Intramolecular Alkene Hydroamination with Hybrid Catalysts Consisting of a Metal Salt and a Neutral Organic Base

Article abstract of DOI:10.1002/ejic.202000671

Hybrid catalysts consisting of alkaline earth iodides (AeI₂) and the Schwesinger base ^tBuP4 catalyse the intramolecular alkene hydroamination of H₂C=CHCH₂CR₂CH₂NH₂ [CR₂

Full text of DOI:10.1002/ejic.202000671

European Journal of Inorganic Chemistry
Accepted Article
Title: Intramolecular Alkene Hydroamination with Hybrid Catalysts
Consisting of a Metal Salt and a Neutral Organic Base
Authors: Philipp C. Stegner, Christian A. Fischer, D. Thao Nguyen,
Andreas Rösch, Johanne Penafiel, Jens Langer, Michael
Wiesinger, and Sjoerd Harder
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Eur. J. Inorg. Chem. 10.1002/ejic.202000671
Link to VoR: https://doi.org/10.1002/ejic.202000671
01/2020

10.1002/ejic.202000671
European Journal of Inorganic Chemistry
Intramolecular Alkene Hydroamination with Hybrid Catalysts Consisting of a
Metal Salt and a Neutral Organic Base
Philipp C. Stegner, Christian A. Fischer, D. Thao Nguyen, Andreas Rösch, Johanne Penafiel, Jens Langer,
Michael Wiesinger and Sjoerd Harder*
Inorganic and Organometallic Chemistry, Friedrich-Alexander-Universität Erlangen-Nürnberg,
Egerlandstraße 1, 91058 Erlangen, Germany.
Mr. P. C. Stegner, Mr. C.A. Fischer, Dr. D. T. Nguyen, Mr. A. Rösch, Dr. J. Penafiel, Dr. J. Langer. Mr. M.
Wiesinger, Prof. Dr. S. Harder
E-mail: sjoerd.harder@fau.de
Twitter: @harder_research
Key words: Calcium; intramolecular alkene hydroamination; Schwesinger base; enantioselective catalysis
Key topic: Hybrid Calcium Catalysts
Table of Contents: Influences of metal,
ligand, and organic base in hybrid
catalysts for the intramolecular alkene
hydroamination
investigated.
have
been
Abstract: Hybrid catalysts consisting of alkaline earth iodides (AeI2) and the Schwesinger base tBuP4
catalyse the intramolecular alkene hydroamination of H2C=CHCH2CR2CH2NH2 (CR2 = CPh2, C(CH2)5, CMe2).
Activities decrease along the row: Ca > Sr >> Mg > Ba. Hybrid catalysts consisting of tBuP4 and ZnI2, AlI3,
FeCI3 or NaI were found to be fully inactive. Also the hybrid catalyst tBuP3/CaI2 was not active which means
that the base strength of the non-nucleophilic organic base must be higher than that of tBuP3 (pKa BH+ =
38.6). Combinations of tBuP4 with CaX2 (X = Cl, Br, OiPr, OTf, NTf2) were found to be fully inactive which
may in part be explained by poor solubility. The hybrid catalysis method is therefore limited to the
combination tBuP4/CaI2 but the iodide ligands may be partially or fully replaced by chiral ligands. Chiral
modifications of the hybrid catalysts gave in intramolecular alkene hydroamination ee values up to 33%.
1
This article is protected by copyright. All rights reserved.

10.1002/ejic.202000671
European Journal of Inorganic Chemistry
Introduction
Whereas the majority of catalysts in the fine chemical production are still based on the more noble and
expensive transition metals (Pd, Pt, Ir, Rh, Ru), there is an increasing tendency towards the use of more
earth-abundant metals (Mn, Fe, Ni, Cu, Zn).[1] This trend is not limited to transition metals but also includes
catalysis with earth-abundant main group metals (Ca, Mg, Al).[2,3] This popularity is fuelled by the many
advantages of using non-poisonous, earth-abundant and cheap metals like Mg and Ca that, unlike precious
metals or lanthanides, are not subjected to monopolies and therefore their price is stable. The fast
growing area of early main group metal catalysis has also been extended to the heaviest group 2 metals
Sr and Ba.[4]
In contrast to transition metal catalysis, the alkaline earth (Ae) metals cannot easily switch between
oxidation states and do not possess d-orbitals for substrate activation. Catalysis with a group 2 metal like
Ca is purely based on polar reaction mechanisms and Lewis-acid activation. A differentiation between
″organocalcium″ and ″Ca2+ ion″ catalyzed reactions should be made. The first class of catalysts are polar
organometallics that are not strictly limited to Ca-C bound compounds but also involve Ca-O, Ca-N, Ca-H,
Ca-P or Ca-Si bound species. Heteroleptic catalysts, LCaR, generally consist of a reactive R group and a
passive spectator ligand L but also homoleptic catalysts like CaR2 are known. The second class of catalysts
relies solely on the Lewis-acidity of “naked” Ca2+ ions, e.g. in Ca2+(NTf2ˉ)(PF6ˉ),[5] and is not only
fundamentally different but also targets different transformations.
Previously, we investigated the role of the metal in double bond functionalization with organometallic
catalysts of type LCaR.[6] Catalytic hydroamination of highly activated C=O or C=N double bonds in
isocyanates or carbodiimides does not need the presence of the Ca2+ (it could be even substituted for
Me4N+). However, in the intramolecular hydroamination of unactivated C=C bonds, the metal cation Ca2+
is essential. This was further confirmed by attempted alkene hydroamination with Schwesinger’s, metal-
free, base tBuP4 (Scheme 1) which in its protonated form has a pKa value of 42.7.[7] Under forcing conditions
this very strong organic base is able to deprotonate the aminoalkene H2NCH2CPh2CH2CH=CH2 but instead
of ring closure to the pyrrolidine product clean formation of 1,1-diphenyl-2-ethylethene was observed
(Scheme 1). Addition of catalytic quantities of CaI2, however, led even under mild conditions to fast ring
closure. We proposed that the role of Ca2+ is twofold: (1) substrate-Ca2+ coordination acidifies the NH2
group enabling facile deprotonation and (2) alkene-Ca2+ coordination activates the C=C bond for
nucleophilic attack. Such hybrid catalysts, consisting of an organic strong base and a metal salt, are
reminiscent of Frustrated Lewis Pair (FLP) catalytic systems which also consist of a base and an acid that
work in concert.[8]
2
This article is protected by copyright. All rights reserved.

10.1002/ejic.202000671
European Journal of Inorganic Chemistry
These hybrid catalysts have clear advantages: they are simple to prepare and the availability of a large
range of potential organic bases and Lewis acidic metal salts forecasts numerous new combinations. We
now present comprehensive investigations on such hybrid catalysts and discuss variation of the organic
base, variation of the metal salt and metal ligand effects. We also report on first applications of this
method in enantioselective intramolecular alkene hydroamination.
Scheme 1. Left: Conversion of an aminoalkene substrate with the organic base tBuP4 under harsh
conditions (80 °C, 5 days, 98%). Right: Catalytic conversion of an aminoalkene substrate with a hybrid
catalyst consisting of CaI2 and the organic base tBuP4 under mild conditions (20 °C, 2 h, 99%).
3
This article is protected by copyright. All rights reserved.

10.1002/ejic.202000671
European Journal of Inorganic Chemistry
Results and Discussion
Structure of the hybrid catalyst
Since the tBuP4 base is not only a very strong Brønsted base but also a potential Lewis base, the interaction
of the hybrid catalyst building blocks, CaI2 and tBuP4, was investigated. Cocrystallization of both partners
was achieved from a mixture of benzene and N,N,N’,N’-tetramethylethylenediamine (TMEDA). The crystal
structure of the 1:1 adduct CaI2·(tBuP4) is shown in Figure 1. The Lewis base tBuP4 efficiently breaks down
the (CaI2)∞ salt. A monomeric CaI2 unit is bound in a tBuP4 pocket which acts as (3)N-donor with two
shorter Ca-N contacts and one rather long interaction (Ca-N2). The Ca-N5 bond of 2.320(2) Å is much
shorter than that to the P-N-P unit: Ca-N1 2.434(2) Å. This is clearly related to the obtuse P1-N1-P2 angle
of 143.6(2)°. Similarly as for silylethers, for which the very poor electron donor character of the Si-O-Si
unit also correlates to obtuse Si-O-Si angles, the P-N-P unit should be considered a poor donor.[9, 10] The
shortest Ca-N contact of 2.320(2) Å is found for interaction with the P=N(tBu) unit which is also the site of
highest Brønsted basicity that is most likely to be protonated. Since CaI2·(tBuP4) crystallized from benzene
in the presence of a tenfold excess of TMEDA, it can be concluded that the Schwesinger base is a stronger
Lewis base than TMEDA. An adduct with TMEDA is also known from work by the Westerhausen group:
CaI2·(TMEDA)2.[11] The Ca-N distances in this complex range from 2.562(5)-2.570(5) Å.
Figure 1. Crystal structure of CaI2·(tBuP4). Selected
bond distances (Å) and angles (°): Ca-I1 3.0133(7),
Ca-I2 3.0472(7), Ca-N1 2.434(2), Ca-N2 2.780(2),
Ca-N5 2.320(2), I1-Ca-I2 97.94(2), P1-N1-P2
143.6(2), P2-N5-C7 126.8(2), P2-N10-P4 152.3(2),
P2-N6-P3 147.7(2).
Variation of the metal salt
A selection of simple metal salts has been investigated regarding their performance as the Lewis acidic
component in hybrid catalysts for the intramolecular alkene hydroamination. This study focussed mainly
on metal iodide salts which generally show higher solubility in organic solvents. In combination with the
4
This article is protected by copyright. All rights reserved.

10.1002/ejic.202000671
European Journal of Inorganic Chemistry
strong organic Schwesinger base tBuP4, the aminoalkene H2NCH2CPh2CH2CH=CH2 was converted to a
pyrrolidine. Conversion and reaction time are summarized in Table 1.
Catalysis in aliphatic solvents was hampered by the complete insolubility of the metal iodide but in
aromatic solvents all metal iodide salts dissolved in the presence of the aminoalkene, except for ZnI2 which
dissolved only after addition of a small amount of THF. This is in agreement with the first step in the
proposed mechanism, i.e. aminoalkene-metal coordination (Scheme 1). Additional evidence is provided
by the fact that the addition of solid CaI2 to an aminoalkene solution in C6D6 resulted in a downfield shift
of the NH2 protons from 0.69 ppm to 0.84 ppm, indicating Ca-substrate coordination. A combination of
the aminoalkene and tBuP4 base did not result in substrate deprotonation. However, immediately after
addition of solid CaI2, the protonated base tBuP4-H+ can be detected in the 31P-NMR-spectrum at room
temperature. By varying the base to calcium stoichiometry, a linear correlation between the amount of
tBuP4-H+ formed and the amount of CaI2 present was observed. Accordingly, any excess of Schwesinger
tBuP4 base with respect to calcium remained unprotonated. Further investigations were run with an
equimolar metal salt/ tBuP4 ratio.
In combination with tBuP4, all AeI2 salts were found to be catalytically active but full conversion within
4.5 h was only obtained with CaI2 (Table 1). The activities do not follow the trends in group 2 but decrease
in the order: CaI2 > SrI2 >> MgI2 >> BaI2. This observation is in accordance with previously published trends
in alkaline earth metal catalyzed hydroamination.[12-18] The superior performance of CaI2 may be explained
by two antiparallel trends in group 2. On the one hand, the reactivity of the Ae-N bond increases down
the group: Mg < Ca < Sr < Ba. On the other hand, the Lewis acidity of the metal increases in reverse order:
Ba > Sr > Ca > Mg. Therefore, Ca shows an ideal combination of relatively high reactivity and Lewis acidity.
Replacing the solvent from benzene to THF generally led to faster conversion (Table 1). This contrasts with
earlier results, showing a decelerating effect of the polar solvent THF, which was explained by competition
between metal-THF and metal-substrate coordination.[18]
To gain further insight in the influence of the metal, the metal salts ZnI2, AlI3, FeCI3 and NaI were also
tested as the Lewis acidic catalyst component. In both solvents, benzene and THF, no catalytic activity was
observed. Therefore, further investigations focussed on use of CaI2 or other Ca complexes.
5
This article is protected by copyright. All rights reserved.

10.1002/ejic.202000671
European Journal of Inorganic Chemistry
Table 1. Selected metal iodides in catalytic cooperative hydroamination for the substrate
H2NCH2CPh2CH2CH=CH2. Conditions: 10 mol% metal iodide, 10 mol% tBuP4, 20 °C, C6D6.
Entry Metal salt
Time [h]
4.5
Conversion [%]
1
2
3
4
5
MgI2
CaI2
SrI2
34
> 98
71
4.5
4.5
BaI2
ZnI2
36
< 10
0
36
Scheme 2. Selected phosphazene and cyclopropenimine bases tested in intramolecular alkene
hydroamination of H2NCH2CPh2CH2CH=CH2 (10 mol% CaI2, 10 mol% base, C6D6, 20 °C).
6
This article is protected by copyright. All rights reserved.

10.1002/ejic.202000671
European Journal of Inorganic Chemistry
Variation of the organic base
The organic base is crucial for deprotonation of the aminoalkene substrate. In order to establish the
minimum base strength, a variety of different non-nucleophilic bases was tested. It was found that bases
(B) with low base strength did not give any conversion: PtBu3 (pKa BH+ = 11.4), the NHC ligand IiPr (IiPr =
[HCN(iPr)]2C, pKa BH+ = 22.8) and the Verkade base (pKa BH+ = 29.3). This means that, despite acidification
of the aminoalkene by metal coordination, the base strength should be clearly higher than that of the
Verkade base. A further series of phosphazene and cyclopropenimine bases was tested (Scheme 2).
Smooth ring closure towards the pyrrolidine cyclization product was only observed for Schwesinger’s
phosphazene base tBuP4 (pKa BH+ = 42.7). Even the somewhat weaker base tBuP3 (pKa BH+ = 38.6) did not
show any conversion. The method is therefore limited to Schwesinger’s tBuP4 base.
Variation of the ligand
To shed light on the role of the ligands coordinated to calcium, the iodide ligands in CaI2 were replaced by
various ligands of low basicity. Hybrid catalysts consisting of CaX2 (X = Cl, Br, OiPr, OTf, NTf2) combined
with Schwesinger’s tBuP4 base were tested in the intramolecular alkene hydroamination of
H2NCH2CPh2CH2CH=CH2 (10 mol% catalyst, THF-d8, 20 °C). 1H NMR monitoring of these reaction mixtures
showed in all cases no conversion. For catalysts with CaCl2 and CaBr2 this might be attributed to the poor
solubilities of these salts. Catalysts with OTf, NTf2 or OiPr ligands, however, are largely soluble in THF-d8.
It is remarkable that hybrid catalysts with strongly Lewis acidic components like Ca(OTf)2 or Ca(NTf2)2 are
+
not active. Also a combination of Ca(NTf2)2 and [nBu4 ][PF6ˉ], which is a highly active Lewis acid catalyst,[16]
showed in combination with the tBuP4 base no activity in intramolecular alkene hydroamination. The
inactivity of strong Lewis acids with highly electron withdrawing OTf and NTf2 ligands complexes may be
explained by Ca-O interactions with the sulfonic oxygen atoms, reducing the ability of Ca2+ to bind the
aminoalkene. The good performance of CaI2 as the Lewis acidic component in hybrid catalysis may also
be explained by its unexpectedly high Lewis acidity. Recent Fluoride-Ion-Affinity (FIA) calculations on gas
phase AeI2 monomers show FIA values (kJ/mol) of: Mg 426, Ca 436, Sr 423, Ba 410.[20] The FIA of CaI2 was
found to be even higher than that of Ca[N(C6F5)2]2 (429 kJ/mol). Although aggregation will influence the
Lewis acidities of Ca salts, these values demonstrate the intrinsically high Lewis acidity of CaI2.
7
This article is protected by copyright. All rights reserved.

10.1002/ejic.202000671
European Journal of Inorganic Chemistry
As CaI2 was found to be the only active Lewis acidic component, we aimed to synthesize heteroleptic chiral
calcium iodide complexes of composition L*-Ca-I with the goal to investigate these in enantioselective
hybrid catalysis. The prerequisites of ligand L* are, apart from being chiral and optically active, its
capability to stabilize the heteroleptic complex by preventing ligand redistribution to the homoleptic
complexes L*2Ca and CaI2. The latter Schlenk equilibrium is known to lead to loss of enantiocontrol in Ca
catalysed alkene hydroamination.[21] It is also imperative that L* is not too basic which would cause ligand
loss by protonation. For this reason, the literature-known hydroxyl based ligand L1-H[22] with a pKa below
18 and a pendant chelating arm to stabilize the heteroleptic species was chosen (Scheme 3a). The chiral
alcohol, which is based on the optically active natural building block (-)-fenchone, was converted to its
potassium salt by reaction with KH in THF. Salt metathesis with CaI2 to give the target complex L1-CaI,
however, did not lead to the heteroleptic species. Despite a 50% excess of CaI2, only the homoleptic
complex (L1)2Ca was isolated. Exchanging the pyridyl pendant arm in L1 for a N,N-dimethylaminobenzyl
functionality (L2),[23] however, gave exclusively the heteroleptic complex which crystallized as the THF
solvate (L2)CaI·(THF)3. The latter complex is also in solution stable towards ligand exchange reactions:
NMR investigations on a solution of (L2)CaI·(THF)3 in C6D6 show even at +60 °C no indications for a Schlenk
equilibrium. Although very recent theoretical insights in the Schlenk equilibrium certainly led to a better
understanding,[24] it is hitherto still not possible to predict why one ligand leads to stable heteroleptic
complexes, while another similar ligand gives exclusively the homoleptic species. The crystal structures of
(L1)2Ca and (L2)CaI·(THF)3 (Figure 2) do also not give further indications.
Scheme 3. Syntheses of chiral Lewis acidic Ca complexes with (a) monoanionic ligands or (b) dianionic
ligands.
8
This article is protected by copyright. All rights reserved.

10.1002/ejic.202000671
European Journal of Inorganic Chemistry
Complex (L1)2Ca crystallized in the space group P212121 and, although the complex was crystallized from
THF solution, there is no coordinating solvent bound to Ca. The two N,O-chelating ligands completely
encapsulate the metal which is also partially due to Ca···Ph interactions. Compound (L2)CaI·(THF)3
crystalized in the same space group. The Ca metal exhibits an octahedral coordination in which the
negatively charged alkoxide and iodide ligands are in axial positions while the neutral THF ligands and the
pendant Me2N-arm are bound in the equatorial plane.
Figure 2. Crystal structures. (a) (L1)2Ca; selected bond distances (Å): Ca-O1 2.106(2), Ca-O2 2.110(2), Ca-
N1 2.486(2), Ca-N2 2.455(2), Ca···C27 3.142(3), Ca···C50 2.965(2). (b) (L2)CaI·(THF)3; selected bond
distances (Å): Ca-I 3.2144(7), Ca-N 2.527(2), Ca-O1 2.120(2), Ca-O2 2.416(3), Ca-O3 2.429(2), Ca-O4
2.425(3). (c) (L6)Ca·(THF)4; selected bond distances (Å): Ca-O1 2.253(2), Ca-O2 2.239(2), Ca-O3 2.393(2),
Ca-O4 2.376(2), Ca-O5 2.443(2), Ca-O6 2.394(2).
In combination with the strong base tBuP4, these chiral Ca complexes were tested as the Lewis acidic
component in hybrid catalysts for intramolecular alkene hydroamination. Although most of the chiral
complexes contain THF ligands, it should be realized that these are only loosely bound and can easily be
exchanged by aminoalkene substrates. Three standard aminoalkenes have been tested:
H2NCH2CR2CH2CH=CH2 (CR2 = CPh2, C(cyclohexane) or CMe2). Depending on the steric bulk of the
substituents inposition, the substrates show varying reaction times (Table 2). While smooth ring
closure was observed for the bulky diphenyl substituted derivative at room temperature, 60°C was
necessary for the cyclohexane-substituted substrate. In the case of the dimethyl-substituted derivative,
90°C is necessary for successfully cyclization to the corresponding heterocycle. This is in agreement with
the Thorpe-Ingold-effect which states that larger substituents accelerate intramolecular ring closure.[25]
9
This article is protected by copyright. All rights reserved.

10.1002/ejic.202000671
European Journal of Inorganic Chemistry
The enantioselectivities were determined by conversion of the chiral pyrrolidine products to
1
diastereomers by derivatization with O-acetyl mandelic acid and subsequent analysis by H NMR
spectroscopy. In case of insufficient peak separation, derivatization was carried out using Mosher chloride
and the optical purity was determined by integration of the corresponding diastereomeric 19F NMR
signals.
The homoleptic complex (L1)2Ca showed in combination with tBuP4 no activity. This is possibly due to the
inaccessibility of Ca for aminoalkene coordination. The more open heteroleptic complex, (L2)CaI·(THF)3,
catalysed cyclization of all three substrates in essentially quantitative yield with ee’s varying between 8
and 15%. The low values may be explained by poor stereocontrol of the chiral building blocks. Ligand
exchange reactions are, however, also not excluded. Although (L2)CaI·(THF)3 itself is stable towards ligand
exchange, possible heteroleptic intermediates on the catalytic pathway may be instable towards such
detrimental transformations. For those reasons, we became interested in Ca complexes with chiral,
bidentate, dianionic ligands like C2-symmetric BINOL (Scheme 3b).
A variety of literature-known (S)-BINOL ligands (L3-L6) with larger substituents in the 3,3’-positions has
been prepared and was converted by reaction with Ca(OiPr)2 to the corresponding Ca complexes (Scheme
3b). Although ligands L3-L5 carry larger or potentially coordinating substituents, in all cases only undefined
coordination polymers that are even insoluble in polar solvents like THF were isolated. This clearly shows
the challenge to design ligands that saturate the coordination sphere of large Ca2+ ions. Only in case of
larger Ph3Si-substituents a defined complex was isolated. Complex (L6)Ca·(THF)4 crystallizes in space
group P21 with four THF ligands at Ca (Figure 2c). By exposing the compound to vacuum, the total number
of THF ligands can only be reduced to three. The torsion angle (N)C-C-C-C(N) determines the twist
between the naphthyl rings. In the unmetallated BINOL it measures 76.7(3)°[26] while in the Ca complex it
is clearly enlarged to 86.9(3)°. This can be explained by the large size of the Ca2+ ion (ionic radius for 6-
coordinate Ca2+: 1.00 Å).[27]
Complex (L6)Ca·(THF)4 gave in combination with tBuP4 full conversion in the catalytic ring closure of
aminoalkenes (Table 2). The enantioselectivities, improve slightly from 8%-15% to 23%-33%. This is a
consequence of the catalyst design, based on a dianionic ligand. This leads to suppression of ligand
exchange processes and a higher degree of stereo control. The general trend is that enantioselectivities
increase with steric bulk of the CR2 groups in the aminoalkene. The temperature has no effect on the
enantioselectivities and only influences the conversion times. The generally observed poor stereocontrol
10
This article is protected by copyright. All rights reserved.

10.1002/ejic.202000671
European Journal of Inorganic Chemistry
in Ca catalysed reactions may be explained by the long and dynamic Ca-ligand bonds. Therefore,
enantioselective catalysis with chiral Ca catalysts remains challenging.
Table 2. Intramolecular alkene hydroamination of H2C=CHCH2CR2CH2NH2 (CR2 = CPh2, C(CH2)5 or CMe2)
with hybrid catalysts consisting of tBuP4 (10 mol%) and a Lewis acidic Ca complex (10 mol%) in C6D6
(conversion determined by 1H NMR).
Product
[Ca]
Conv. [%]
>99
T [°C]
20
60
20
60
20
90
90
90
60
60
60
T [h]
5
ee [%]
(L2)CaI·(THF)3
(L2)Ca·(THF)4
(L6)Ca·(THF)4
(L6)Ca·(THF)4
CaI2
15
15
33
33
-
>99
1
>99
5
>99
1
>98
4.5
18
24
24
5
(L2)CaI·(THF)3
(L6)Ca·(THF)4
CaI2
>98
8
>97
26
-
>97
(L2)CaI·(THF)3
(L6)Ca·(THF)4
CaI2
>98
8
>98
5
23
-
>98
4.5
Conclusions
We demonstrated that the combination of alkaline earth iodides (AeI2) and the Schwesinger base tBuP4
gives rise to a cooperative reactivity that catalyses the intramolecular hydroamination of different
11
This article is protected by copyright. All rights reserved.

10.1002/ejic.202000671
European Journal of Inorganic Chemistry
aminoalkenes: H2C=CHCH2CR2CH2NH2. Not only the activated aminoalkene substrate (CR2 = CPh2) but also
more challenging aminoalkenes could be converted; e.g. CR2 = C(alkyl)2.
The influence of the metal was investigated and activities decrease along the row: Ca > Sr >> Mg > Ba. The
good performance of Ca may be explained by an ideal combination of relatively high reactivity (increasing
from Mg < Ca < Sr < Ba) and high Lewis acidity (increasing from Ba < Sr < Ca < Mg). Other metals salts, like
ZnI2, AlI3, FeCI3 and NaI, were found to be fully inactive in combination with tBuP4. Further research
therefore concentrated on hybrid catalysts with Lewis acidic Ca compounds.
The strength of the non-nucleophilic base was varied and it was found that the method is limited to use
of the highly basic tBuP4 base (pKa BH+ = 42.7). Since the tBuP3 base (pKa BH+ = 38.6) gave no conversion,
the base strength should be: pKa BH+ > 38.6.
Variation of the ligand X in CaX2 was also investigated. Combinations of tBuP4 with CaX2 (X = Cl, Br, OiPr,
OTf, NTf2) were found to be fully inactive. This may in part be due to poor solubility. This limits the hybrid
catalysis method to the combination tBuP4/CaI2. It was, however, found that the iodide ligands may be
partially or fully replaced by chiral ligands. These chiral modifications of the hybrid catalysts gave in
intramolecular alkene hydroamination ee values up to 33%. The poor enantioselecivities for these chiral
Ca catalysts may find their origin in the generally rather long and weak bonds to Ca which makes the three-
dimensional structures dynamic and undefined. Enantioselective catalysis with Ca catalysts is therefore a
challenge that still needs to be solved.
Experimental Section
General Experimental Procedures
All experiments were carried out in flame dried glassware under N2 using standard Schlenk techniques or
a glovebox. All solvents were degassed and subsequently dried over an aluminum oxide column, except
for THF which was dried over Na and redistilled. Anhydrous AeI2 (Ae = Ca, Sr, Ba) salts were synthesized
by reacting the metals with I2 in dry THF, removing excess of metal by filtration and evaporation of all
solvents under high vacuum at 60 °C. In the synthesis of MgI2, 1,2-diiodoethane instead of I2 has been
used. The salts ZnI2, CaBr2, CaCl2, Ca(OiPr)2, [nBu4N+][PF6ˉ], AlI3, FeCl3 and NaI and Ca(NTf2)2 were
purchased by Sigma Aldrich and thoroughly dried under high vacuum at 60 °C. Ca(OTf)2 was prepared from
CaCO3 and HOTf in toluene and thoroughly dried under high vacuum. Trifluoromethanesulfonic acid
(HOTf) was purchased from Fluorochem. The N-heterocyclic carbene, 1,3-diisopropyl-1H-imidazol-3-ium-
12
This article is protected by copyright. All rights reserved.

10.1002/ejic.202000671
European Journal of Inorganic Chemistry
2-ide,[28] as well as the strong bases G3, GC2 and GP2 were synthesized according to published
protocols.[29] The phosphazene bases tBuP4 and tBuP3 were synthesized according to the procedure of
Schwesinger et al.[30] The P-bases P(tBu)3, Verkade base and tBuP2 were purchased from Sigma Aldrich. The
amino-pentene substrates 2,2-dimethyl-pent-4-enylamine, 2,2-diphenyl-pent-4-enylamine and (1-allyl-
cyclohexyl)-methylamine were prepared using literature procedures.[31] Ligands L1-H,[22] L2-H,[23] L3-H2,[32]
[34]
[35]
L4-H2,[33] L5-H2 and L6-H2 were prepared according to literature known procedures. Unspecified
starting materials and research chemicals were obtained from commercial suppliers where appropriate
and used without further purification. NMR spectra were measured on Bruker Avance III HD 400 MHz and
Bruker Avance III HD 600 MHz spectrometers. Chemical shifts (δ) are given in ppm (parts per million)
values, coupling constants (J) in Hz. For describing signal multiplicities common abbreviations were used:
s (singlet), d (doublet), t (triplet), q (quartet), m (multiplet), dm (doublet of multiplet) and br (broad).
Spectra were referenced due to solvent residual signals.[36] Elemental analysis was performed with a
Hekatech Eurovector EA3000 analyzer. All crystal structures have been measured on a SuperNova
(Agilent) diffractometer with dual Cu and Mo microfocus sources and an Atlas S2 detector.
Crystallographic data have been deposited with the Cambridge Crystallographic Data Centre as
supplementary publication numbers: 2015270 for tBuP4/CaI2, 2015271 for (L1)2Ca, 2015272 for
(L2)CaI·(THF)3 and 2015273 for (L6)Ca·(THF)4.
Synthesis of CaI2(tBuP4): CaI2 (9.3 mg, 31.6 µmol, 1 eq.) was suspended in benzene (500 µL). Subsequently
tBuP4 (20.1 mg, 31.6 µmol, 1 eq.) and TMEDA (48 µL, 316 µmol, 10 eq.) were added. The suspension was
heated to 80°C for 2h and the hot solution was filtered through a pre heated microglass fiber filter. The
clear solution was put in a thermostat at 80°C and slowly cooled to room temperature. Within 24h
formation of crystalline materials was observed. The crystals were isolated, washed with n-pentane
1
(2x 300 µL) and dried under vacuum. Yield: 10 mg, 10.8 µmol, 34 %. H NMR (600 MHz, C6D6, 25°C): δ =
2.53 (d, 3JPH = 10.0 Hz, 54H, CH3), 1.29 (s, 9H, C(CH3)3) ppm. 31P NMR (243 MHz, C6D6, 25°C): δ = +13.46 (d,
JPH = 49.0 Hz, P=NC(CH3)3), -22.80 (d, JPH = 49.0 Hz, P(N(CH3)2)3) ppm. 13C{1H} APT NMR (151 MHz, C6D6,
25°C): δ = 49.4 (s, C0), 37.6 (s, CH3), 31.9 (s, C(CH3)3) ppm. Anal. Calcd. for C22H63CaI2N13P4 (MW = 927.62
g/mol): C, 28.49; H, 6.85; N, 19.63. Found: C, 28.82 H, 6.85; N, 19.78. Note: Similar attempts using tBuP3,
tBuP2 and cyclopropenimine bases (Scheme 2) failed to yield any crystalline products.
Synthesis of (L1)2Ca: A Schlenk flask was charged with a slurry of KH (50 mg, 1.25 mmol, 1.5 equiv.) in THF
(6 mL) and ligand L1-H (300 mg, 0.78 mmol, 1 equiv.). The suspension was stirred at room temperature
for 24 h. Subsequently excess potassium hydride was removed by filtration. Solid CaI2 (441 mg, 1.25 mmol,
13
This article is protected by copyright. All rights reserved.

10.1002/ejic.202000671
European Journal of Inorganic Chemistry
1.5 equiv.) was added and the reaction mixture was stirred for another 24h. The suspension was filtered
over Celite and the solvent was removed under reduced pressure. The red residue was extracted with
toluene (8 mL) and filtered over Celite. The solution was concentrated to about 50% of its volume by
removal of solvent under vacuum and the remaining red solution was stored at -30°C. The crystals, formed
over the course of 1 week, were collected, washed with n-hexane (2x 1 mL) and dried under high vacuum.
Yield: 250 mg, 0.31 mmol, 80 %. Crystals suitable for X-ray analysis were grown by slow evaporation of a
dilute toluene solution. 1H NMR (600 MHz, C6D6, 25°C): δ = 7.78 (d, 3JHH = 8.2 Hz, 1H, CHAr), 7.60 (d, 3JHH
7.6 Hz, 2H, CHAr), 7.16-7.13 (m, 3H, CHAr), 7.02-6.97 (m, 2H, CHAr), 6.93-6.85 (m, 2H, CHAr), 6.86 (d, 3JHH
=
=
7.6 Hz, 1H, CHAr), 6.75 (dd, 3JHH = 9.6 Hz, 4JHH = 1.5 Hz, 1H, CHAr), 2.76-2.72 (m, 1H), 2.50 (d, 3JHH = 9.4 Hz,
1H), 1.96-1.91 (m, 2H), 1.55-1.49 (m, 1H), 1.43 (dd, 3JHH = 9.7 Hz, 4JHH = 1.5 Hz, 1H), 0.96 (s, 3H, CH3), 0.94
(dd, 3JHH = 12.1 Hz, 4JHH = 4.3 Hz, 1H), 0.89 (s, 3H, CH3), 0.62 (s, 3H, CH3) ppm. 13C{1H} APT NMR (151 MHz,
C6D6, 25°C): δ = 168.4 (C0, Ar), 154.6 (C0, Ar), 153.6 (C0, Ar), 140.5 (C0, Ar), 137.2 (C0, Ar), 137.2 (CH, Ar),
132.9 (CH, Ar), 130.7 (CH, Ar), 129.4 (CH, Ar), 129.2 (CH, Ar), 126.3 (CH, Ar), 126.1 (CH, Ar), 125.3 (CH, Ar),
124.0 (CH, Ar), 117.8 (CH, Ar), 88.4 (C0-O), 55.4, 51.3, 44.8, 41.4, 38.5, 31.8, 25.3, 23.2, 18.0 ppm. Anal.
Calcd. for C54H56CaN2O2 (MW = 805.13 g/mol): C, 80.56; H, 7.01; N, 3.48. Found: C, 80.21; H, 7.12; N, 3.45.
Synthesis of Compound (L2)CaI·(THF)3: Ligand L2-H (0.50 g, 1.74 mmol) and potassium hydride (0.12 g,
2.99 mmol) were suspended in THF (10 mL). The suspension was stirred at 60 °C for 18 hours. Excess KH
was removed by centrifugation and CaI2 (0.82 g, 1.74 mmol) was added to the clear mother liquor. After
stirring for two days at room temperature, KI was removed by centrifugation and the solution was filtered
through a microglass fiber filter. The clear mother liquor was evaporated to dryness under high vacuum,
yielding an off-white solid. The product was extracted with toluene (2 x 10 mL) and filtered through a
microglass fiber filter. The solvent was evaporated under high vacuum and the off-white raw product was
recrystallized from a mixture of hexane (4 mL) and THF (7 mL) yielding a white crystalline solid. Yield: 0.43
1
3
g, 0.64 mmol, 37 %. H NMR (400 MHz, THF-d8, 25°C): δ = 7.59 (d, 3JHH = 8.0 Hz, 1H, CHAr), 7.09 (d, JHH
7.8 Hz, 4JHH = 1.6 Hz, 1H, CHAr), 7.00 (td, 3JHH = 7.6 Hz, 4JHH = 1.6 Hz, 1H, CHAr), 6.90 (d, 3JHH = 7.3 Hz, 4JHH
1.2 Hz, 1H, CHAr), 4.70 (d, 3JHH = 10.4 Hz, 1H, NCH2), 3.58 (m, 12H, THF), 2.81-2.76 (m, 1H), 2.58 (d, 3JHH
=
=
=
10.4 Hz, 1H, NCH2), 2.55 (broad s, 3H, NCH3), 2.46 (broad s, 3H, NCH3), 2.25 (d, 3JHH = 12.2 Hz, 1H), 1.73 (m,
12H, THF), 1.61 (m, 1H), 1.52-1.51 (m, 1H), 1.23 (s, 3H, CH3), 1.16 (d, 3JHH = 11.6 Hz, 2H), 1.12 (s, 3H, CH3),
0.79 (td, 3JHH = 12.0 Hz, 4JHH = 3.7 Hz, 1H), 0.32 (s, 3H, CH3) ppm. 13C APT NMR (101 MHz, THF-d8, 25°C): δ=
155.1 (C0, Ar), 137.8 (C0, Ar), 135.7 (CH, Ar), 128.8 (CH, Ar), 125.5 (CH, Ar), 123.8 (CH, Ar), 89.0 (C0-O), 68.0
(THF), 66.6 (NCH2), 55.3, 52.2, 44.7, 41.1, 35.4, 31.4, 26.2, 26.2, 25.1 (THF + THF-d8), 24.8, 20.2 ppm. Anal.
Calcd. for C31H52CaINO4 (MW = 669.74 g/mol): C, 55.59; H, 7.83; N, 2.09; Found: C, 55.37; H, 7.62; N, 2.00.
14
This article is protected by copyright. All rights reserved.

10.1002/ejic.202000671
European Journal of Inorganic Chemistry
Synthesis of Compound (L6)Ca·(THF)4: Ligand L6-H2 (600 mg, 0.747 mmol) and Ca(OiPr)2 (120 mg
0.759 mmol) were suspended in THF (10 mL). The white suspension was heated in a laboratory microwave
to 140 °C for 6 h yielding a bright yellow turbid solution. Traces of unreacted Ca(OiPr)2 were removed by
centrifugation and the clear orange solution was filtered over Celite. The solvent was removed under
vacuum yielding a yellow-orange powder. Subsequent washing with pentane (2 x 5 mL) and drying under
reduced pressure at 60 °C gave the crude product as a yellow powder. The powder was dissolved in THF
(10 mL), filtered through a short Celite plug and stored at -30 °C for prolonged time. Bright yellow crystals
were collected and dried under high vacuum at 60 °C. Crystals suitable for X-ray analysis were grown from
1
a hot THF solution by slow cooling to room temperature. Yield: 0.54 g, 0.478 mmol, 64 %. H NMR (400
MHz, THF-d8, 25°C): δ = 7.69-7.67 (m, 12H), 7.46 (s, 2H), 7.24-7.15 (m, 20H), 6.75-6.69 (m, 4H), 6.64-6.60
(dt, 3JHH = 5.5 Hz, 4JHH = 2.1 Hz, 2H) ppm. 13C APT NMR (101 MHz, THF-d8, 25°C): δ = 171.6 (C0, Naph), 140.5
(C0, Naph), 139.8 (CH, Naph), 139.2 (C0, Ph), 137.3 (CH, Ph), 128.8 (CH, Ph), 127.9 (CH, Naph), 127.7 (CH,
Ph), 127.3 (C0, Naph), 126.7 (C0, Naph), 125.4 (CH, Naph), 125.1 (CH, Naph), 118.2 (CH, Naph), 118.0 (C0,
Naph) 67.2 (THF + THF-d8), 25.1 (THF + THF-d8) ppm. Anal. Calcd. for C72H72CaO6Si2 (MW = 1129.61 g/mol):
C, 76.56; H, 6.42; Found: C, 76.39; H, 6.27.
Catalytic intramolecular alkene hydroamination with hybrid catalysts: A J. Young NMR-tube was charged
with a metal salt or complex (0.025 mmol, 0.1 equiv.), a strong organic base (0.025 mmol, 0.1 equiv.) and
H2C=CHCH2CPh2CH2NH2 (59.4 mg, 0.25 mmol, 1.0 equiv.). The mixture was dissolved either in C6D6 or THF-
1
d8 (550 µL). The conversion to the corresponding pyrrolidine product was monitored by H NMR
spectroscopy.
Asymmetric catalytic intramolecular alkene hydroamination with hybrid catalysts: A J. Young NMR-tube
was charged with a chiral Ca complex (0.025 mmol, 0.1 equiv.), the Schwesinger base tBuP4 (15.9 mg,
0.025 mmol, 0.1 equiv.), H2C=CHCH2CR2CH2NH2 (0.25 mmol, 1.0 equiv.) and C6D6 (550 µL). The cyclization
to the corresponding pyrrolidine product was monitored by 1H NMR spectroscopy. After completion, the
NMR tube was opened to the atmosphere and filtered over Celite. Subsequently, O-acetyl mandelic acid
(72.8 mg, 0.375 mmol, 1.5 equiv.) was added and the reaction mixture was stirred for 30 min in a vortex
stirrer. The enantiomeric excess was determined via integration of the corresponding diastereomer
1
signals using H NMR spectroscopy. In case of an insufficient resolution, the cyclization experiment was
repeated and Mosher´s chloride (94.7 mg, 0.375 mmol, 1.5 equiv.) and triethylamine (37.9 mg, 0.375
mmol, 1.5 equiv.) were added under inert conditions and the reaction mixture was stirred for 1 h in a
vortex. The ammonium chloride precipitate was removed by centrifugation and filtration and the
15
This article is protected by copyright. All rights reserved.

10.1002/ejic.202000671
European Journal of Inorganic Chemistry
enantiomeric excess was determined via integration of the corresponding 19F NMR signals. In order to
obtain the best signal separation, NMR spectra were measured at various temperatures between 20°C
and 70 °C and the NMR spectrum with the best signal separation was chosen.
Acknowledgments
We kindly acknowledge the Deutsche Forschungsgemeinschaft (DFG) for funding of this project (HA
3218/8-1) and thank Antigone Roth and Christina Wronna for multiple elemental analyses and Jochen
Schmidt and Dr. Christian Färber for support with NMR measurements.
Declaration
The authors declare no competing financial interest.
References
[1] Reviews: (a) L. L. Schafer, P. Mountford, W. E. Piers, Dalton Trans. 2015, 44, 12027-12028. (b) P.
Chirik, Acc. Chem. Res. 2015, 48, 1687-1695. (c) V. Rodriguez-Ruiz, R. Carlino, S. Bezzenine-Lafollée,
R. Gil, D. Prim, E. Schulz, J. Hannedouche, Dalton Trans. 2015, 44, 12029-12059. (d) M. Albrecht, R.
Bedford, B. Plietker, Organometallics 2014, 3, 5619-5621. (e) J. R. McKone, S. C. Marinescu, B. S.
Brunschwig, J. R. Winkler, H. B. Gray, Chem. Sci. 2014, 5, 865-878.
[2] Reviews: (a) S. Harder, Chem. Rev. 2010, 110, 3852-3876. (b) A. G. M. Barrett, M. R. Crimmin, M. S.
Hill, P. A. Procopiou, Proc. R. Soc. A 2010, 466, 927-963. 8C9 m: Westerhausen; J. Langer, S. Krieck,
C. Glock, Rev. Inorg. Chem. 2011, 31, 143-184. (d) M. R. Crimmin, M. S. Hill, Topics in Organometallic
Chemistry, Ed. S. Harder, 2013, vol. 45, p. 191.
[3] L. A. Berben, Chem. Eur. J. 2015, 21, 2734-2742.
[4] S. Harder, in “Early Maingroup Metals in Catalysis: Concepts and Reactions” Wiley-VCH, Weinheim,
Germany 2019, Chapter 10.
[5] (a) J.-M. Begouin, M. Niggemann, Chem. Eur. J. 2013, 19, 8030-8041. (b) S. Kobayashi, Y. Yamashita,
Acc. Chem. Res. 2011, 44, 58-71.
[6] J. Penafiel, L. Maron, S. Harder, Angew. Chem. Int. Ed. 2015, 54, 201-206.
[7] R. F. Weitkamp, B. Neumann, H. G. Stammler, B. Hoge, Angew. Chem. 2019, 131, 14775–14780.
[8] J. Lam, K. M. Szkop, E. Mosaferi, D. W. Stephan, Chem. Soc. Rev. 2019, 48, 3592-3612.
16
This article is protected by copyright. All rights reserved.

10.1002/ejic.202000671
European Journal of Inorganic Chemistry
[9] S. Grabowsky, M. F. Hesse, C. Paulmann, P. Luger, J. Beckmann, Inorg. Chem. 2009, 48, 4384-4393.
[10] J. Pahl, H. Elsen, S. Harder, Chem. Commun. 2018, 54, 7846-7849.
[11] J. Langer, S. Krieck, R. Fischer, H. Görls, M. Westerhausen, Z. Anorg. Allg. Chem. 2010, 636, 1190–
1198.
[12] M. R. Crimmin, I. J. Casely, M. S. Hill, J. Am. Chem. Soc. 2005, 127, 2042-2043.
[13] M. R. Crimmin, M. Arrowsmith, A. G. M. Barrett, I. J. Casely, M. S. Hill, P. A. Procopiou, J. Am.
Chem. Soc. 2009, 131, 9670–9685.
[14] N. Romero, S. Rosca, Y. Sarazin, J.-F. Carpentier, L. Vendier, S. Mallet-Ladeira, C. Dinoi, M. Etienne,
Chem.–Eur. J. 2015, 21, 4115-4125.
[15] B. Liu, T. Roisnel, J.-F. Carpentier and Y. Sarazin, Chem.–Eur. J. 2013, 19, 13445-13462.
[16] A. G. M. Barrett, M. R. Crimmin,M. S. Hill, P. B. Hitchcock, G. Kociok-Köhn, P. A. Procopiou, Inorg.
Chem. 2008, 47, 7366-7376.
[17] A. G. M. Barrett, M. R. Crimmin, M. S. Hill, P. B. Hitchcock, G. Kociok-Köhn, P. A. Procopiou, Inorg.
Chem. 2009, 48, 4445–4453.
[18] B. Freitag, C. A. Fischer, J. Penafiel, G. Ballmann, H. Elsen, C. Färber, D. F. Piesik, S. Harder, Dalton
Trans. 2017, 46, 11192-11200.
[19] a.) J. N. Begouin, M. Niggemann, Chem. Eur. J. 2013, 19, 8030-8041. b.) L. Fu, M. Niggemann, Chem.
Eur. J. 2015, 21, 6367-6370. c.) V. J. Meyer, C. Ascheberg, M. Niggemann, Chem. Eur. J. 2015, 21,
6371-6374. d.) T. Haven, G. Kubik, S. Haubenreiser, M. Niggemann, Angew. Chem. Int. Ed. 2013, 52,
4016-4019. e.) V. J. Meyer, L. Fu, F. Marquardt, M. Niggemann, Adv. Syn. Cat. 2013, 335, 1943-1947.
[20] C. A. Fischer, A. Rösch, H. Elsen, G. Ballmann, M. Wiesinger, J. Langer, C. Färber, S. Harder, Dalton
Trans. 2019, 48, 6757–6766.
[21] (a) F. Buch, S. Harder, Z. Naturforsch. 2008, 63b, 169-177. (b) J. Wixey, B. Ward, Dalton Trans. 2011,
40, 7693-7696. (c) J. Wixey, B. Ward, Chem. Commun. 2011, 47, 5449-5451. (d) D. Nixon, B. Ward,
Chem. Commun. 2012, 48, 11790-11792.
[22] M. Leven, N. E. Schlörer, J. M. Neudörfl, B. Goldfuss, Chem. Eur. J. 2010, 16, 13443 – 13449.
[23] B. Goldfuss, M. Steigelmann, S. I. Khan, K. N. Houk, J. Org. Chem. 2000, 65, 77-82.
[24] (a) R. M. Peltzer, O. Eisenstein, A. Nova, M. Cascella, J. Phys. Chem. B 2017, 121, 4226−4237. (b) R.
Mathias, J. Gauss, O. Eienstein, M. Cascella. J. Am. Chem. Soc. 2020, 142, 2984−2994.
[25] M. E. Jung, G. Piizzi, Chem. Rev. 2005, 105, 1735-1766.
[26] T. Lee, J. F. Peng, Crys.l Growth Des. 2010, 10, 3547-3554.
[27] R.D. Shannon, Acta Crystallogr. 1976, A32, 751-767.
17
This article is protected by copyright. All rights reserved.

10.1002/ejic.202000671
European Journal of Inorganic Chemistry
[28] M. Niehues, G. Kehr, G. Erker, B. Wibbling, R. Fröhlich, O. Blaque, H. Berke, J. Orgomet. Chem. 2002,
663, 192-203.
[29] M. H. Dunn, M. Konstandaras, M. L. Cole, J. B. Harper, J. Org. Chem. 2017, 82, 7324-7331.
[30] R. Schwesinger, H. Schlemper, Angew. Chem. 1987, 99, 1212-1214.
[31] P. Horrillo Martínez, K. C. Hultzsch, F. Hampel, Chem. Commun. 2006, 21, 2221−2223.
[32] R. R. Milburn, H. S. M. Shakil, O. Prien, Z. Ahmed, V. Snieckus, Organic Letters 2007, 9, 4403-4406.
[33] Y. Zhang, C. Lim, S. Boon, S. Derek, H. Pan, Y. Zhao, Angew. Chem. Int. Ed. 2013, 126, 1423-1427.
[34] W. Zhang, X. Zhang, Angew. Chem. Int. Ed. 2006, 45, 5515-5518.
[35] R. I. Storer, D. E. Carrera, Y. Ni, D. W. MacMillan, J. Am. Chem. Soc. 2006, 128, 84-86.
[36] G. R. Fulmer, A. J. M. Miller, N. H. Sherden, H. E. Gottlieb, A. Nudelman, B. M. Stoltz, J. E. Bercaw,
K. I. Goldberg, Organometallics 2010, 29, 2176-2179.
18
This article is protected by copyright. All rights reserved.