPO isomer like
ACS Paragon Plus Environment
17

ACS Catalysis
Page 18 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
31
1
found in solution by NMR. In the P NMR spectrum it resonates at ꢀ3.4 ppm ( J = 760 Hz) in
P,H
1
CD Cl and ꢀ2.3 ppm ( J = 778 Hz) in our hydrolysis environment.
2
2
P,H
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 9. Molecular structure of L2DP4 (thermal ellipsoids at the 30 % probability level).
Selected bond lengths [Å] and angles [°]: P1ꢀO1 1.575(2), P1ꢀO2 1.456(2), P1ꢀO3 1.585(2), O1ꢀ
P1ꢀO2 113.2(1), O2ꢀP1ꢀO3 112.1(1), O1ꢀP1ꢀO3 104.36(8).
3
.3. DFT calculations of the hydrolysis of L2
The question “Which PꢀO bond in L2 is most fragile against water?” can be predicted by
means of DFT calculations by comparing the free energy for each hydrolysis pathway. Scheme 4
illustrates the thermodynamic stability of all the possible hydrolysis products of L2. For
simplicity water was chosen as solvent. The computed results give evidence that the exocyclic Pꢀ
O bond is less stable. Therefore, it should be preferentially cleaved, followed by the cyclic ester
bond. The anhydride PꢀO bond is the most stable one. Among the three pathways only the
cleavage of the exocyclic PꢀO bond is thermodynamically favored giving rise to the
corresponding HASPO, which could be in equilibrium with the corresponding phosphorous acid
diester, here abbreviated as PADE (Scheme 5).
ACS Paragon Plus Environment
18

Page 19 of 56
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
Scheme 4. Changes of Gibbs free energy (ΔG, kcal/mol) for cleavage of every PꢀO bond in L2
Scheme 5. Tautomeric equilibrium of L2DP4
A controversial result to the experimental findings is the calculated result that the PADE
isomer is thermodynamically more favored. It should be therefore predominant in the
equilibrium with the HASPO isomer. It holds also for products from the second hydrolysis step.
Due to the calculations, breakage of one of the PꢀO bonds in PADE is not favored, which would
finally lead to phosphorous acid and salicylic acid. Another thermodynamically favored product
is besides salicylic acid the phosphorous acid monoester L2DP1. Together with the sixꢀ
membered cyclic HASPO/PADE and phosphorous acid, these three Pꢀcontaining products are
the only energetically preferred compounds emerging from the hydrolysis.
ACS Paragon Plus Environment
19

ACS Catalysis
Page 20 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Concerning the experimental observations shown in Figure 6, that L2DP1 was the
predominant product at the beginning of hydrolysis, we conclude that the liberating of salicylic
acid should be kinetically controlled.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
It was found that the tautomeric equilibrium of the cyclic products could be reversed by a
change of the chemical environment. For example, our calculation showed that in the 7ꢀ
membered phenolphosphite unit L1DP4 the equilibrium was shifted from the PADE to the
HASPO isomer in an aqueous solution. Nevertheless in gas phase the PADE isomer was
calculated to be more stable. Another explanation of the disparate results obtained by theoretical
and experimental investigations could be the fact that the calculations refer to the hydrolysis in
neutral conditions. However, it should be taken in mind, that in the experiment protons were
increasingly generated from salicylic acid and phosphorous acid during progressing degradation
process. We assume that the presence of protons leads to an alteration of the hydrolysis and
2
1
favors simultaneously the tetravalent oxide isomer (Scheme 6). By lowering the energy barrier,
it leads the reaction preferentially to the direction of the HASPO isomer. Thus, HASPO L2DP4
and its decomposition intermediate L2DP5 cannot be detected until a late stage of the hydrolysis
as recorded in Figure 5. Depicted is there an autocatalysis process triggered by progressively
generated protons. It seems that protons have not only a strong impact on the PADE/HASPO
equilibrium but also on the stability of these intermediates. Although L2DP4 was continuously
formed, its concentration still kept constant at a low level in comparison to L2DP1. This
observation can be rationalized by the fast subsequent hydrolysis of L2DP4 via L2DP5 to
phosphorous acid. (Scheme 3)
ACS Paragon Plus Environment
20

Page 21 of 56
ACS Catalysis
1
2
3
4
5
6
7
8
9
Scheme 6. Possible Hydrolysis mechanism of phosphite in the presence of protons
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
From calculated bond lengths, both the cyclic PꢀO bonds of L2 become weaker after opening
of the ring structure. In the consequence, the anhydride bond is weakened as expected, and
becomes less stable than the ester bond. In contrast, the anhydride bond is more stable than the
ester in the cyclic PADEꢀisomer. (results in supporting information)
Obviously, the intracyclic PꢀO bonds of the PADE isomer of L2 get an extra stabilization by
the almost planar sixꢀmembered ring and the anellation with the aryl ring. An electronꢀdonating
MeOꢀgroup in pꢀposition to the acyl group of L2 does not exert a significant stabilizing effect on
the stability towards hydrolysis, which is in agreement with the experimental finding (vide infra).
When it is replaced by the electronꢀwithdrawing Clꢀgroup, all three PꢀO bonds are activated.
3.4. Stabilization effects by substituents
In order to study the effect of substituents on the hydrolysis stability of the phosphite, a MeOꢀ
or tBuꢀgroup was incorporated in the parent structure. Alternatively, as an electronꢀwithdrawing
substituent Cl was chosen.
The influences of these substituents in the periphery of salicylic acid (Chart 1) on the
hydrolysis stability in comparison to L2 are summarized in Table 1. Under the conditions
detailed above it can be concluded that the electronꢀdonating MeOꢀgroup in the para position to
the acyl moiety marginally improves the stability of the phosphite towards water. In contrast, the
electronꢀwithdrawing Cl group in the same position destabilizes L2 significantly. A clear
improvement of hydrolysis resistance has been realized by application of the corresponding 3,5ꢀ
diꢀtꢀbutyl substituted salicylic acid derivative.
ACS Paragon Plus Environment
21

ACS Catalysis
Page 22 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Chart 1. Derivatives of L2
a
Table 1. Hydrolysis of L2 and its derivatives
Derivative
L2
t1/2 (h)
17.8
L2-OMe
L2-tBu
L2-Cl
21.8
144.4
After 1 h only 58.7%
of ligand left
a
[
L] = 0.049 M, 10 eq. of water, r.t.
3.5. Hydrolysis of L2 in Rhꢀcomplex (RhL2)
Treatment of racemic L2 with Rh(CO) (acac) in a molar ratio of 2:1 in toluene led to the
2
3
1
complex Rh(L2) (acac) (RhL2). The P NMR spectrum of the complex displayed two sets of
2
doublets centered at 122.0 and 122.4 ppm with couplings JP,Rh = 305 and 307 Hz, as expected for
a mixture consisting of the meso isomer and the R,Rꢀ and S,Sꢀenantiomers.
3
1
The time dependent in situ P NMR spectra of the hydrolysis of RhL2 at r.t. in the presence
of 10ꢀfold amount of water are depicted in Figure 10. The figure illustrates that besides the
phosphite peaks of decreasing intensity there is a new set of doublets appearing at δ = 124.7 ppm
with a splitting of 301 Hz. Considering its typical chemical shift and coupling constant, the
doublet can be attributed to the PADE of L2 that has been coordinated to Rh. This observation
ACS Paragon Plus Environment
22

Page 23 of 56
ACS Catalysis
1
2
3
4
5
6
7
8
9
gives proof that the degradation product of the monoꢀphosphite is able to coordinate to the metal.
Simultaneously, the coordination retains the PADE form in equilibrium with its corresponding
HASPO. Meanwhile, a very broad signal appeared around δ = 80 ppm. It might stem from a
trivalent form of L2DP1 with strong molecular dynamics. A new signal at δ = 1.1 ppm without
PꢀH splitting was found additionally, however, its origin is still unclear up to now. A resonance
assigned to phosphorous acid could also be observed when the reaction was carried out at 90 °C
for 5 h. The change of the concentration of the phosphite indicates that the hydrolysis obeys a
first order kinetic with respect to L2. The halfꢀlife time is 173 h, which is much longer than the
halfꢀtime observed in the hydrolysis of the “free” ligand. This result gives clear evidence that
coordination of the monodentate ligand to Rh stabilizes it towards hydrolysis. This stabilizing
effect is manifested in the crystal structure of RhL2 (Figure 11) by a shortening of two of the
three PꢀO bonds of L2 in comparison to the uncoordinated phosphite L2.
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
1
Figure 10. In situ P NMR spectra of hydrolysis of RhL2 with 10 eq. of water at r. t. (phosphite
region).
ACS Paragon Plus Environment
23

ACS Catalysis
Page 24 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 11. Molecular structure of RhL2 (thermal ellipsoids at the 30 % probability level;
hydrogen atoms have been omitted for clarity). Selected bond lengths [Å] and angles [°]: P1ꢀO3
1
.632(1), P1ꢀO4 1.614(1), P1ꢀO5 1.594(1), P2ꢀO8 1.633(1), P2ꢀO9 1.607(1), P2ꢀO10 1.601(1),
Rh1ꢀP1 2.1335(4), Rh1ꢀP2 2.1322(4), O3ꢀP1ꢀO4 100.16(6), O4ꢀP1ꢀO5 97.14(6), O3ꢀP1ꢀO5
01.78(6), O8ꢀP2ꢀO9 100.85(6), O9ꢀP2ꢀO10 96.76(6), O8ꢀP2ꢀO10 102.02(6).
1
3
.6. Hydrolysis of the “right arm” of L1
The “right arm” of L1 is represented by the model phosphite L3. (Scheme 7) The
corresponding hydrolysis product L3DP1 has been prepared and structurally analyzed previously
1
8
in our group. In its PADEꢀisomer this degradation product was able to act as a modifying
ligand in the rhodium catalyzed hydroformylation. The superior stabilizing effect of the biphenol
in contrast to salicylic acid is demonstrated by comparing stabilities of the corresponding
monophosphites, i.e. the acylphosphite L2 and the phenolphosphite L3. Much higher resistance
toward hydrolysis of the latter has been found. The phosphite remained almost intact in the
3
1
presence of 10 equivalents of water at 90 °C. In total, three products were detected by in situ P
NMR spectroscopy, which were assigned to the HASPO of L3 (L3DP1), a biphenol
phosphorous acid monoester (L3DP2), and phosphorous acid, respectively (Scheme 7).
According to this result, the hydrolysis pathway of L3 can be unambiguously revealed. Unlike
ACS Paragon Plus Environment
24

Page 25 of 56
ACS Catalysis
1
2
3
4
5
6
7
8
9
L2 its exocyclic PꢀO bond is firstly cleaved selectively followed by cleavage of one of the PꢀO
bonds in the HASPO. A clear autocatalysis process can be concluded from the concentration
profile of the components of the hydrolysis of L3. As the amount of H PO increased the
3
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
decomposition of the phosphite was dramatically accelerated. (results in supporting information)
Scheme 7. Hydrolysis pathway of L3
3.7. Hydrolysis of Rhꢀcomplex of L1 (RhL1)
Rh(acac)(L1) was hydrolyzed under the same conditions as employed for L1. (Figure 12)
Surprisingly, the bidentate diphosphite was decomposed faster after coordination to the metal.
As shown in Figure 13 decrease of the halfꢀlife time of L1 was apparent, especially for the trial
at 120 °C, where the destabilizing effect of Rh is remarkable. This finding is in strong contrast to
that what has been observed for the monophosphite L2. Irrespective of this observation a similar
31
degradation pattern for the hydrolysis of both ligands was found. The in situ P NMR spectra
showed a doublet signal (δ = 130.5 ppm) with a splitting of 293 Hz appearing in the phosphite
region as well, together with another doublet at 163.9 ppm with a splitting of 248 Hz. With
ACS Paragon Plus Environment
25

ACS Catalysis
Page 26 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
respect to the observations with RhL2, these signals can be clearly attributed to the hydrolysis
V
III
products which coordinate to Rh after P to P tautomerization e.g. HASPOs of L2 and L3.
Unfortunately, under our conditions no species having a PꢀH bond could be detected. Only a
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
31
singlet at 0.6 ppm in P NMR spectrum was observed.
31
Figure 12. In situ P NMR spectra of hydrolysis of RhL1 (100 eq. of water, 120 °C), where two
III
new formed doublets appear in P region.
OMe
OMe
OMe
OMe
O
O
O
O
tBu
tBu
tBu
O
O
P
O
O
P
O
tBu
O
tBu
P
tBu
P
tBu
tBu
O
O
tBu
O O
tBu
MeO
OMe
MeO
OMe
L4
L1
a
Table 2. Hydrolysis of L1 and L4
Derivative
t1/2 / 90 °C (h) t1/2 / 120 °C
h)
(
L1
60.3
24.5
ACS Paragon Plus Environment
26

Page 27 of 56
ACS Catalysis
1
2
3
4
5
6
7
8
9
L4
91.2
45.6
a
[
L] = 0.0245 M, 100 eq. water.
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
t(1/2)ꢀ
60.26
54.57
8
0
0
0
0
0
6
4
2
24.49
4.65
L1 (90°C)
L1
RhL1
RhL1
(120°C) (90°C) (120°C)
Figure 13. Comparison of halfꢀlife times of L1 and RhL1 at different hydrolysis temperatures.
a
b
Table 3. Hydroformylation of nꢀoctenes
c
d
ꢀ1
Entry
1
ligand
Rh/L/nꢀoctenes Aldehydes
%)
nꢀselectivity
(%)
kobs (min )
(
L1
L4
1:5:2168
1:5:2131
98
88
60.8
0.13363
0.06568
2
70.7
a
ꢀ3
[
Rh] = 0.77 x 10 mol/l; T = 120 °C, p = 20 bar total, t = 4 h, solvent propylene carbonate.
b
3
.3% 1ꢀoctene, 48.4% Z/Eꢀ2ꢀoctene, 29.2% Z/Eꢀ3ꢀoctene, 16.4% Z/Eꢀ4ꢀoctene, 2.1% skeletal
C8ꢀolefinic isomers, 0.6% nꢀoctane. determined by GC with toluene as internal standard.
pseudoꢀfirstꢀorder rate constant derived from exponential curve fitting of gas consumption
versus time.
c
d
a
Table 4. Hydroformylation of 2ꢀpentene
b
c
d
ꢀ1
Entry
ligand
Rh/L/2ꢀ
pentene
Aldehydes
(%)
nꢀselectivity
kobs (min )
(%)
1
2
L1
L4
1:5:666
1:5:645
100
100
71.4
81.4
0.1769
0.0456
ACS Paragon Plus Environment
27

ACS Catalysis
Page 28 of 56
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
a
ꢀ3
[
Rh] = 1.07 x 10 mol/l; T = 120 °C, p = 20 bar total, t = 1.5 h, solvent propylene carbonate.
b
c
determined by GC with toluene as internal standard. share of nꢀhexanal among aldehydes
formed. pseudoꢀfirstꢀorder rate constant derived from exponential curve fitting of gas
consumption versus time.
d
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
.8. Optimization of the structure of L1: L4
As found above an ortho placed tꢀbutyl group in the acylphosphite structure has a stabilizing
effect on the hydrolysis resistance of the isolated “left” part of the bidentate ligand. In order to
prove, if this conclusion is also valid for the corresponding bidentate ligand, we synthesized
ligand L4. As demonstrated in Table 2, the tꢀbutyl decorated diphosphite showed a superior
stability in comparison to L1 at different temperatures. Notably, the stabilizing effect became
more significant when the temperature was increased from 90 °C to 120 °C, which is especially
important for the conditions of technical application.
3
.9. Hydroformylation of nꢀoctene and 2ꢀpentene: A comparison of L1 and L4
In order to elucidate which consequences arise from the application of a more stable ligand on
the catalytic performance, we compared L1 and L4 in the hydroformylation of nꢀoctenes and 2ꢀ
pentene, respectively. Results are listed in Table 3 and Table 4. Interestingly, the results are
dependent on the substrate used. L1 induced higher yields of aldehydes than L4 in the
hydroformylation of nꢀoctenes. Noteworthy, the selectivity towards the formation of the desired
nꢀaldehyde is superior in comparison with the less decorated ligand.
A similar result was obtained, when 2ꢀpentene was submitted to the reaction. The tertꢀbutylꢀ
decorated ligand L4 produced a less active but more selective catalyst in comparison to the
parent ligand L1.
4
. Conclusions
ACS Paragon Plus Environment
28

Page 29 of 56
ACS Catalysis
1
2
3
4
5
6
7
8
9
The complete hydrolysis pathway of the bidentate hybrid diphosphite L1 was clarified. The
diphosphite is of interest for the industrial application in the rhodium catalyzed hydroformylation
of olefins. In situ NMR spectroscopy gave evidence that salicylic acid is firstly released from the
structure of L1. The resulting phosphorous acid monoester L1DP1 undergoes hydrolysis
unselectively. By comparing the stability of the two monophosphite units in L1, it can be
concluded that the “right arm” of the ligand is stabilized by its phenolphosphite moiety. In
contrast, the acylphosphite L2 is quite sensitive towards water. Hydrolysis of L2 proceeds
unselective and gives salicylic acid and phosphorous acid as main final products. The
spectroscopic results were supported by DFT calculations. Both degradations of L1 and L2
follow firstꢀorder kinetics with respect to the phosphite with an excessive amount of water.
Salicylic acid decorated with sterically demanding tꢀbutyl groups in 3ꢀ and 5ꢀposition gives a
ligand (L4) of much higher hydrolysis stability compared with unsubstituted L1. This is in line
with the fact that the overwhelming amount of phosphites suggested as ligands in rhodium
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
7
catalyzed hydroformylation bear tꢀbutyl groups in ortho position to the phosphite unit.
Obviously, this effect is general also for acylphosphites with smaller ring in the backbone.
Interestingly, L4 allows also the achievement of a higher nꢀregioselectivity in the
hydroformylation of nꢀoctenes and 2ꢀpentene. Unfortunately, inferior reaction rates resulted with
L4, but this can be counteracted in a technical scale either by longer reaction times or by an
increase of the temperature.
One of the most striking results of our study is that coordination to rhodium has different
impacts on the hydrolysis stability of bidentate and monodentate ligands. The monodentate
ligand L2 is stabilized in its Rh complex, whilst in strong contrast, the bidentate ligand L1 is
faster hydrolyzed after coordination to Rh. It will be investigated in next studies, whether this
ACS Paragon Plus Environment
29