Structural diversity in pyridine and polypyridine adducts of ring slipped manganocene: Correlating ligand steric bulk with quantified deviation from ideal hapticity

Article abstract of DOI:10.1039/c8dt00537k

We have synthesized several new manganocene-adduct ([Cp₂Mn(L)] = 1-L) complexes using pyridine and polypyridine ligands and report their molecular structures and characterization data. Consistent with other molecules in this class [(η^x-Cp)₂MnL_n] or [(η^x-Cp)₂Mn(L-L)] (n = 1, 2; x = 1, 3, or 5), the manganese-cyclopentadienide interaction deviates from the classical η^x interactions (x = 3 or 5). Such deviations have been ascribed to steric factors and often called non-ideal hapticity. However, there is no quantification of this non-ideal hapticity and thus it is difficult to evaluate the extent of ring slippage or assign hapticity. Furthermore, the hypothesis that non-ideal hapticity in high-spin Mn^II complexes is induced by steric interactions has not been systematically evaluated. Therefore, we report herein a quantified scale for deviation from ideal hapticity between zero (ideal η⁵ interaction) and one ( η¹ interaction). This quantified deviation from ideal hapticity has an empirical relationship with the ligand's steric properties, which strongly supports the premise that steric interactions cause the deviations in ionic M-Cp interactions.

Full text of DOI:10.1039/c8dt00537k

Dalton
Transactions
View Article Online
View Journal
PAPER
Structural diversity in pyridine and polypyridine
adducts of ring slipped manganocene: correlating
ligand steric bulk with quantified deviation from
ideal hapticity†
Cite this: DOI: 10.1039/c8dt00537k
Anthony F. Cannella, a Suman Kr Dey,a Samantha N. MacMillanb and
a
David C. Lacy
*
We have synthesized several new manganocene-adduct ([Cp2Mn(L)] = 1–L) complexes using pyridine and
polypyridine ligands and report their molecular structures and characterization data. Consistent with
other molecules in this class [(ηx-Cp)2MnLn] or [(ηx-Cp)2Mn(L–L)] (n = 1, 2; x = 1, 3, or 5), the manganese-
cyclopentadienide interaction deviates from the classical ηx interactions (x = 3 or 5). Such deviations have
been ascribed to steric factors and often called non-ideal hapticity. However, there is no quantification of
this non-ideal hapticity and thus it is difficult to evaluate the extent of ring slippage or assign hapticity.
Furthermore, the hypothesis that non-ideal hapticity in high-spin MnII complexes is induced by steric
interactions has not been systematically evaluated. Therefore, we report herein a quantified scale for devi-
ation from ideal hapticity between zero (ideal η5 interaction) and one (“η1” interaction). This quantified
deviation from ideal hapticity has an empirical relationship with the ligand’s steric properties, which
strongly supports the premise that steric interactions cause the deviations in ionic M–Cp interactions.
Received 8th February 2018,
Accepted 3rd March 2018
DOI: 10.1039/c8dt00537k
rsc.li/dalton
tion is called “ring slippage” which involves a change in hapti-
city of a Cp ring that remains planar. Veiros and coworkers
Introduction
Cyclopentadienide (Cp) is undoubtedly one of the most impor- have discussed the differences between folding and ring slip-
tant ligands in organometallic chemistry. Since the discovery page showing that, in the latter case, the extent of slippage can
of its η5 coordination to iron by structural means,1 Cp is found be ambiguous and often described as non-ideal η5 coordi-
as an auxiliary ligand in a variety of applications that include nation (rather than η3, etc.).8
electron transfer studies,2 asymmetric polymerizations,3 and
This type of ring slippage is especially prevalent for manga-
medicine.4 One quality of Cp and related cyclic η5-dienyl nese(II) cyclopentadienyl complexes because of the ionic
ligands is their ability to impact chemical reactivity by means nature of the Mn(II)–carbon bond.9 For instance, the ionic
of ring slip isomerization from a complex that has η5 coordi- character makes the Cp ligand in manganocene (Cp2Mn = 1)
nation to one that is η3-allyl or monodentate (“η1”) and can more labile10 and susceptible to attack by an acid compared to
even completely dissociate to form an outer sphere counter other first row metallocenes.11 Additionally, manganocene will
anion (Chart 1).5 NMR spectroscopy is often used to reveal form manganate anions to induce ring slippage by the
changes in hapticity because the dynamic nature of η3-allyl Cp addition of weak nucleophiles (e.g., Cp) to the metal center.12
ligands makes them difficult to crystallize.6 Known examples Furthermore, the reaction of neutral monodentate (L) and
of structurally characterized complexes with bona fide hapto- bidentate (L–L) Lewis bases such as ethers,13 carbenes,14
tropic shifts display “ring folding”, which indicates an auth- amines,15 and phosphines16 with manganocene is also
entic η3-allyl binding mode.7 An alternative form of isomeriza-
aDepartment of Chemistry, University at Buffalo, State University of New York,
Buffalo, New York 14260, USA. E-mail: dclacy@buffalo.edu
bDepartment of Chemistry and Chemical Biology, Cornell University, Ithaca,
New York 14853, USA
†Electronic supplementary information (ESI) available. CCDC 1822539–1822546.
For ESI and crystallographic data in CIF or other electronic format see DOI:
10.1039/c8dt00537k
Chart 1 Diverse coordination chemistry of Cp anion.
This journal is © The Royal Society of Chemistry 2018
Dalton Trans.

View Article Online
Paper
Dalton Transactions
reported to result in simple adducts of the type [(η5-Cp)2MnL]
or [(ηx-Cp)2Mn(L–L)] (x = 1 or 5) with ring slippage.
These ring slips are often accompanied by tilts of the
planar Cp rings relative to the perpendicular axis. It has been
shown that electronic arguments17 do not apply to high-spin
manganese(II) compounds and that steric factors are at play.
Wilkinson, Hursthouse, and coworkers first forwarded this
hypothesis for their high-spin 1-L and 1-L–L adducts, stating
that electronic factors are likely a minor component to ring
tilts because of negligible metal–ligand covalency.16 Soon after
this, Bottomley used the bond angles in 1-THF to show that
steric factors were indeed responsible for the observed ring
tilts.13 The factors that result in slippage are therefore likely to
also be steric in nature. More recent observations made by
Walter, Wright, and Layfield for their CpnMn-L (n = 1,2) type
complexes also deviate from ideal η5 coordination and provide
further support for this ionic model.18 Finally, the chelating
diphosphine and diamine ligands dmpe (dmpe = 1,2-bis(di-
methylphosphino)ethane) and TMEDA (TMEDA = N,N,N′N′-
tetramethylethylene-1,2-diamine) have both been used to
prepare 1-(L–L) compounds. The 21 electron 1-dmpe complex
contains two nominal η5-Cp ligands, whereas the 17 electron
1-TMEDA complex contains one η5-Cp ligand and one η1-Cp
ligand.15,16 Tertiary amine ligands are more bulky than their
phosphine congeners and further demonstrate the effects of
steric bulk in these otherwise similar molecules.
Additional compounds with varying degrees of hapticity are
required to confirm the role of steric effects in ring slip and to
gather guiding principles to control the extent of slippage.
Herein, we describe the synthesis and characterization of
additional compounds of the 1-L class using mono-, di-, and tri-
dentate (poly)pyridine ligands. Using these new compounds and
data from an extensive list of known manganocene adducts, we
structurally parameterize the deviation from ideal η5 coordination
in this family of compounds. In doing so, a strategic method to
control ring slip has emerged in addition to providing a novel
method to empirically evaluate the steric bulk of neutral ligands.
Scheme 1 Synthesis of Cp2Mn (1) adducts: L = 2,2’-bipy (bipy, R1 = R2
= H); 4,4’-Me2bipy (R1 = H, R2 = Me); 6,6’-Me2bipy (R1 = Me, R2 = H);
6,6’-Ph2bipy (R1 = Ph, R2 = H); 1,10-phenanthroline (phen, R1 = R2 = H);
4,7-Ph2phen (R1 = H, R2 = Ph); 2,9-Me2phen (R1 = Me, R2 = H); py2 (py =
pyridine); terpy (terpyridine).
ether, dried open to the glovebox nitrogen atmosphere, and
finally sealed in briefly evacuated glass ampules were we able to
obtain reliable CHN analysis for most of the compounds. The
instability and difficulty associated with obtaining reliable CHN
for these 1-L adducts was also encountered by Howard et al. for
1-PR3 (PR3 = PMePh2, PMe3, dmpe).16
THF solutions of 1 are colorless, but the 1-L adducts com-
plexes are yellow or red in THF. The UV-vis spectra of 1-L are gen-
erally featureless save for intense bands in the UV region associ-
ated with ligand π–π* transitions and ill-defined bands in the
visible region that are likely CT in nature owing to the fact that
ligand field transitions in high-spin Mn(II) are spin forbidden
(Fig. S1†). A similar color change from colorless to red was
observed when 1 was treated with 6,6′-diphenylbipyridine in THF.
However, attempts to isolate and characterize 1-6,6′-2,2′-diphenyl-
bipyridine only resulted in the isolation of free ligand and 1.
The magnetic properties of the adducts were explored with
EPR and Evans’ method. The solution state magnetic
moments for some of the adducts were determined in
benzene, but low solubility and relative instability of the com-
plexes precluded most of the compounds from being
measured. In the instances where data were collected, the
values were near those expected for high-spin 1 complexes (µeff
= 6.1µB for 1-2,9-Me2phen). The EPR spectra at low tempera-
ture (6–15 K) exhibit features nearly identical to free mangano-
cene (Fig. S1†). Overall the conclusion is that the compounds
exhibit high-spin electronic configurations that are well within
the expectations for most organometallic Mn(II) complexes.
Results
Synthesis and characterization of the 1-L adducts
The [Cp2Mn-L] adducts (1-L) were synthesized by mixing Cp2Mn
(1) and the desired ligand (L) in a 1 : 1 stoichiometry in THF at
Molecular structures of the 1-L adducts
room temperature (Scheme 1). The 1-L adducts were purified by The 1-L complexes exhibit typical binding modes between the
crystallization from THF or toluene at −35 °C and afforded crys- metal center and the polypyridine ligand (Fig. 1). However, the
tals suitable for diffraction. We attempted to synthesize the Cp–Mn interaction varies depending on the polypyridine
mono-, di-, and tri-pyridine adducts 1-pyn (n = 1, 2, and 3) by ligand and is the subject of the remainder of this manuscript.
adding the corresponding equivalents of pyridine to 1, but we The 1-L adducts crystallized in a variety of space groups and
only obtained 1-py2 for all three reactions. The compounds are often contained multiple molecules per asymmetric unit
not indefinitely stable and therefore prepared fresh in small (Fig. S2–S9†). For instance, the asymmetric unit of 1-bipy con-
batches and stored in the crystallization mother liquor. The tained three molecules of the form [Cp2Mn(bipyridine)] and
combustion CHN analysis of week-old samples or samples each of these three molecules have their own unique bond
dried extensively under vacuum provided carbon percentages metrics and are therefore treated separately in our analysis.
substantially lower than expected. Only when samples were
In order to rationalize the observed structural diversity in
freshly prepared and washed sparingly with cold petroleum the 1-L adducts we operationally define several key structural
Dalton Trans.
This journal is © The Royal Society of Chemistry 2018

View Article Online
Dalton Transactions
Paper
Discussion
Single crystal X-ray diffraction analysis reveals that each of
the eight new 1-L adducts contain typical coordination
between the metal ion and the (poly)pyridine ligand but have
ambiguous hapticity Cp ligation. A notable characteristic is
the diversity in the coordination of the Cp rings that manifest
in a degree of ring-slippage (S), M–Cpcentroid distance (l), and
tilts (τ) from complex to complex (Fig. 2). Upon initial inspec-
tion, there was no obvious trend between the (poly)pyridine
ligand in 1-L and the extent of ring-slippage except that the 1-
terpy adduct contains the largest M–Cpcentroid distance. A cor-
relation was indeed found between the steric bulk of the
ligand and the extent of ring-slippage, but this required a
quantification of the extent of hapticity change in addition to
quantification of the ligand steric bulk, each of which is
described in the following sections. The Tolman cone angle
(ΘT) and buried volume parameters (%Vbur) have been exten-
sively used and conveniently defined and therefore constitute
our preferred choices in parameterizing the adduct ligand.
Finally, in addition to the eight new compounds synthesized
and characterized herein, we also compiled the important
bond metrics for all of the known relevant manganocene
adduct complexes (Table 1) to effectively discover a corre-
lation with our new non-ideal hapticity parameter and the
steric bulk of the ligands.
Fig. 1 Molecular structure of one of the molecules in the asymmetric
unit for compounds 1-bipy (left) and 1-terpy (right); ellipsoids shown at
50% probability and solvent molecules and H-atoms in the crystal struc-
ture are not displayed. Atom colors: grey = C; blue = N; magenta = Mn.
See Fig. S2–9† for remaining molecular structures.
Fig. 2 Geometric parameters used in this study (left) and traditional
binning of hapticity (right).
Non-idealized hapticity parameter (D)
components (Fig. 2). The first of these important metrics that
describe the 1-L complexes is the Mn–Cpcentroid distance (l)
(each complex having two values of l) and is defined as the dis-
tance from the Mn ion to the center of the Cp ring (M–
Cpcentroid distance). The slip parameter (S)19 for planar Cp
rings, a characteristic of all the Cp rings studied here, is
defined as the distance from the Cpcentroid to a point on the Cp
ring that coincides with a normal vector (N) from the metal
ion to the plane of the Cp ring. The slip metric S forms the
short edge of a right triangle with l as the hypotenuse and N as
the second largest length. The angle between l and N is thus
the tilt angle (τ) and is easily obtained by using the sine law as
shown in eqn (1).
The M–Cp interaction, l, in the 1-L adducts does not fall into
idealized hapto environments (η5, η3, η2, or η1, Fig. 2) and, at
first glance, does not appear to follow any trend, nor was there
any evidence that electronic factors are at play. We are not the
first to notice this seemingly ambiguous interaction, which
has been described before as non-idealized η5 interactions and
attributed to steric factors.13,16,18 Prior to this report, a sys-
tematic evaluation of this proposal was not possible and there-
fore we formulated a new parameter, D, that quantifies the
deviation from idealized hapticity. For a given system of com-
pounds, this value D is defined by eqn (2),
x ꢀ xmin
D ¼
ð2Þ
xmax ꢀ xmin
ꢀ ꢁ
S
l
τ ¼ sinꢀ1
ð1Þ
where x is some geometric value, or a metric that is deter-
mined from geometric values, measured from X-ray crystal
structures. At the onset, it was not obvious which geometric
value or metric to use for the calculation of D. To solve this
issue, we independently quantified the steric bulk properties
of the adduct ligand ΘT and %Vbur and plotted them against
various geometric values or metrics. These are discussed
later, but briefly here we found that the best correlation of D
was obtained using the metric l × N for the value of x in eqn
(2). The restriction of S to inside the Cp ring creates a
relationship between N and l for which l × N is a metric
(meaning a calculated-measurement). Hence, eqn (2) uses
(l × N)min as the minimum observed value found for the cen-
troid distance multiplied by its corresponding N among the
These three parameters (l, S, and τ) provide the major
points of interest from which the different 1-L adducts are
compared. For example, 1-6,6′-Me2bipy has l = 2.360 and
2.438 Å, S = 0.464 and 0.858 Å, and τ = 11.33° and 20.61°.
These are quite different from the simple 1-THF with l =
2.164 Å, S = 0.093 Å, and τ = 2.47°. A complete list of these
metrics for the complexes that we synthesized and related
compounds from the literature is summarized in Table 1.‡
‡Table 1 is available as an excel file; please contact corresponding author: dcla-
cy@buffalo.edu.
This journal is © The Royal Society of Chemistry 2018
Dalton Trans.

View Article Online
Paper
Dalton Transactions
Table 1 Crystallographic and geometric parameters for manganocene adductsa
lg (Å)
D
Davg
τ (°)
S (Å)
N (Å)
M–L (Å)
Avg ΘT (°)c
Avg %Vbur
Ref.e
d
Entry
Complex
1
2
1-THF
1-PMe3
2.164
2.179
2.238
2.236
2.194
2.334
2.791
2.222
2.708
2.615
2.205
2.744
2.219
2.439
2.212
2.445
2.683
2.704
2.300
2.300
2.251
2.553
2.425
2.211
2.293
2.275
2.326
2.310
2.278
2.262
2.397
2.291
2.360
2.438
2.514
2.277
2.363
2.281
2.391
2.336
2.412
2.378
2.330
2.330
2.478
2.478
2.703
2.524
2.796
2.396
2.620
2.503
0.133
0.000
0.272
0.251
0.194
0.475
0.946
0.250
0.925
0.792
0.217
0.895
0.206
0.565
0.193
0.583
0.920
1.000
0.353
0.353
0.320
0.688
0.552
0.191
0.376
0.336
0.386
0.413
0.341
0.331
0.516
0.310
0.509
0.558
0.645
0.338
0.494
0.332
0.506
0.496
0.580
0.454
0.436
0.436
0.604
0.604
0.882
0.741
0.995
0.563
0.808
0.726
0.133
0.136
2.47
22.02
7.27
10.00
3.46
8.22
35.12
4.55
30.35
27.76
3.45
33.70
11.24
17.99
10.91
20.04
28.63
27.52
13.86
13.86
0.00
26.49
19.28
10.92
9.58
0.093
0.817
0.283
0.388
0.132
0.334
1.605
0.176
1.368
1.218
0.133
1.522
0.430
0.848
0.419
0.838
1.285
1.249
0.551
0.551
0.000
1.139
0.801
0.419
0.382
0.386
0.634
0.383
0.392
0.223
0.724
0.629
0.464
0.858
1.048
0.398
0.550
0.465
0.714
0.216
0.638
0.773
0.485
0.485
0.961
0.961
1.407
0.908
1.568
0.575
1.214
0.825
2.162
2.020
2.220
2.202
2.190
2.310
2.283
2.215
2.337
2.314
2.201
2.283
2.177
2.287
2.172
2.297
2.355
2.398
2.233
2.233
2.251
2.285
2.289
2.171
2.261
2.242
2.238
2.278
2.244
2.251
2.285
2.203
2.314
2.282
2.285
2.242
2.298
2.233
2.282
2.326
2.326
2.249
2.279
2.279
2.284
2.284
2.308
2.355
2.315
2.326
2.322
2.363
2.226
104.15
101.41
17.4
19.0
13
16
2.577f
3
1-PPh2Me
0.222
2.612f
122.91
22.0
16
4
5
1-dmpe
1-TMEDA
0.475
0.598
2.674f
2.354
2.338
2.119f
2.109f
2.226
2.261
2.227
(184.71)
100.96
(201.93)
132.96
(265.91)
103.44
(206.88)
154.08
31.8
38.9
16
15
6
1-(TMG)2
1-(BzEDA)
1-(NHCArMe2Br
1-IMes
0.859
0.556
0.385
0.388
0.960
0.353
0.504
0.372
0.364
21.0
(41.3)
37.9
18a
7
18b
8
)
29.7
30.3
14a
9
2.222
159.21
14a
10
11
12
13
14
1-(NHCMe4
1-IiPrMe2
1-ItBu
)
2
2.214
2.219
2.159
151.19
(302.38)
151.62
23.7
(46.7)
26.6
14a
14b
2.222
2.225
—
34.9
30.3
30.8
14b
1-IMes
156.07
14b
1-bipyb
Mn1
Mn2
Mn3
2.243
2.251
2.236
2.222
2.232
2.257
2.223
2.217
2.233
2.254
2.218
2.212
2.262
2.253
2.255
2.251
2.244
2.253
2.227
2.227
2.210f
2.210f
2.238
2.355
2.282
2.371
2.239
2.357
91.59
(183.18)
This work
9.77
15.81
9.55
9.91
5.65
15
16
17
1-4,4′-Me2bipy
1-6,6′-Me2bipy
1-phenb
0.413
0.533
0.452
17.58
15.93
11.33
20.61
24.64
10.06
13.47
11.77
17.37
5.30
15.35
18.96
12.01
12.01
22.82
22.82
31.37
21.09
34.11
13.88
27.59
19.25
91.91
30.9
37.2
30.6
This work
This work
This work
(183.82)
114.72
(229.44)
91.18
Mn1
Mn2
Mn1
Mn2
(182.37)
18
1-2,9-Me2phenb
0.509
115.78
(231.55)
36.7
30.7
This work
19
20
21
1-4,7-Ph2phen
1-py2
0.436
0.604
0.786
91.17
This work
This work
This work
(182.35)
114.99
(229.98)
125.17
(250.34)
18.6
(36.6)
42.1
1-terpyb
Mn1
Mn2
Mn3
Abbreviations: l = Mn–Cpcent distance (Å); D = non-idealized hapticity parameter; Davg = average of the values of D for a given complex; adduct ΘT
= Tolman cone angle of the adduct (°); %Vbur = percent buried volume of the adduct; τ = tilt angle (°); S = magnitude of slip vector (Å); N =
normal vector from Mn to the plain of the Cp ring. a See Fig. 2 for the definition of parameters. b Asymmetric unit contains multiple molecules
and is designated as Mn1, Mn2, etc. c As determined in this work: steric ligand angles (ΘT) were determined from crystallographic parameters
using eqn (3) and (4). Cone angles in parentheses are ΘT multiplied by two (see Discussion). d As determined in this work: values in parenthesis
are %Vbur calculated using both monodentate ligands as a single entity. The ΘT and %Vbur for compounds with multiple molecules in the
crystallographic unit cell are reported as the average value. e References are for crystallographic information only. f M–L distance is outside one
standard deviation (0.109 Å) of the average M–L distance (2.267 Å) and thereby give potentially erroneous ΘT. Ligand abbreviations: dmpe = 1,2-
bis(dimethylphosphino)ethane; TMEDA = tetramethylethylenediamine; TMG = 1,1,3,3-tetramethylguanidine; BzEDA = N1,N2-dibenzylethane-1,2-
diamine; NHCArMe2Br = 1,3-bis(2,6-dimethyl-4-bromophenyl)-imidazol-2-ylidene; NHCMe4 = 1,3,4,5-tetramethylimidazol-2-ylidene; IMes = 1,3-
dimesitylimidazol-2-ylidene; IiPrMe2 = 1,3-diisopropyl-4,5-dimethylimidazol-2-ylidene; ItBu = 1,3-di-tert-butylimidazol-2-ylidene. g esd ≤ 0.002 Å.
1-L adducts, and (l × N)max is the maximum observed value nature, a comparative metric and must be defined by a
found for the centroid distance multiplied by its corres- specific set of compounds bounded by the two most extreme
ponding N among the 1-L adducts. The D parameter is, by examples. Complex 1-PMe3 has the smallest l × N value and
Dalton Trans.
This journal is © The Royal Society of Chemistry 2018

View Article Online
Dalton Transactions
Paper
was used for the idealized η5 interaction with a D value equal on the ligand. To calculate a full cone angle ΘT, Tolman uti-
to zero for one of the Mn–Cp interactions (Table 1, entry 2). lized eqn (3),
On the other end of the spectrum, the maximum deviation
ꢀ ꢁ
X
2
i
θi
from idealized hapticity is defined as essentially having a so-
called η1 interaction. The largest value of l × N observed
among any known 1-L complex is for one of the M–Cp inter-
ΘT
¼
ð3Þ
2
i
where the “half angles” (θi/2) are the angles formed from the
metal–phosphine bond and outermost van der Waals radii for
the ith substituent (Fig. 3).
Mingos and coworkers later developed a method to obtain
the cone angle directly from crystallographic parameters,
which included the H-atom van der Waals radii.22 The half
angle determined by Mingos’ method requires eqn (3) and (4),
actions in 1-NHCMe4 (Table 1, entry 10). One of the Mn–Cp
2
interactions in 1-terpy is a near second to a maximum devi-
ation with a D value of 0.995 (Table 1, entry 21). One can also
take the average for all the D values (Davg) for each of the
different M–Cp interactions in a single compound. The Davg
parameter actually serves as the more meaningful parameter
with which the steric parameters are correlated because the
adduct ligand affects both Cp rings simultaneously
(vide infra).
ꢂ
ꢃ
θi
180
rH
d
¼ αMingos
þ
sinꢀ1
ð4Þ
2
π
where α and d are the parameters that can be obtained directly
from the crystal structure (Fig. 3). The H-atom van der Waals
radii are 1.0 Å in C–H bonds, which are relevant here.22,23
Since eqn (3) does not mathematically describe a three-
dimensional shape, we can use the same formulae (eqn (3)
and (4)) to calculate a two-dimensional “cone” angle (also
designated ΘT for convenience) for planar monodentate
ligands such as THF and pyridine. Tolman restricted the value
of i from 1–3, but Walter and coworkers effectively used values
of i from 1–5 to determine the cone angles of Cp ligands (ΘW,
Table 2).24 Hence, the value of i for THF and pyridine is simply
two instead of three. Unfortunately, the lack of three-dimen-
sional information in ΘT negatively impacts the comparative
value between ligands such as THF and PPh2Me. Later, this
issue is resolved using percent-buried volume (vide infra).
Many of the ligands used in this study are multidentate and
afford a different problem from which to draw meaningful
comparisons. The historical way to report ΘT for multidentate
ligands is to use eqn (3) with one of the θi/2 being half of the
bite angle and the other θi/2 defined by the bound atom and
van der Waals radii of the outermost atom determined the
same way for phosphine ligands according to eqn (4) (Fig. 4).
ΘT determined this way provides the cone angle for one
side of a bidentate phosphine. A plot of Davg vs. ΘT does not
Cone angle
The ligand cone angle (ΘT) was developed by Tolman to para-
meterize phosphine steric bulk.20,21 This was done using 3D
molecular models and projecting a cone from the metal center
with boundaries defined by the outermost van der Waals radii
Fig. 3 Schematic of geometric entities required for calculation of a
cone angle for a phosphine ligand; figure adaptation from the Tolman
and Mingos’ reports defining cone angles.21,22
Table 2 Comparison of DFT calculated and XRD obtained lavg, adduct ΘT, Cp ΘW, and %Vbur
lavg (Å)a
Adduct ΘT (°)
Cp ΘW (°)a
%Vbur
1-L
Ligand
DFT
XRD
|diff|b
DFT
XRD
|diff|b
DFT
XRD
|diff|b
DFT
XRD
|diff|b
bipy
2.338
2.391
2.334
2.350
2.317
2.398
2.320
2.525
2.644
2.291
2.534
0.047
0.143
187.00
227.80
233.51
186.87
184.00
227.36
184.99
237.87
255.84
183.18
229.45
3.82
1.65
87.70
86.28
88.02
87.46
88.41
86.07
88.08
82.44
79.58
85.56
82.58
2.14
3.70
30.0
36.1
33.8
30.1
29.4
36.0
29.8
35.8
41.9
30.8c
37.2
0.8
7.1
6,6′-Me2bipy
6,6′-Ph2bipye
4,4′-Me2bipy
phen
2,9-Me2phen
4,7-Phphen
py2
2.345
2.359
2.38
2.331
2.479
2.591
0.004
0.043
0.018
0.011
0.046
0.052
183.81
182.37
231.55
182.35
229.98
250.34
3.06
1.63
4.19
2.64
7.89
5.50
83.61
84.23
83.08
84.51
80.57
78.10
3.85
4.18
2.99
3.57
1.87
1.48
30.6
30.9
36.7
30.7
36.6
42.1d
0.5
1.5
0.7
0.9
0.8
0.2
terpy
0.046 0.043
3.797 2.106
2.973 1.016
0.8 0.4
a The Cp cone angle provided (ΘW) is average of both M–Cp interactions.24 b Absolute value of the difference of DFT–XRD. c NiBr2bipy = 35.7 d Ru
(H2O)(bipy)(terpy) = 47.7.25 e The adduct 1-6,6′-Ph2bipy could not be isolated.
This journal is © The Royal Society of Chemistry 2018
Dalton Trans.

View Article Online
Paper
Dalton Transactions
entheses. The correlation between deviations from idealized
hapticity (Davg) and the steric bulk (ΘT) is intuitive, with the
largest ligands showing the largest deviation from idealized
hapticity (Fig. 5).
The Davg parameter provides a substantially better corre-
lation when N × l is used for the metric in eqn (2) than with
other geometric parameters or metrics. For instance, a plot of
Davg that uses the average tilt angle (τ) or the average magni-
tude of the slip vector (S) versus the steric properties of the
ligand have poor correlations (Fig. S10 and Table S2†). Despite
the poor relationships, there are still observable trends in
these complexes with bulkier ligands having larger average S
or τ values.
Fig. 4 Geometric location of half angles (θi/2) for planar multidentate
ligands.
Finally, it is also important to discuss the effect of M–L dis-
tance on the value of ΘT. Tolman standardized the M–P dis-
tance to 2.28 Å and found that deviations in M–P distances of
0.1 Å only changed ΘT by ≤5°.21 Mingos explored the relation-
ship between M–P and ΘT and showed that the average of
many ΘT for M–PPh3 complexes determined using the crystallo-
graphic M–P distance was statistically equivalent to the ΘT
determined by Tolman. Nonetheless, Mingos demonstrated
that a large spread of ΘT values exists over the full range of
M–P distances. Conveniently, the 1-L adducts studied here
have an average M–L bond distance of 2.267 Å with a standard
deviation of 0.109 Å, very close to the standard 2.28 Å used for
conventional ΘT and hence there is no need to standardize the
M–L distances in this study. Some of the M–L distances do fall
outside one standard deviation of the M–L distances investi-
gated here and these are indicated in Table 1 with footnote f.
While the trend between Davg and ΘT reveals a correlation
between steric bulk and deviation from ideal hapticity, the
comparison has some shortcomings. First, Davg varies between
molecules of the same compound and suggests that other
factors, such as crystal packing, are at play. To address this
issue, we used density functional theory to optimize the 1-L
adducts and compare the computed values of D and compare
them to the experimental values. We also used the optimized
geometries to determine ΘT for the adducts. These results are
summarized in Table 2 and show that the computational
results match closely to the experimental values. The small
deviations from DFT and XRD are probably due to subtle
crystal packing effects.
correlate when the traditional ΘT used for bidentate ligands is
used (Fig. 5, blue dots). However, a linear correlation (R2
=
0.93) is obtained when one plots Davg vs. ΘT in which all of the
multidentate and bis(monodentate) ligands are multiplied by
two. Hence, ΘT for multidentate ligands and complexes with
more than one of the same monodentate ligands (e.g., 1-py2)
were summed in order to fully account for the steric bulk of
the ligand; these summed ΘT are reported in Table 1 in par-
Percent buried volume
A second shortcoming in the Davg and ΘT comparison results
Fig. 5 A plot of the deviation from idealized hapticity (Davg) versus the from the failures of ΘT, which are a lack of three-dimensional
Tolman cone angle (ΘT). The circles represent data points with ΘT values
determined directly from eqn (3) and (4). The squares represent data
points with ΘT values determined from eqn (3) and (4) but also take into
consideration both halves of multidentate ligands (e.g., bipy) or twice
information and its inability to address changes in the M–L
distance. Attempts in the scientific community to find a more
inclusive steric parameter have led to the discovery and appli-
cation of the percent buried volume (%Vbur) parameter.26%
the value for compounds with two monodentate ligands (e.g., pyridine).
The inscribed numbers correspond to the entries in Table 1. The red line Vbur is defined as the space occupied by the ligand in the first
is a linear regression fit for the square data points and is shown to high-
coordination sphere of the metal center and has since been
light the trend (R2 = 0.93). The Davg value (Table 1) was obtained by
shown to correlate well with a variety of chemical properties
taking the average of all Mn–Cp interactions in a unit cell for a given
compound’s crystal structure (e.g., a Davg value for a compound in
including reactivity.27 The primary advantage of %Vbur over ΘT
is that it can be universally applied and compared across all
which the crystal structure contains three molecules is the average of six
Mn–Cp D values). The same operation was performed for ΘT.
types of M–L interactions with the freeware SambVca.25
Dalton Trans.
This journal is © The Royal Society of Chemistry 2018

View Article Online
Dalton Transactions
Paper
Cavallo and coworkers found that a sphere radius of 3.5 Å pro-
vided the best fit for a large collection of ligands that included
phosphines and N-heterocyclic carbenes. This insight made by
Cavallo and Nolan is that since a majority of M–L bond dis-
tances fall within the range of 2.0–2.5 Å, a sphere radius of
3.5 Å is sufficient to account for the van der Waals radius of
the atoms directly coordinating to the metal center in most
cases. Proceeding with the same method outlined by Cavallo
and coworkers using SambVca we calculated the %Vbur of the
ligands in the 1-L adducts relevant to this work (Table 1) and
used the same program to generate steric maps for the ligands
(Fig. 6).28 Additionally, we compared the %Vbur determined
from the XRD and those of the DFT computed structures for
the new eight 1-L adducts described herein (Table 2). The devi-
ation from ideal hapticity correlates well with the %Vbur deter-
mined for the 1-L adducts with an essentially equal correlation
compared to ΘT (Fig. 7, R2 = 0.92).
For complexes with multiple monodentate ligands (e.g. 1-
py2), the %Vbur was determined for a single ligand in addition
to both ligands as a single entity, just as it was performed for
the ΘT values (Table 1). For example, the %Vbur of one of the
pyridine ligands in 1-py2 is 18.6%, but if one includes both
pyridine ligands the %Vbur is 36.6%. The correlation shown in
Fig. 7 uses the %Vbur for both ligands as a single entity, as they
are both contributing to the overall steric bulk causing ring
slip. We note that the %Vbur of pyridine in 1-py2 (18.6%, entry
20) is very close to other monodentate ligands such as the
single THF ligand 1-THF (17.4%, entry 1), NHCMe4 in 1-
Fig. 7 A plot of the deviation from idealized hapticity (Davg) versus the
%Vbur. The red line is a linear regression fit to the data shown (R2 = 0.92).
The inscribed numbers correspond to the entries in Table 1. The Davg
value (Table 1) was obtained by taking the average of all Mn–Cp inter-
actions in a unit cell for a given compound’s crystal structure (e.g., a
Davg value for a compound in which the crystal structure contains three
molecules is the average of six Mn–Cp D values). The same operation
was performed for %Vbur
.
(NHCMe4
) (23.7%, entry 10), and TMG in 1-TMG2 (%Vbur
2
Conclusions
21.0%, entry 6). Why the smallest ligand in this list is the only
1-L adduct with a single monodentate ligand is uncertain. We
attempted to synthesize the mono and tris pyridine adduct of
1 but only 1-py2 was isolated from these reactions.
We have confirmed that Cp ring slippage in manganocene
adducts is the result of steric factors by correlating the steric bulk
of the coordinating ligand and the extent of ring slippage. This
required a defined non-ideal hapticity parameter (D) for which we
are the first to execute and did so by compiling our own and
most other known adducts relevant to the current work. The cor-
relations between the deviation from ideal hapticity (Davg) and the
steric bulk, either ΘT or %Vbur, of the adduct ligand clearly
demonstrate that steric effects are the predominant forces
causing the ring slip in ionic metallocenes. Additionally, the simi-
larity with DFT calculated values (D, ΘT, ΘW, %Vbur) shows that
crystal-packing effects are minor in causing ring slip. This brings
about an interesting possible application of the correlation that
we have found in that a predictable degree of ring slip can be
designed into an ionic metallocene by tuning the steric bulk.
Considering the importance of molecular structure in controlling
the properties of catalysts or single molecular magnets, this type
of structural control could potentially find application in a variety
of fields where cyclopentadienide ligands are employed.
Experimental
General methods
All manipulations were performed under a dry, anaerobic
argon atmosphere using Schlenk line techniques or in a nitro-
Fig. 6 Representative steric maps for 1-L adducts. The x-axis and left
y-axis are in units of Å. See Fig. S11 and S12† for remaining maps.
This journal is © The Royal Society of Chemistry 2018
Dalton Trans.

View Article Online
Paper
Dalton Transactions
gen-filled VAC Atmosphere Genesis glovebox. The reagents Concentrated solutions of the adduct in THF stored at −35 °C
were purchased from commercial vendors and used without afforded XRD quality crystals. UV-vis (THF) (ε = M−1 cm−1): 420
further purification unless otherwise specified. Terpyridine, (891), 341 (1388). ATR-FTIR (cm−1): 3073, 3055, 1611, 1557,
4,7-diphenyl-1,10-phenanthroline, and 4,4′-dimethyl-2,2′-bipyr- 1441, 762. Anal. Calcd (found) for C22H22MnN2: %C, 71.54
idine were generously donated by Professor Jim Atwood (70.40); %H, 5.96 (5.83); %N, 7.58 (7.32). Carbon percentages
(University at Buffalo). 3 Å molecular sieves were activated by are consistently low despite several attempts.
heating ≥ 200 °C under vacuum (≈100 mTorr) for 48 h.
Synthesis of 1-6,6′-Me2bipy. The procedure and workup for
Anhydrous solvents were purified using Pure Process 1-6,6′-Me2bipy was the same as that for 1-bipy with the follow-
a
Technology solvent purification system and stored in a glove- ing modifications: 6,6′-dimethyl-2,2′-bipyridine (91 mg,
1
box over 3 Å molecular sieves for at least 24 h before use. H 0.49 mmol), 1 (90 mg, 0.49 mmol). Yield: 119 mg, 66%.
NMR spectra were recorded on a Varian Mercury-300 or Varian Concentrated solutions of the adduct in THF stored at −35 °C
Inova-400 MHz spectrometer. The values of chemical shifts afforded XRD quality crystals. UV-vis (THF) (ε = M−1 cm−1): 415
(ppm) are referenced to the residual solvent proton reso- (408). ATR-FTIR (cm−1): 3069, 3029, 1597, 1576, 1443, 791.
nances. UV-vis spectra were collected using an Agilent Cary Anal. Calcd (found) for C22H22MnN2: %C, 71.54 (71.31); %H,
8454 spectrophotometer. Solid ATR-FTIR spectra were collected 6.00 (6.41); %N, 7.58 (7.25). µeff (benzene, 298 K) = 5.0µB.
inside of an argon filled Omni VAC glovebox using a Bruker
Synthesis of 1-phen. The procedure and workup for 1-phen
Alpha IR spectrophotometer (PLATINUM-ATR insert module). was the same as that for 1-bipy with the following modifi-
The synthesis of MnCp2 was performed according to the pub- cations: 1,10-phenanthroline (95 mg, 0.53 mmol), 1 (95 mg,
lished protocol11 except that NaH, rather than Na metal, was 0.53 mmol). Yield: 132 mg, 68%. Concentrated solutions of
used to prepare NaCp.
the adduct in toluene stored at −35 °C afforded XRD quality
Synthesis of 6,6′-diphenyl-2,2′-bipyridine. This procedure is crystals. UV-vis (THF) (ε = M−1 cm−1): 435 (1552). ATR-FTIR
modified after a literature report for 6,6′-dimesityl-2,2′-bipyri- (cm−1): 3062, 1621, 1575, 1513, 1421, 727. Anal. Calcd (found)
dine.29 To a toluene suspension (30 mL) of 6,6′-dibromo-2,2′- for C22H18MnN2: %C, 72.33 (69.07); %H, 4.97 (5.09); %N, 7.67
bipyridine30 (0.433 g, 1.38 mmol), an excess of phenylboronic (6.81). Carbon percentages are consistently low despite several
acid (0.505 g, 4.14 mmol, ≥97% HPLC grade) dissolved in attempts.
5 mL of dimethylformamide (DMF) was added. A 10 mL
Synthesis of 1-4,7-Ph2phen. The procedure and workup for
sample of 2 M Na2CO3, Pd(OAc)2 (2 mol%), and triphenyl- 1-4,7-Ph2phen was the same as that for 1-bipy with the follow-
phosphine (10 mol%) were added to the reaction flask and the ing modifications: 4,7-diphenyl-1,10-phenanthroline (187 mg,
reaction mixture was refluxed for 24 h open to air, cooled to 0.56 mmol), 1 (102 mg, 0.55 mmol). Yield: 158 mg, 55%.
room temperature, and allowed to settle to separate the Concentrated solutions of the adduct in THF stored at −35 °C
aqueous and organic layers. The organic layer was washed afforded XRD quality crystals. UV-vis (THF) (ε = M−1 cm−1): 468
with brine (3 × 100 mL) and the aqueous layer was washed (1039). ATR-FTIR (cm−1): 3052, 3036, 1557, 1517, 1416, 760.
with chloroform (3 × 100 mL) and the organic fractions were Anal. Calcd (found) for C34H26MnN2: %C, 78.91 (74.91); %H,
combined and reduced to an oil on a rotary evaporator. The oil 5.06 (5.29); %N, 5.41 (4.45). Carbon percentages are consist-
was triturated with 30 mL hexane and the resulting solid was ently low despite several attempts.
filtered and washed with water (3 × 15 mL) and hexanes (3 ×
10 mL) and dried overnight under vacuum (product does 1-2,9-Me2phen was the same as that for 1-bipy with the follow-
sublime). Spectroscopic data matches the literature (CDCl3).31
ing modifications: 2,9-dimethyl-1,10-phenanthroline (107 mg,
Synthesis of 1-bipy. solution of MnCp2 (95 mg, 0.51 mmol), 1 (94 mg, 0.51 mmol). Yield: 115 mg, 57%.
Synthesis of 1-2,9-Me2phen. The procedure and workup for
A
0.51 mmol) in THF (5 mL) was added dropwise to a solution of Concentrated solutions of the adduct in THF stored at −35 °C
2,2′-bipyridine (76 mg, 0.49 mmol) in THF (5 mL) at room afforded XRD quality crystals. UV-vis (THF) (ε = M−1 cm−1): 453
temperature. The reaction mixture turned dark red/brown and (341.36), 328 (1064). ATR-FTIR (cm−1): 3046, 1591, 1560, 1501,
was stirred at room temperature for 45 minutes and then 1424, 731. Anal. Calcd (found) for C24H22MnN2: %C, 73.28
recrystallized by storing at −35 °C. The crystals obtained at (73.58); %H, 5.64 (5.68); %N, 7.12 (7.08). µeff (benzene, 298 K)
−35 °C were filtered cold and washed once with 5 mL of cold = 6.1µB.
petroleum ether (117 mg, 70%). Saturated solutions of the
Synthesis of 1-py2. The procedure and workup for 1-py2 was
adduct in toluene stored at −35 °C afforded XRD quality crys- the same as that for 1-bipy with the following modifications:
tals. UV-vis (THF, nm (ε, M−1 cm−1)): 417 (654), 343 (972). pyridine (80 μL, 1 mmol), 1 (90 mg, 0.49 mmol), bright yellow
ATR-FTIR (cm−1): 3069, 1593, 1576, 1556, 747. Anal. Calcd solution. Yield: 92 mg, 55%. Concentrated solutions of the
(found) for C20H18MnN2: %C, 70.38 (69.29); %H, 5.28 (5.22); adduct in toluene stored at −35 °C afforded XRD quality crys-
%N, 8.21 (8.74). Carbon percentages are consistently low tals. UV-vis (THF) (ε = M−1 cm−1): featureless, shoulder
despite several attempts.
300 nm. ATR-FTIR (cm−1): 3068, 3055, 1599, 1570, 1485, 1441,
Synthesis of 1-4,4′-Me2bipy. The procedure and workup for 759. Anal. Calcd (found) for C20H20MnN2: %C, 69.97 (69.93);
1-4,4′-Me2bipy was the same as that for 1-bipy with the follow- %H, 5.87 (5.90); %N, 8.16 (8.36). µeff (benzene, 298 K) = 5.1µB.
ing modifications: 4,4′-dimethyl-2,2′-bipyridine (90 mg,
Synthesis of 1-terpy. The procedure and workup for 1-terpy
0.49 mmol), 1 (90 mg, 0.49 mmol). Yield: 30 mg, 17%. was the same as that for 1-bipy with the following modifi-
Dalton Trans.
This journal is © The Royal Society of Chemistry 2018

View Article Online
Dalton Transactions
Paper
cations: terpyridine (95 mg, 0.41 mmol) in toluene (5 ml) was 10 M. E. Switzer and M. F. Rettig, J. Chem. Soc., Chem.
added dropwise to a solution of 1 (75 mg, 0.41 mmol) in Commun., 1972, 687–688.
toluene (5 ml). Yield: 150 mg, 87%. Vapor diffusion of hexane 11 G. Wilkinson, F. A. Cotton and J. M. Birmingham, J. Inorg.
into a saturated solution of 1-terpy in THF afforded crystals Nucl. Chem., 1956, 2, 95–113.
suitable for XRD. UV-vis (THF) (ε = mol−1 cm−1): 235 (26 680), 12 (a) A. D. Bond, R. A. Layfield, J. A. MacAllister,
278 (17 958). ATR-FTIR (cm−1): 3056, 1593, 1570, 1474, 770.
Anal. Calcd (found) for C25H21MnN3: %C, 71.77 (71.54); %H,
5.06 (5.53); %N, 10.04 (10.15).
%Vbur calculation. The SambVca 2.0 web application was
used to calculate the %Vbur and also generate the steric maps.
M. McPartlin, J. M. Rawson and D. S. Wright, Chem.
Commun., 2001, 1956–1957; (b) C. S. Alvarez, A. Bashall,
E. J. McInnes, R. A. Layfield, R. A. Mole, M. McPartlin,
J. M. Rawson, P. T. Wood and D. S. Wright, Chem. – Eur. J.,
2006, 12, 3053–3060.
The xyz coordinate files were created using IQmol v2.7.1 for 13 F. Bottomley, P. N. Keizer and P. S. White, J. Am. Chem.
DFT structures or Mercury v3.9 for XRD determined structures. Soc., 1988, 110, 137–140.
The sphere radius we employed was 3.5 Å and chosen based 14 (a) C. D. Abernethy, A. H. Cowley, R. A. Jones,
on the original work by Cavallo and Nolan.27 Hydrogen atoms
C. L. B. Macdonald, P. Shukla and L. K. Thompson,
Organometallics, 2001, 20, 3629–3631; (b) M. D. Walter,
D. Baabe, M. Freytag and P. G. Jones, Z. Anorg. Allg. Chem.,
2016, 642, 1259–1263.
were included in the calculation of the %Vbur
.
15 J. Heck, W. Massa and P. Weinig, Angew. Chem., Int. Ed.
Engl., 1984, 23, 722–723.
Conflicts of interest
The authors declare no competing financial interest.
16 C. G. Howard, G. S. Girolami, G. Wilkinson,
M. Thorntonpett and M. B. Hursthouse, J. Am. Chem. Soc.,
1984, 106, 2033–2040.
17 J. W. Lauher and R. Hoffmann, J. Am. Chem. Soc., 1976, 98,
1729–1742.
Acknowledgements
Financial support was provided by the University of Buffalo
(UB) start-up funds and an ACS Petroleum Research Fund
(ACS-PRF-57861-DN13). The authors would like to thank the
UB Center for Computational Research for access to compu-
tation facilities. The authors thank Kelly Smeeth for insightful
discussions.
18 (a) C. Soria Alvarez, A. Bashall, A. D. Bond, D. Cave,
E. A. Harron, R. A. Layfield, M. E. G. Mosquera,
M. McPartlin, J. M. Rawson, P. T. Wood and D. S. Wright,
Dalton Trans., 2003, 3002–3008; (b) C. S. Alvarez, S. R. Boss,
J. C. Burley, S. M. Humphry, R. A. Layfield, R. A. Kowenicki,
M. McPartlin, J. M. Rawson, A. E. Wheatley, P. T. Wood and
D. S. Wright, Dalton Trans., 2004, 3481–3487.
19 J. W. Faller, R. H. Crabtree and A. Habib, Organometallics,
1985, 4, 929–935.
20 C. A. Tolman, J. Am. Chem. Soc., 1970, 92, 2956–2965.
Notes and references
1 G. Wilkinson, M. Rosenblum, M. C. Whiting and 21 C. A. Tolman, Chem. Rev., 1977, 77, 313–348.
R. B. Woodward, J. Am. Chem. Soc., 1952, 74, 2125–2126.
2 R. R. Gagne, C. A. Koval and G. C. Lisensky, Inorg. Chem.,
1980, 19, 2854–2855.
22 T. E. Muller and D. M. P. Mingos, Transition Met. Chem.,
1995, 20, 533–539.
23 S. S. Batsanov, Inorg. Mater., 2001, 37, 871–885.
3 H. H. Brintzinger, D. Fischer, R. Mülhaupt, B. Rieger and 24 A. Glockner, H. Bauer, M. Maekawa, T. Bannenberg,
R. M. Waymouth, Angew. Chem., Int. Ed. Engl., 1995, 34,
1143–1170.
4 G. Gasser, I. Ott and N. Metzler-Nolte, J. Med. Chem., 2011,
54, 3–25.
5 J. M. O’Connor and C. P. Casey, Chem. Rev., 1987, 87, 307–
318.
6 W. Simanko, V. N. Sapunov, R. Schmid, K. Kirchner and
S. Wherland, Organometallics, 1998, 17, 2391–2393.
C. G. Daniliuc, P. G. Jones, Y. Sun, H. Sitzmann,
M. Tamm and M. D. Walter, Dalton Trans., 2012, 41, 6614–
6624.
25 (a) A. Poater, B. Cosenza, A. Correa, S. Giudice, F. Ragone,
V. Scarano and L. Cavallo, Eur. J. Inorg. Chem., 2009, 2009,
1759–1766; (b) L. Falivene, R. Credendino, A. Poater,
A. Petta, L. Serra, R. Oliva, V. Scarano and L. Cavallo,
Organometallics, 2016, 35, 2286–2293.
7 (a) G. Huttner, H. H. Brintzinger, L. G. Bell, P. Friedrich, 26 (a) A. C. Hillier, W. J. Sommer, B. S. Yong, J. L. Petersen,
V. Bejenke and D. Neugebauer, J. Organomet. Chem., 1978,
145, 329–333; (b) E. U. van Raaij, H. H. Brintzinger,
L. Zsolnai and G. Huttner, Z. Anorg. Allg. Chem., 1989, 577,
217–222.
L. Cavallo and S. P. Nolan, Organometallics, 2003, 22, 4322–
4326; (b) A. Poater, F. Ragone, S. Giudice, C. Costabile,
R. Dorta, S. P. Nolan and L. Cavallo, Organometallics, 2008,
27, 2679–2681.
8 (a) L. F. Veiros, Organometallics, 2000, 19, 5549–5558; 27 H. Clavier and S. P. Nolan, Chem. Commun., 2010, 46, 841–
(b) M. J. Calhorda, C. C. Romao and L. F. Veiros, Chem. –
Eur. J., 2002, 8, 868–875.
9 R. A. Layfield, Chem. Soc. Rev., 2008, 37, 1098–1107.
861.
28 A. Poater, F. Ragone, R. Mariz, R. Dorta and L. Cavallo,
Chem. – Eur. J., 2010, 16, 14348–14353.
This journal is © The Royal Society of Chemistry 2018
Dalton Trans.

View Article Online
Paper
Dalton Transactions
29 (a) M. Schmittel, A. Ganz, W. A. Schenk and M. Hagel, Z. 30 K. Kamata, A. Suzuki, Y. Nakai and H. Nakazawa,
Naturforsch., B: J. Chem. Sci., 1999, 54, 559–564; Organometallics, 2012, 31, 3825–3828.
(b) M. D. Sampson, A. D. Nguyen, K. A. Grice, C. E. Moore, 31 D. E. Stephens, J. Lakey-Beitia, J. E. Burch, H. D. Arman
A. L. Rheingold and C. P. Kubiak, J. Am. Chem. Soc., 2014,
136, 5460–5471.
and O. V. Larionov, Chem. Commun., 2016, 52, 9945–
9948.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2018