Electrochemistry of transition metal hydride diphosphine complexes trans-MH(X)(PP)2 and trans-[MH(L)(PP)2]+, M = Fe, Ru, Os; PP = chelating phosphine ligand

Article abstract of DOI:10.1016/j.ica.2020.120124

A series of over 30 iron, ruthenium, and osmium hydride phosphine complexes are reported, along with their M^III/II redox potentials. The complexes are of the type MH(PP)_n(X) or [MH(PP)_n(L)]⁺, where PP is one of

Full text of DOI:10.1016/j.ica.2020.120124

Inorganica Chimica Acta xxx (xxxx) xxx
Contents lists available at ScienceDirect
Inorganica Chimica Acta
journal homepage: www.elsevier.com/locate/ica
Research paper
Electrochemistry of transition metal hydride diphosphine complexes
trans-MH(X)(PP)2 and trans-[MH(L)(PP)2]+, M = Fe, Ru, Os; PP = chelating
phosphine ligand
Samantha D. Drouin, Patricia A. Maltby, Benjamin E. Rennie, Caroline T. Schweitzer,
Adina Golombek, E. Paul Cappellani, Robert H. Morris*
The Department of Chemistry, University of Toronto, 80 Saint George St., Toronto, ON M5S 3H6, Canada
A R T I C L E I N F O
A B S T R A C T
A series of over 30 iron, ruthenium, and osmium hydride phosphine complexes are reported, along with their
MIII/II redox potentials. The complexes are of the type MH(PP)n(X) or [MH(PP)n(L)]+, where PP is one of the
following bidentate phosphine ligands: dppe, dtpe, depe, and dtfpe, with n = 2; or the tetradentate phosphine
ligand meso-tet-1, with n = 1. The electrochemical data of these complexes and those from the literature are used
to determine the Lever EL parameter of ꢀ 0.65 V for the hydride ligand for iron and ruthenium. For osmium,
however, the EL value for the hydride ligand is found to be more positive at only ꢀ 0.37 V, an increase which is
Dedicated to Professor Maurizio Peruzzini
Keywords:
Transition metals
Coordination chemistry
Paramagnetism
Synthesis
attributed to Os-H
σ bond strengthening due to relativistic effects. The correlation holds for irreversible oxida-
tions as well as reversible ones. These EL values can now be used along with Lever’s equations to predict redox
Hydride
Electrochemistry
Mechanism
potentials of other iron-group hydride complexes.
Acidity
1. Introduction
derstanding of its chemical reactions [3,13]. Pickett and co-workers
found that the E1/2 ox of Cr(CO)5L was very sensitive to the nature of L
There has been considerable interest in correlating electrochemical
potentials with the effects of ligand substitution [1–10]. Electrochemical
potentials have been related to a variety of properties of transition metal
complexes such as charge-transfer transition energies, infrared stretch-
ing frequencies of ligands, HOMO energies ligand substitution effects
[1] and the stability of dihydrogen complexes [4]. However, the elec-
trochemistry of transition metal hydride complexes has not been studied
systematically. Understanding the electronic nature of these hydride
complexes is an important handle in the rational development of cata-
lysts for such processes as carbon dioxide reduction and hydrogen
evolution [11,12]. In this paper the electrochemical properties of a se-
ries of metal hydride complexes, formulated as trans-MH(X)(PP)2 and
trans-[MH(L)(PP)2]+ (M = Fe, Ru and Os; PP = dppe, depe, dtpe, dtfpe,
1/2 meso-tet-1; X = anionic ligands and L = neutral monodentate li-
gands) will be reported. This data will be used in conjunction with
literature data to determine the Lever electronic parameter for the hy-
dride ligand for Fe, Ru and Os complexes.
and they developed a series of ligand constants, PL, by the use of Eq. (1)
[14].
ox
ox
PL = E [Cr(CO)5L] ꢀ E [Cr(CO)6]
(1)
1/2
1/2
The ligand parameters, PL, represent the change in energy of the
HOMO of Cr(CO)6 when one CO ligand is replaced by a ligand, L. In
some cases, in addition to HOMO energies, structural re-organization
energies would have to be included as well. Ligands with the weakest
σ
-donor and strongest
most positive PL constants while strong
such as the halides and pseudo-halides have the most negative PL. Thus
π
-acceptor properties such as CO and N2 have the
σ
-donor and -donor ligands
π
PL appears to reflect the net donor properties (σ -and π - donor + π
-acceptor) of the ligand L. Pickett also applied the PL parameter to other
ox
metal systems by graphing E
versus PL for a series of complexes
1/2
[ML’5L]. The linear correlation produced is shown in eq (2), where Es
ox
equals to the E1/2 of the complex M(CO)6 and ß is the slope of the E vs.
1/2
PL plot [14,15].
The electrochemical potential of a compound is valuable in the un-
* Corresponding author.
E-mail address: RMORRIS@CHEM.UTORONTO.CA (R.H. Morris).
https://doi.org/10.1016/j.ica.2020.120124
Received 23 August 2020; Received in revised form 1 November 2020; Accepted 2 November 2020
Available online 7 November 2020
0020-1693/© 2020 Elsevier B.V. All rights reserved.
Please cite this article as: Samantha D. Drouin, Inorganica Chimica Acta, https://doi.org/10.1016/j.ica.2020.120124

S.D. Drouin et al.
Inorganica Chimica Acta xxx (xxxx) xxx
Table 1
Selected EL values. Unless specified, the data is from ref. [2].a
Ligand
EL (V)
Ligand
EL (V)
Ligand
EL (V)
Ligand
EL (V)
Ligand
EL (V)
Ligand
EL (V)
CO
0.99
0.68
0.58
0.52
P(OMe)3
meso-tet-1b
dtpe b
0.42
0.35
0.34
0.36
depe b
PMe2Ph
MeCN
PMe3
0.27
0.34
0.34
0.33
MeIm
CNꢀ
0.08
0.02
Clꢀ
ꢀ 0.24
ꢀ 0.25
ꢀ 0.30
ꢀ 0.37
SPhꢀ
ꢀ 0.53
ꢀ 0.55
ꢀ 0.65
N2
NCOꢀ
N3ꢀ
StBuꢀ
P(OPh)3
NCSꢀ c
Brꢀ
ꢀ 0.06
Hꢀ (Fe, Ru) b
dtfpe b
dppe
ꢀ 0.22
Hꢀ (Os) b
a See table S4 for a complete list. b EL value is derived here. c The EL for NCSꢀ refers to the nitrogen-bound ligand, which is applicable for the complexes studied here.
See ref. [16] for a crystal structure of nitrogen-bound NCSꢀ of the related iron complex Fe(NCS)2(PPh2(CH2)2PPh(CH2)3PPh(CH2)2PPh2)
from the Picket/Pletcher ligand parameter, might be too low [2]. The EL
Table 2
of the hydride was calculated to be ꢀ 0.76 V using the computed ligand
The slope, intercept data, and equations for E1/2(d5/d6) (V vs. NHE) versus ΣEL
electronic parameter (CEP) method, via an equation that relates the CEP
for low spin six coordinate iron group complexes in organic solvents.
to the Lever electronic parameter [8]. In our work, experimental data is
Couple
Source
Slope
Intercept
IM/SD
R2
E1/2
=
EL of
used to calculate an EL value for the hydride ligand based on iron-group
data. Up to now, the electrochemical behavior of ReCl(L)(diphos)2,
SM/SD
S
MΣEL + IM
Hydride /
Error
[MoXL(diphos)2]n+
,
[RuL(NH3)5]n+
,
[ReLN2(dppe)2]2+
,
[FeH(L)
FeIII/II
RuIII/II
OsIII/II
FeIII/II
RuIII/II
OsIII/II
Lever a
Lever
Lever
1.10 /
0.05
ꢀ 0.43 /
0.99
0.99
0.98
0.91
0.84
0.88
Ecalc
=
–
–
–
(dppe)2]+, and multi- and mono- nuclear complexes of the type [FeH
(CN)(dppe)2] with a bridging cyanide or with benzoyl isocyanide, have
been examined by Pickett et al. [15,17] and Almeida et al. [18]
respectively. Here we extend the study to other ditertiaryphosphines as
well as to other transition metals Ru and Os.
0.12
1.10ΣEL
ꢀ
+
ꢀ
ꢀ
+
–
0.43
0.97 /
0. 01
0.04 /
0.03
Ecalc =
0.97ΣEL
0.04
1.01 /
0.02
ꢀ 0.40 /
Ecalc =
0.11
1.01ΣEL
2. Experimental
0.40
this
1.08 /
0.05
ꢀ 0.46 /
Eobs
=
ꢀ 0.65 /
work b
0.05
1.08ΣEL
0.02
2.1. Materials and methods
0.46
this
0.77 /
0.05
0.15 /
0.03
Eobs
=
ꢀ 0.65 /
Abbreviations: MeIm = 1-methylimidazole, dppe = 1,2-bis(diphe-
nylphosphino)ethane, depe = 1,2-bis(diethylphosphino)ethane, dtpe =
work
0.77ΣEL
0.02
0.15
this
1.04 /
0.10
ꢀ 0.44 /
Eobs
=
ꢀ 0.37 /
1,2-bis(di(para-tolyl)phosphino)ethane,
ylphosphines, meso-
diphos
=
bistertiar-
work
0.11
1.04ΣEL
0.04
0.44
tet-1 = (R,S)-PPh2(CH2)2P(Ph)(CH2)2P(Ph)(CH2)2PPh2, dtfpe = 1,2-
a Lever refers to Lever’s equations discussed in ref. [2]. b Refers to the hydride
bis(di(para-trifluorotolyl)phosphino)ethane, bpy
=
2,2′-bipyridine,
plots discussed in this work.
THF = tetrahydrofuran, MeOH = methanol, CH2Cl2 = dichloromethane,
NBA = 4-nitrobenzyl alcohol, NPOE = 2-nitrophenyloctylether, FAB MS
= Fast atom bombardment mass spectrometry, Epa = anodic peak po-
tential of an irreversible oxidation wave; Fc+/Fc = ferrocenium/ferro-
cene couple, NHE = Normal Hydrogen Electrode.
ox
E
= Es + ßPL
(2)
1/2
A more general additive ligand method has been proposed by Lever
based on the change in the ruthenium(III)/ruthenium(II) couple as a
function of bound ligands in organic solvents [2]. He defined an additive
ligand parameter EL, which represents a separate contribution for each
ligand of the complex. For example, the EL contribution of bpy in the
complex [Ru(bpy)3]2+ (E1/2 = 1.53 V vs. NHE) is defined to be 1.53/6 or
0.255 V because the complex contains six identical Ru-N bonds. In this
fashion the EL values for over 200 ligands were defined (Table 1 lists
some of these values in order of decreasing potential which will be used
in this study). Therefore, the electrochemical potential of the general
complex, RuXxYyZz, can be calculated by using the EL values as in Eq. (3)
or its abbreviated form [2].
All operations were conducted in an argon atmosphere by use of
Schlenk line techniques, or in a purified nitrogen atmosphere using a
Vacuum Atmospheres glove box. All solvents were dried over appro-
priate reagents and distilled under N2 before use. Benzene, THF, diethyl
ether, and hexanes were dried over and distilled from sodium-
benzophenone ketyl. MeOH was dried over magnesium methoxide,
and ethanol over magnesium ethoxide. Acetone was dried over potas-
sium carbonate. CH2Cl2 was distilled from calcium hydride. Deuterated
solvents were dried over Linde type 4 Å molecular sieves and degassed
with argon using several evacuate/refill cycles. Reagent-grade chem-
icals were used as purchased from Aldrich Chemical Company, Inc.
unless otherwise stated. Phosphine ligands, dppe, depe and
Cl2PCH2CH2PCl2 were purchased from Strem Chemical Co. or Digital
Specialty Chemicals Ltd. Sodium thiophenoxide, NaSPh, was prepared
in 93% yield from thiophenol and sodium hydride using a Et2O/THF (7/
3 v/v) solvent and 1.75 h reaction time; the white solid was filtered and
washed with Et2O. Sodium thio-tert-butoxide was prepared from HStBu
and NaH in THF. Infrared spectra were recorded as Nujol mulls on NaCl
plates using a Nicolet 5DX FTIR spectrometer. NMR spectra were
recorded on Varian XL–400 (400 MHz for 1H, 162 MHz for 31P), Varian
Gemini 200 (200 MHz for 1H) or Varian Gemini 300 (300 MHz for 1H,
120.5 MHz for 31P) spectrometers. Chemical shifts refer to room tem-
perature conditions (20 ◦C) unless specified otherwise. 31P chemical
shifts were measured in the proton decoupled mode and referenced
either internally or externally to 1% P(OMe)3 in C6D6 sealed in coaxial
capillaries, but are reported relative to 80% H3PO4 by use of δ (P
(OMe)3) = 140.4 ppm. 1H chemical shifts were measured relative to
Ecalc = xEL(X) + yEL(Y) + zEL(Z) = ΣEL
(3)
Lever has also demonstrated that the EL ligand parameters can be
used for many other metal ions in a given oxidation state [2]. Plots of the
∑
observed potentials for certain complexes (MXxYyZz) against EL were
linear as expressed in Eq. (4).
Eobs = SM[ΣEL] + IM
(4)
The slope, SM and intercept, IM data for the metal ion being studied
should reflect the specific M(n)/M(n-1) couple. The values of the slope
(SM) and intercept (IM) constants determined by Lever for the M(III)/M
(II) couple of the iron group d6 complexes MXxYyZz, M = Fe, Ru, Os are
summarized in Table 2. Also in table 2 is the hydride data from this
work, which is discussed later. These slope and intercept data can be
used to predict unknown electrochemical potentials. Conversely,
observed potentials may be used to calculate as yet unknown EL values.
Lever suggested that the EL of the hydride, ꢀ 0.30, which was derived
2

S.D. Drouin et al.
Inorganica Chimica Acta xxx (xxxx) xxx
Table 3
Lever additive electrochemical parameters, ΣEL, calculated E1/2(d5/d6) and observed E1/2 (or Epa) values for the complexes trans-[MH(L)(PP)2]+ and trans-MH(X)(PP)2;
THF solvent, 0.2 M NBu4PF6, 250 mV sꢀ 1 scan rate.
No.
Metal
Bisphosphine
L or Xꢀ
ΣEaL
Ecalc V vs. NHE b
Eobs V vs. Fc+/Fc c
Eobs V vs. NHE d
1a
1b
1c
1d
1e
1f
Fe
Fe
Fe
Fe
Fe
Fe
Fe
Fe
Fe
Fe
Fe
Fe
Fe
Fe
Fe
Fe
Fe
Fe
Fe
Ru
Ru
Ru
Ru
Ru
Ru
Ru
Ru
Ru
Ru
Ru
Ru
Ru
Ru
Ru
Os
Os
Os
Os
Os
Os
Os
Os
depe
N3ꢀ
0.13
ꢀ 0.29
ꢀ 0.23
ꢀ 0.22
ꢀ 0.20
ꢀ 0.02
0.79
1.13
0.09
0.08
0.29
0.37
0.73
1.01
1.44
0.87
1.08
1.17
1.52
1.80
ꢀ 0.08
ꢀ 0.17
0.21
0.22
0.24
0.40
0.79
0.14
0.53
0.85
1.10
1.10
0.89
1.37
1.76
0.13
0.15
0.46
0.76
1.01
1.04
1.10
1.27
ꢀ 0.98
ꢀ 0.96
ꢀ 0.98
ꢀ 0.95
ꢀ 0.75
0.31
ꢀ 0.33
ꢀ 0.31
ꢀ 0.33
ꢀ 0.30
ꢀ 0.10
0.96
depe
NCOꢀ
Clꢀ
0.18
0.19
0.21
0.37
1.11
1.42
0.47
0.46
0.65
0.73
1.05
1.31
1.70
1.18
1.37
1.45
1.77
2.03
ꢀ 0.12
ꢀ 0.22
0.18
0.19
0.21
0.37
0.77
0.10
0.51
0.83
1.09
1.09
0.88
1.37
1.77
0.52
0.54
0.85
1.15
1.40
1.43
1.49
1.65
depe
depe
Brꢀ
depe
NCSꢀ
N2
depe
1h
1j
depe
CO
0.76
1.41
dtpe
NCOꢀ
Clꢀ
ꢀ 0.68
ꢀ 0.72
ꢀ 0.43
ꢀ 0.32
ꢀ 0.06
(0.65)
(0.43)
(0.24)
(0.36)
(0.38)
(0.53)
(-0.03) e
ꢀ 0.90
ꢀ 0.50
ꢀ 0.40
ꢀ 0.30
ꢀ 0.38
ꢀ 0.27
0.27
ꢀ 0.03
ꢀ 0.07
0.22
1k
1l
dtpe
dtpe
NCSꢀ
1m
1n
1o
1p
1r
dtpe
CNꢀ
0.33
dtpe
NCMe
H2
0.59
dtpe
(1.30)
(1.08)
(0.89)
(1.01)
(1.03)
(1.18)
(0.62)
ꢀ 0.25
0.15
dtpe
CO
dtfpe
dtfpe
dtfpe
dtfpe
dtfpe
depe
NCOꢀ
NCSꢀ
CNꢀ
1t
1u
1v
1w
2a
2b
2c
2d
2e
2f
MeCN
H2
StBuꢀ
Hꢀ
depe
depe
NCOꢀ
Clꢀ
0.25
depe
0.35
depe
Brꢀ
0.27
depe
NCSꢀ
NCMe
Hꢀ
0.38
2g
2h
2i
depe
0.92
meso-tet-1
meso-tet-1
meso-tet-1
meso-tet-1
meso-tet-1
dtfpe
dtfpe
dtfpe
dppe
ꢀ 0.47
ꢀ 0.35
ꢀ 0.10
0.42
0.18
Clꢀ
0.30
2j
MeIm
PMe2Ph
NCMe
StBuꢀ
NCSꢀ
MeCN
StBuꢀ
SPhꢀ
Brꢀ
0.55
2k
2l
1.07
0.46
1.11
2m
2o
2p
3a
3b
3e
3f
0.48
1.13
(0.67)
(0.59)
ꢀ 0.77
ꢀ 0.65
ꢀ 0.11
0.31
(1.24)
(1.32)
ꢀ 0.12
0.00
dppe
dppe
0.54
dppe
MeIm
PMe3
MeCN
P(OMe)3
P(OPh)3
0.96
3g
3h
3i
dppe
0.68f
1.33
dppe
0.58f
1.23
dppe
(0.97)
(1.62)
3j
dppe
>1.2 g
>1.85
aΣEL = 4 EL(PP) + EL(L) + EL(Hꢀ ). b Lever’s calculated equations for Fe, Ru and Os are listed in Table 2. c ±0.05 V. Values in brackets represent irreversible processes
which are reported as Epa peak potentials. d Converted from Fc+/Fc reference to the NHE reference by adding 0.65. e Epa value in CH2Cl2 solvent is 0.48 V vs. Fc+/Fc, or
1.13 V vs NHE. f CH2Cl2 solvent. g No OsIII/II redox processes were observed in the electrochemical window.
partially deuterated solvent peaks, but are reported relative to tetra-
methylsilane. FAB-MS spectra were obtained with a VG 70-250S mass
spectrometer using a NBA or NPOE matrix. Microanalyses were per-
formed on a sample handled under an inert atmosphere by Canadian
Microanalytical Services, Vancouver. Electrochemical Apparatus: Elec-
trochemical measurements were made on a Princeton Applied Research
Model 273 Potentiostat/Galvanostat (without IR compensation). All
cyclic voltammograms were obtained using an electrochemical cell
consisting of a platinum working electrode, tungsten secondary elec-
trode and silver wire reference electrode in a Luggin-Haber probe
capillary. All measurements were made under Ar gas using 0.2 M
NBu4PF6 in freshly distilled THF or CH2Cl2 as the electrolyte solution.
The cyclic voltammograms were collected as two cycles of negative to
positive to negative potential sweeps. The applied potential was varied
over a 2 V range: ꢀ 0.8 to 1.2 V vs. Ag+/Ag for THF solutions, or ꢀ 0.3 to
1.7 V vs. Ag+/Ag for CH2Cl2 solutions. The E1/2 and Epa values are re-
ported relative to the ferrocenium/ferrocene couple. Initially, a back-
ground scan of 6 mL of the electrolyte solution was made in order to
insure that no oxygen or water was present. Under a positive pressure of
argon, approximately 10 mg of the complex were added to the electro-
lyte. A small amount of solution was drawn up around the reference
electrode and the argon was turned off. The electrochemistry cell was
enabled and the acquisition program started. After the initial measure-
ment, a small amount of ferrocene was added and the process was
repeated. All the E1/2 values for the complexes trans-MH(X)(PP)2 and
trans-[MH(L)(PP)2]+ are summarized in Table 3.
2.2. Synthesis
Preparations of the ligands dtpe [19], meso-tet-1 [20] and dtfpe [19]
have been reported. (NH4)2[OsBr6] was prepared from OsO4 according
to the literature method [21]. The following complexes were prepared
by previously reported methods: FeCl2(depe)2 [22], trans-FeHCl(PP)2
(1k: PP = dtpe [23]; 1s: dtfpe [23]; 1c: depe [24]), Ru(cod)(cot) [25],
cis- and trans-Ru(H)2(depe)2 [26], trans-RuH(Cl)(depe)2 [24], RuH(X)
(meso-tet-1) (2 h: Xꢀ = Hꢀ ; 2i: Xꢀ = Clꢀ ) and [RuH(L)(meso-tet-1)]+ (2j:
L = MeIm; 2k: L = PMe2Ph; 2 l: L = NCMe) [20].
trans-FeH(X)(depe)2, 1a, 1b, 1d, 1e: The complex FeHCl(depe)2 (0.2
g, 0.4 mmol) and a slight excess of NaX (Xꢀ = N3ꢀ , NCOꢀ , NCSꢀ , Brꢀ )
were stirred overnight in acetone (20 mL). The fine sodium salts that
formed were filtered off through Celite and the filtrate taken to dryness.
The solid was dissolved in benzene (5 mL) and hexanes (5 mL) added.
The volume of this solution was reduced under vacuum and placed in the
refrigerator. The solid that formed was filtered off and the filtrate was
3

S.D. Drouin et al.
Inorganica Chimica Acta xxx (xxxx) xxx
slowly taken to dryness to leave the product as needles. trans-FeH(N3)
(depe)2 (1a): 1H NMR(C6D6): ꢀ 30.8 ppm (quint, 2JHP = 47.6 Hz, FeH);
31P{1H} NMR(THF) 89.5 ppm (s). FAB MS calc. for C20H49FeN3P4 : 511;
obs. 511 (M+). trans-FeH(NCO)(depe)2 (1b): 1H NMR(acetone‑d6):
ꢀ 28.7 ppm (quint, 2JHP = 47.1 Hz, FeH); 31P{1H} NMR(THF) 89.5 ppm
(s). FAB MS calc. for C21H49FeNOP4 : 511; obs. 511 (M+). trans-FeH(Br)
(depe)2 (1d): 1H NMR(acetone‑d6): ꢀ 31.5 ppm (quint, 2JHP = 50.9 Hz,
FeH); 31P{1H} NMR(THF) 85.2 ppm (s). FAB MS: calc. for C20H49FeP4Br
: 548; obs. 548 (M+). trans-FeH(NCS)(depe)2 (1e): 1H NMR(acetone‑d6):
ꢀ 25.7 ppm (quint, 2JHP = 47.3 Hz, FeH); 31P{1H} NMR(THF) 89.1 ppm
(s). FAB MS calc. for C21H49FeNP4S : 527; obs. 527 (M+).
trans-[FeH(CO)(dtfpe)2]BF4 , 1x: The pale yellow powder, trans-[FeH
(H2)(dtfpe)2]BF4 , was placed in a flask under argon. The flask was
evacuated and heated in vacuum for 3 min with a heat gun at approx-
imately 170 ◦C. The pale yellow solid turned navy blue. The powder was
then placed under 1 atm CO(g) and the color immediately changed to
pale yellow (100% yield). IR: 1967 (
ν
CO). 1H NMR(acetone‑d6): ꢀ 7.64
ppm (quint, 2JHP = 47.3 Hz, FeH); 31P{1H} NMR(acetone‑d6) 90.7 ppm
(s). FAB MS calc. for C61H41F24FeOP4: 1425; obs. 1425 (M+).
trans-RuHX(depe)2, 2a, 2c, 2e, 2f: The complexes 2a, 2c, 2e, 2f were
obtained by the same procedure as described for the preparation of 1a
by using RuHCl(depe)2 as starting material. trans-RuH(StBu)(depe)2, 2a:
trans-[FeH(L)(depe)2]BF4 (1f, 1h): The complexes were prepared by
the method of Bancroft and Mays [27] except that NaBF4 was used
instead of NaBPh4. trans-[FeH(N2)(depe)2]BF4 (1f): 1H NMR(ace-
tone‑d6): ꢀ 18.2 ppm (quint, 2JHP = 47.3 Hz, FeH); 31P{1H} NMR(THF)
80.8 ppm (s). trans-[FeH(CO)(depe)2]BF4 (1h): 1H NMR(acetone‑d6):
ꢀ 10.9 ppm (quint, 2JHP = 47.8 Hz, FeH); 31P{1H} NMR(THF) 65.7 ppm
(s). FAB MS calc. for C21H49FeOP4: 497; obs. 497 (M+).
yellow solid, yield 80%. 1H NMR(acetone‑d6): ꢀ 18.4 ppm (quint, 2JHP
=
21.1 Hz, RuH); 31P{1H} NMR(THF) 61.8 ppm (s). FAB MS: calcd for
C
24H58P4RuS: 604; obs.: 547 (M+ꢀ tBu), 546 (M+ꢀ tBu,H). trans-RuH
(NCO)(depe)2, 2c: yellow solid, yield 80%. 1H NMR(acetone‑d6): ꢀ 17.8
ppm (quint, JHP = 18.9 Hz, RuH); 31P{1H} NMR(THF) 65.2 ppm (s).
2
FAB MS: calcd for C21H49NOP4Ru: 557; obs.: 556 (M+ꢀ H), 515 (M+ꢀ H,
1
NCO). trans-RuHBr(depe)2, 2e: pale yellow solid, yield 80%. H NMR
2
trans-FeH(X)(dtpe)2, 1j, 1l, 1m: The complexes 1j-1m were obtained
by the same procedure as described in preparation of 1a by using FeHCl
(dtpe)2 as the starting material. trans-FeH(NCO)(dtpe)2, 1j: orange solid,
(acetone‑d6): ꢀ 20.0 ppm (quint, JHP = 19.6 Hz, RuH); 31P{1H} NMR
(THF) 61.7 ppm (s). FAB MS calcd for C20H49BrP4Ru: 594; obs.: 594
(M+), 515 (M+ꢀ Br). trans-RuH(NCS)(depe)2, 2f: yellow solid, yield
2
yield 60%. 1H NMR(acetone‑d6): ꢀ 26.5 ppm (quint, JHP = 46.3 Hz,
80%. 1H NMR(acetone‑d6): ꢀ 19.0 ppm (quint, 2JHP = 19.5 Hz, RuH); 31
P
FeH); 31P{1H} NMR(THF) 82.9 ppm (s). FAB MS calc. for C61H65Fe-
{1H} NMR(THF) 64.7 ppm (s). FAB MS calcd for C21H49NP4RuS: 573;
obs.: 572 (M+ꢀ H), 515 (M+ꢀ HNCS).
NOP4: 1007; obs. 1007 (M+). trans-FeH(NCS)(dtpe)2, 1l: orange solid,
yield 59%. 1H NMR(C6D6): ꢀ 24.4 ppm (quint, 2JHP = 43.9 Hz, FeH); 31
P
trans-[RuH(NCMe)(depe)2]BF4, 2g: The complex 2g (yellow solid,
yield 80%) was obtained by the same procedure as described in the
{1H} NMR(THF) 83.5 ppm (s). FAB MS calc. for C61H65FeNP4S: 1023;
obs. 1023(M+). trans-FeH(CN)(dtpe)2, 1m: yellow solid, yield 59%. IR:
preparation of 1n by using RuHCl(depe)2 as the starting material. H
1
2
2056 (
ν
CN). 1H NMR(acetone‑d6): ꢀ 15.1 ppm (quint, 2JHP = 44.2 Hz,
NMR(acetone‑d6): ꢀ 18.2 ppm (quint, JHP = 19.7 Hz, RuH); 31P{1H}
FeH); 31P{1H} NMR(THF) 89.1 ppm (s). FAB MS calc. for C61H65FeNP4:
NMR(THF) 62.2 ppm (s). FAB MS calcd for C22H52NP4Ru: 556; obs.: 515
(M+ꢀ MeCN).
991; obs. 991 (M+).
trans-[FeH(NCMe)(dtpe)2]BPh4 , 1n: trans-Fe(H)(Cl)(dtpe)2 (0.10 g,
0.10 mmol) was stirred for 5 h in 10 mL acetonitrile with NaBPh4 (0.05
g, 0.44 mmol). The solution was filtered through THF saturated Celite.
The solvent was removed under vacuum. Methanol (10 mL) was added
and the resulting yellow solid was filtered off and washed with meth-
anol. 1n (0.16 g, yield 86%) was obtained. 1H NMR(acetone‑d6): ꢀ 20.8
ppm (quint, 2JHP = 48.1 Hz, FeH); 31P{1H} NMR(THF) 80 ppm (s). FAB
MS calc. for C62H68FeNP4: 1006; obs. 965 (M+- MeCN).
trans-[Ru(H)(MeIm)(meso-tet-1)]BF4 , 2j: The complex 2j (cream
colored solid, 0.035 g, 60%) was made by the same procedure as
described in the preparation of 1n by using trans-RuHCl(meso-tet-1) as
the starting material. 1H NMR(acetone‑d6): 8.0–6.5 (m, PC6H5); 6.6, 6.1,
5.95 (3 m, –CHCHNCH-); 3.0–1.1 (6 m, PCH2CH2P); 3.05 (s, –NCH3);
ꢀ 19.5 (quint, 2JHP = 20.1 Hz, RuH). 31P{1H} NMR(THF) 98.7 (dm), 59.1
(dm), 2JPP = 230.1 Hz. FAB MS calcd for C46H49N2P4Ru: 855; obs.: 772
(M+ꢀ H,MeIm).
trans-[FeH(CO)(dtpe)2]BF4 , 1p: trans-[FeH(Cl)(dtpe)2] (0.15 g, 0.15
mmol) was dissolved in 20 mL acetone under a CO atmosphere to give a
dark red solution. NaBF4 (0.016 g, 0.15 mmol) was then added and the
mixture was stirred overnight. This pale yellow solution was filtered
through acetone-saturated Celite. The solvent was evaporated to 3 mL
and the addition of 15 mL diethyl ether caused the precipitation of a pale
yellow solid which was filtered off and washed with diethyl ether. The
trans-RuH(X)(dtfpe)2, 2m, 2o: A mixture of [RuH(dtfpe)2]BF4 (0.2 g)
and an excess amount of MX (KNCS for 2o, NaStBu for 2 m) in MeOH
(10 mL) was stirred at room temperature for 5 min to give a colorless
solution. The MeOH was removed completely under vacuum. The res-
idue was extracted with CH2Cl2 (10 mL). Addition of hexane (20 mL)
produced a white solid. trans-RuH(StBu)(dtfpe)2, 2m: white solid, yield
80%. 1H NMR(C6D6): ꢀ 18.9 ppm (quint, 2JHP = 19.1 Hz, RuH); 31P{1H}
NMR(THF) 62.2 ppm (s). FAB MS calcd for C64H50F24RuP4S: 1532; obs.:
1532 (M+). trans-RuH(NCS)(dtfpe)2, 2o: 1H NMR(CD2Cl2): ꢀ 17.1 ppm
(quint, 2JHP = 20.7 Hz, RuH); 31P{1H} NMR(THF) 65.6 ppm (s). FAB MS
calcd for C61H41F24RuNP4S: 1501; obs.: 1501 (M+).
yield was 0.15 g (93%). IR: 1940 (
ν
CO). 1H NMR(acetone‑d6) ꢀ 8.05 ppm
(quint, 2JHP = 47.6 Hz, FeH); 31P{1H} NMR(THF) 82.9 ppm (s). FAB MS
calc. for C61H64FeOP4: 993; obs. 993 (M+).
trans-FeH(X)(dtfpe)2, 1r, 1 t, 1u: The complexes 1r, 1t and 1u were
obtained by the same procedure as described in the preparation of 1a by
using FeHCl(dtfpe)2 as the starting material. trans-FeH(NCO)(dtfpe)2,
trans-[RuH(NCMe)(dtfpe)2]BF4 , 2p: The complex 2p (white solid,
yield 80%) was obtained from same procedure as described in prepa-
ration of 1n. 1H NMR(acetone‑d6): ꢀ 15.9 ppm (quint, 2JHP = 18.8 Hz,
RuH); 31P{1H} NMR(THF) 62.6 ppm (s). FAB MS calcd for
1
1r: orange solid, yield: 53%. H NMR(acetone‑d6): ꢀ 26.2 ppm (quint,
2JHP = 44.2 Hz, FeH); 31P{1H} NMR(THF) 81.7 ppm (s). FAB MS calc. for
C
61H41F24FeNOP4: 1439; obs. 1439 (M+). trans-FeH(NCS)(dtfpe)2, 1t:
C
62H44F24RuNP4: 1471; obs.: 1471 (M+)
orange solid, yield: 54%. 1H NMR(C6D6): ꢀ 24.9 ppm (quint, 2JHP = 42.8
trans-[RuH(CO)(dtfpe)2], 2q: The complex 2q (white solid, yield
Hz, FeH); 31P{1H} NMR(THF) 85.6 ppm (s). FAB MS calc. for
80%) was obtained from same procedure as described in preparation of
1h by using [RuH(dtfpe)2]BF4 as the starting material. 1H NMR(CD2Cl2):
ꢀ 7.1 ppm (quint, 2JHP = 19.7 Hz, RuH); 31P{1H} NMR(THF) 63.2 ppm
(s). FAB MS calcd for C61H41F24RuOP4: 1484; obs.: 1484 (M+)
trans-OsHCl(depe)2: The complex was prepared using a modified
version of Chatt and Hayter’s method [24] which is described in refer-
ence [28]. NMR data were not given when this complex was first
described. 1H NMR(C6D6) –22.8 (quint, 2JHP = 15.3 Hz); 31P{1H} NMR
(acetone) 33.4 (s).
C
61H41F24FeNP4S: 1455; obs. 1455 (M+). trans-FeH(CN)(dtfpe)2, 1u:
1
yellow solid, yield: 56%. IR: 2083 (
ν
CN). H NMR(C6D6): ꢀ 15.3 ppm
(quint, 2JHP = 46 Hz, FeH); 31P{1H} NMR(THF) 91.5 ppm (s). FAB MS
calc. for C61H41F24FeNP4: 1423; obs. 1423 (M+).
trans-[FeH(NCMe)(dtfpe)2]BF4 , 1v: The complex 1v (yellow solid,
yield 80%) was obtained by the same procedure as described in prepa-
1
ration of 1n by using FeHCl(dtfpe)2 as the starting material. H NMR
(acetone‑d6): ꢀ 21.3 ppm (quint, JHP = 43.5 Hz, FeH); 31P{1H} NMR
2
(THF) 84.1 ppm (s). FAB MS calc. for C62H44F24FeNP4: 1438; obs. 1397
(M+- MeCN).
trans-OsH(StBu)(dppe)2 , 3a: The complex was prepared from trans-
OsHBr(dppe)2 (0.122 g, 0.144 mmol) and NaStBu (0.07 g, 0.62 mmol) in
4

S.D. Drouin et al.
Inorganica Chimica Acta xxx (xxxx) xxx
THF (20 mL). See below for the preparation of the OsHBr(dppe)2 com-
plex. The reaction mixture was refluxed for 2 h. After completion of the
reaction, the reaction mixture was slowly filtered through THF-
saturated Celite. Solvents were removed from the bright yellow filtrate
by vacuum distillation. Acetone (~4 mL) was added to the orange-
Table 4
List of 31P NMR chemical shifts of some complexes trans-[MH(L)(PP)2]+ and
trans-MH(X)(PP)2 (M = Fe, Ru and Os).
P
P
L or Xꢀ
δ
δ
δ
Δ (Fe-
Δ (Ru-
Δ (Fe-
–
(PFe)
(PRu)
(POs)
Ru)
Os)
Os)
dppe a
dppe a,c
dppe
H2
92.5
68.6
83.4
62.3
–
37.5
49.8
28
23.9
31.1
33.6
34.3
–
55
–
yellow residue, and the tangerine orange solid was filtered and dried
1
in vacuum. Yield was 33% (0.40 g). IR: 2040 (br w,
ν
OsH). H NMR
Hꢀ
–
–
–
–
SPhꢀ
–
–
(C6D6): 7.6 to 6.8 (m, 40H, Ph), 2.89 (m, 4H, CH2), 2.06 (m, 4H, CH2),
1.08 (s, 9H, tBu), ꢀ 17.8 (quint, 1H, 2JHP = 16.7 Hz, OsH). 31P{1H} NMR
(THF) 24.4 ppm (s). FAB MS calcd for C56H58SP4Os: 1078; obs. 1079
(weak, MH+), 990 (MH+ -tBuS), 988 (MH+ -tBuS, 2H).
dppe
NCMe
84.3 a
33
51.3
b
dppe
Clꢀ
–
62.5 d
68.7
84.4
64.4
61.7
64.7
62.2
61.8
71.4
84.8
65.6
62.6
63.2
30.5
30.7
48.3
32.0
–
–
32.0
38.0
36.1
32.4
–
–
depea
depe a,c
depe
H2
92.9
–
24.2
62.2
Hꢀ
–
trans-OsH(SPh)(dppe)2 , 3b: A slurry of trans-OsHBr(dppe)2 (0.218 g,
0.204 mmol), NaSPh (0.048 g, 0.36 mmol) and THF (10 mL) was stirred
for 3 h, and the 31P NMR spectrum recorded; trans-OsHBr(dppe)2 was
the major species present. More NaSPh was added (0.050 g, 0.038
mmol) and the reaction mixture was refluxed for 3 h and then stirred for
50 h. The 31P NMR spectrum showed that the reaction was complete.
The suspension was chilled, filtered through THF-saturated Celite, and
the solvent was removed from the yellow filtrate under reduced pres-
sure. The resulting bright yellow powder was precipitated from THF (3
mL) by addition of cold Et2O (10 mL) to give 0.277 g of crude product.
This was dissolved in benzene (20 mL) and the salts were removed by
filtering through benzene-saturated Celite. The volume of the filtrate
reduced in vacuum to 2 mL, Et2O (4 mL) were added, and the mixture
briefly cooled. The bright yellow solid (0.181 g, 80% yield) was filtered,
Clꢀ
87.2
85.2
89.1
89.1
89.1
94.7
–
85.6
84.1
90.7
22.8
23.5
24.4
26.9
27.3
23.3
–
20.0
21.5
27.5
24.1
55.2
depe
Brꢀ
–
depe
NCSꢀ
MeCN
StBuꢀ
H2
–
–
–
depe
–
–
–
–
31.5
32.6
–
–
–
33.5
–
depe
–
dtfpe a
dtfpe a,c
dtfpe
39.9
52.2
–
54.8
Hꢀ
–
NCSꢀ
MeCN
CO
–
dtfpe
–
–
–
dtfpe
–
Average
56.5
a Ref. [23]. bNo NMR data were given by Chatt et al. [15] when this complex was
first described. c Data for trans isomer is given; d Ref. [33].
obs. 1071 (weak, M+), 989 (M+ꢀ MeIm). trans-[OsH(PMe3)(dppe)2]PF6,
washed with Et2O (5 mL) and dried in vacuum. IR: 2020 (m,
ν
OsH). 1H
3g: yellow solid, 0.07 g, yield 66%. IR: 2038 (w,
ν
OsH). 1H NMR(ace-
NMR(C6D66): 7.5 to 6.75 (m, 40H, PC6H5), 6.63 (m, 2H, SC6H5), 6.26 (m,
3H, SC6H5), 2.80 (m, 4H CH2), 2.08 (m, 4H, CH2), 16.6 (quint, 1H, 2JHP
= 16.6 Hz, OsH). 31P{1H} NMR(C6D6) 28.0 ppm (s). FAB MS calcd for
tone‑d6) 7.4 to 6.9 (m, 40H, Ph), 2.64 (m, 4H, CH2), 2.30 (m, 4H, CH2),
0.49 (d, 9H, 2JHP = 7 Hz, Me), ꢀ 11.9 (d quint, 1H, 2JHP = 41.7, 19.2 Hz,
OsH). 31P{1H} NMR(acetone‑d6): 29.4 (d, dppe), 73.9 (quint, PMe3, 2JPP
= 14.2 Hz). FAB MS calcd for C55H58P5Os: 1065.3; obs. 1065.2 (M+),
989 (M+ꢀ PMe3) and 987 (M+ -PMe3,2H). trans-[OsH(NCMe)(dppe)2]
C
58H54SP4Os: 1098; obs. 1098 (M+), 989 (M+ -SPh).
trans-OsHCl(dppe)2 3d: This complex was prepared by reacting trans
[Os(
η
2-H2)H(dppe)2]PF6 with LiCl as described in reference [28]. NMR
PF6, 3h: white solid, 0.045 g, yield 70%. IR: 2066 (w, νOsH), 2267 (w,
data were not given when this complex was first described [24]. 1H NMR
(C6D6) 7.6 to 6.8 (m, 40H, Ph), 2.57 (m, 4H, CH2), 2.00 (m, 4H, CH2),
ꢀ 20.4 (quint, 1H, 2JHP = 15.4 Hz, OsH); 31P{1H} NMR(C6D6) 30.5 (s).
trans-OsHBr(dppe)2 3e: This complex was prepared from
(NH4)2[OsBr6] and 2.2 equivalents of dppe as described in reference
[28]. 31P{1H} NMR(CH2Cl2) 28.7 (s). 1H NMR(CD2Cl2) 7.35 to 6.95 (m,
ν
CN). 1H NMR(CD2Cl2): 7.4 to 7.1 (m) and 6.7 (br s) (40H, Ph), 2.54 (m,
4H, CH2), 2.06 (m, 4H, CH2), 1.81 (s, 3H, NCMe), 16.6 (quint, 1H, 2JHP
= 16.2 Hz, OsH). 31P{1H} NMR(CH2Cl2) 33.0 ppm (s). FAB MS calcd for
C54H52NP4Os: 1030; obs. 1030 (weak, M+), 989 (M+ -NCMe), 987 (M+
-NCMe, 2H). trans-[OsH(P(OMe)3)(dppe)2]PF6, 3i: white solid, 0.08 g,
yield 75%. IR: 2002 (w,
ν
OsH). 1H NMR(CD2Cl2): 7.45 to 7.0 (m) and 6.8
40H, Ph), 2.66 (m, 4H, CH2), 2.10 (m, 4H, CH2), ꢀ 20.4 (quint, 2JHP
=
(br s) (40H, Ph), 2.76 (m, 4H, CH2), 2.38 (m, 4H, CH2), 2.86 (d, 9H, 2JHP
= 10.4 Hz, Me), ꢀ 10.6 (d quint, 1H, 2JHP = 86.7, 19.0 Hz, OsH). 31P{1H}
NMR(acetone‑d6) 28.1 (d, dppe), 91.7 (quint, P(OMe)3, 2JPP = 23.0 Hz).
FAB MS calcd for C55H58O3P5Os: 1113; obs. 1113 (M+), 987 (M+ -P
(OMe)3).
15.3 Hz, OsH)
trans-[Os(
η
2-H2)H(dppe)2]PF6: The complex was prepared in 92%
yield from OsH2(dppe)2 (0.660 g in 40 mL Et2O) and 0.35 mL of 60–65
wt% HPF6(aq), using a method analogous to that described for the BF4ꢀ
salt [26]. After 15 min of reaction the solvent was decanted from the
white product, and the solid was treated with 2x40 mL of Et2O (stirring
followed by decantation). The product was filtered, washed with 3x15
mL of Et2O and dried in vacuo; 0.696 g were obtained. The complex is
air-stable enough to be filtered through the air, though long term storage
trans-[OsH(P(OPh)3)(dppe)2]PF6, 3j: white solid, 0.08 g, yield 77%.
1
IR 2062, 2013 (w). H NMR(acetone‑d6): 7.4 to 6.9 (m, 40H, PC6H5),
6.84 and 6.23 (m, 15H, OPh), 3.02 (m, 4H, CH2), 2.64 (m, 4H, CH2),
ꢀ 10.4 (d quint, 1H, JHP = 108.9, 19.8 Hz, OsH). 31P{1H} NMR
2
(acetone/THF: 3/2 v/v) 25.3 (d, dppe), 76.5 (quint, 2JPP = 22.1 Hz, P
(OPh)3). Anal. Calcd for C70H64F6P6Os: C, 58.25; H, 4.47. Found: C,
57.81; H, 4.66. FAB MS calcd for C70H64P5Os: 1299; obs. 1299 (M+), 989
(M+ -P(OPh)3) and 987 (M+ -P(OPh)3, 2H).
1
(weeks) requires a N2 or Ar atmosphere. H and 31P NMR(acetone or
dichloromethane solvents) as described for trans-[Os(
η
2-H2)H(dppe)2]
BF4; a PFꢀ6 resonance is also observed by 31P NMR.
trans-[OsH(L)(dppe)2]PF6,
3f-3j:
trans-[Os(H)(H2)(dppe)2]PF6
(0.106 g, 0.093 mmol) was dissolved in 10 mL of acetone and a slight
excess amount of L was added. The solution was refluxed for 50 min; a
31P NMR spectrum indicated that the reaction was complete. The solvent
was removed under reduced pressure, and the white residue dissolved in
3 mL of CH2Cl2. Three volumes of Et2O were added, and the solution was
cooled for 30 min, followed by stirring to produced a white solid. The
volume was reduced in vacuum by one third and the solid was filtered,
washed with 2 mL of Et2O and dried in vacuum.
3. Results and discussion
3.1. Synthesis and characterization
The complexes trans-MH(X)(PP)2 (M = Fe, PP = depe, dtpe and
dtfpe, X = NCSꢀ , NCOꢀ , Nꢀ3 , CNꢀ , Brꢀ ; M = Ru, PP = depe, 1/2 meso-tet-
1, X = Clꢀ , Brꢀ , NCSꢀ , NCOꢀ , StBuꢀ ; M = Os, PP = dppe, X = StBuꢀ ,
SPhꢀ ) were prepared in high yield from the starting complexes trans-MH
(X)(PP)2 (X = Clꢀ or Brꢀ ) and an excess of the appropriate potassium or
sodium salt in acetone at 20 ◦C in 4 h as shown in Eq. (5). The 31P NMR
data for some of these complexes and ones described below are listed
Table 4 to illustrate the trend δ Fe > δ Ru > δ Os.
trans-[OsH(MeIm)(dppe)2]PF6, 3f: yellow solid, 0.089 g, yield 75%.
IR: 2066 (w,
ν
OsH). 1H NMR(acetone‑d6) 7.7 to 6.7 (m, 40H, Ph), 6.46,
6.21, 6.08 (s, –CHCHNCH-), 3.07 (s, 3H, NMe), 2.79 (m, 4H, CH2), ~2.0
(m, ~4H, CH2), ꢀ 18.4 (quint, 1H, 2JHP = 16.6 Hz, OsH). 31P{1H} NMR
(acetone) 34.8 ppm (s). Anal. Calcd for C56H55F6N2OsP5: C, 55.35; H,
4.56. Found: C, 55.69; H, 4.87. FAB MS calcd for C56H55N2OsP4: 1071;
5

S.D. Drouin et al.
Inorganica Chimica Acta xxx (xxxx) xxx
(5)
Field and co-workers have reported the preparation of structurally
similar thiolate iron hydride complexes MH(SR)(dmpe)2 [29,30]. The
dihydride complexes M(H)2(dmpe)2 (M = Fe, Ru) are protonated by
arene or alkanethiols in THF to initially yield the molecular hydrogen
complexes [MH(H2)(dmpe)2]+ which react with the thiolate to give a
mixture of cis and trans-MH(SR)(dmpe)2. These thiolate complexes react
with HBF4•Et2O to form protonated thiol complexes, rather than dihy-
drogen complexes [31]. The thiolate complex trans-RuH(SPh)(dppe)2
was reported to crystallize with the thiolate trans to the hydride [31].
Some literature data for thiolate complexes are included in the Sup-
porting Information (SI).
The complexes trans-[MH(L)(PP)2]BF4 (M = Fe, PP = depe, dtpe,
dtfpe, L = NCMe, CO, N2, H2; M = Ru, PP = depe, 1/2 meso-tet-1, L =
NCMe, CO, H2, PMe2Ph, MeIm) were also prepared from trans-MHCl
(PP)2 by direct ligand substitution in the presence of NaBF4 as Eq. (6)
shows. Analogous iron nitrile complexes have been prepared in a similar
fashion; these complexes tend to crystallize in a trans configuration
[32]. The electrochemistry of these complexes has been reported by Lee
et al. [32] and is included in the SI.
Fig. 1. Cyclic voltammograms of trans-OsH(SPh)(dppe)2 in THF solution con-
taining 0.2 M NBu4PF6, at scan rates (n) of 50 and 500 mVsꢀ 1
.
(
η
2-H2)(PP)2]PF6 (M = Fe, Ru, Os) are prepared in a pure form in high
yield. The pathway for the substitution reaction is likely via a five co-
ordinate hydride species. Once the electron-deficient five-coordinate
metal hydride is formed, the substitution goes quickly. Preparation of
the carbonyl derivative of trans-[FeH(H2)(dtfpe)2]BF4, however,
required generation of the five coordinate intermediate in the solid state,
as Eq. (9) shows.
(6)
The five-coordinate ruthenium hydride complex [RuH(dtfpe)2]BF4
of unknown stereochemistry can also be used to prepare trans-RuH(X)
(dtfpe)2 (X = StBuꢀ and NCSꢀ ) and trans- [RuH(NCMe)(dtfpe)2]+ in high
yield as Eq. (7) shows.
(9)
The FAB MS results show that for anionic ligands, complexes trans-
MH(X)(PP)2 can either lose Hꢀ or Xꢀ to give five-coordinate fragments
[MH(PP)2]+ or [MX(PP)2] whereas for the neutral ligands, the com-
plexes trans-[MH(L)(PP)2]+ lose the neutral ligand to give fragments
[MH(PP)2]+. The observation of the unsaturated complexes support the
proposal that they are intermediates in the substitution reactions rep-
resented by Eqs. (8) and (9).
(7)
The complexes trans-[MH(L)(PP)2]+ and trans-MH(X)(PP)2 prepared
in this work were shown by 1H and 31P NMR spectroscopy to exist solely
as the trans isomer. In preparing the cationic members of the series trans-
[MH(L)(dppe)2]+, tetraphenylborate (BPhꢀ4 ) was avoided as a counter-
ion because it is redox active in the electrochemical potential region
of interest. Tetrafluoroborate (BFꢀ4 ) or hexafluorophosphate (PF6ꢀ ) are
appropriate counter-ions.
However, it is noteworthy that if the ligand Xꢀ is basic enough, the
deprotonation of the dihydrogen complex may occur. For example, the
reaction of trans-[OsH(H2)(dppe)2]PF6 and excess NaStBu gives a
mixture of cis and trans-Os(H)2(dppe)2 (94%) and trans-OsH(StBu)
(dppe)2 (6%). That the pKa of the dihydrogen complex is about 13 while
the pKa of HStBu is about 10 in water, but higher in less polar solvents
gives a reasonable explanation [23].
Some complexes trans-[MH(L)(PP)2]+ and trans-MH(X)(PP)2 (M =
Fe, PP = dtfpe, L = CO; M = Ru, PP = dppe, X = SPhꢀ ; M = Os, PP =
dppe, X = Clꢀ , StBuꢀ , L = MeIm, NCMe, PMe3, P(OMe)3, P(OPh)3) can
be prepared from their dihydrogen precursors in acetone as shown in Eq.
(8). Similar iron-group dihydrogen complexes have been synthesized,
and it has been shown that the dihydrogen can be substituted with tri-
flate [34]. The complex trans-OsH(Br)(dppe)2, however, is prepared
directly from (NH4)2[OsBr6] and dppe in refluxing MeOH/EtOH [28].
Dichloromethane was not used to prepare solutions of the electron
rich thiolate complexes trans-MH(SR)(dppe)2 because this solvent reacts
with these compounds. Most of these metal hydride complexes are very
soluble in most organic solvents.
All of the products were characterized by 31P, 1H NMR, FAB MS and
electrochemistry. The 31P NMR data (Table 4) show systematic trends
where for complexes with the same ligands but different metals, the
ruthenium chemical shifts are 20 to 30 ppm downfield of the those of
iron while the osmium chemical shifts are 51 to 62 ppm downfield of he
ruthenium ones. There is a good correlation between the 31P chemical
shifts of ruthenium and osmium (Figure S4) but a poor correlation be-
tween those of iron and ruthenium or iron and osmium.
(8)
3.2. Electrochemistry of trans-[MH(L)(PP)2]+ and trans-MH(X)(PP)2
Since we are interested in the oxidation of the M(II) state, the region
between ꢀ 1 to 1.6 V vs. Fc+/Fc was examined. The cyclic voltammo-
grams of most of the metal hydride complexes showed pseudo-reversible
The advantage of this route is that the precursor materials, trans-[MH
6

S.D. Drouin et al.
Inorganica Chimica Acta xxx (xxxx) xxx
relatively high, which makes the oxidized complex stable to proton loss.
For example , the FeIII complex [FeH(Cl)(depe)2]+ has a pKLaAC of 19,
which is too high to be acidic enough for proton loss, and the voltam-
mogram is reversible. Conversely, complexes that are irreversibly
oxidized tend to have more acidic metal hydrides upon oxidation. For
example, the FeIII complex [FeH(CO)(dtpe)2]2+ has a pKaLAC value of ꢀ 1,
and the complex is irreversibly oxidized. The oxidation is irreversible
likely because the complex loses a proton upon oxidation. On the same
grounds, some reversible voltammograms are expected to be irrevers-
ible, yet display partially reversible behavior, possibly due to slow
proton loss.
The dihydrogen complexes in Table 3 (entries 1o and 1w) also
display irreversible oxidations. In these cases, the complex likely loses
dihydrogen upon oxidation, as discussed elsewhere [35].
Fig. 2. Plot of anodic and cathodic peak currents (ipa and ipc) versus scan rate
ν1/2 for a solution of OsH(SPh)(dppe)2.
The EL values for the depe, dtpe, and meso-tet-1 ligands were not
reported by Lever, and are determined here from experimental data. The
EL values of all of these ligands are included in Table 1.
waves in this region. A typical cyclic voltammogram is shown in Fig. 1.
The cyclic voltammograms (CVs) of the complex trans-OsH(SPh)
(dppe)2 were obtained by cycling the potential of the working electrode,
relative to Fc+/Fc, between ꢀ 0.8 and 0.7 V at a scan rate of 50 or 500
mVsꢀ 1. An oxidation wave at the anionic potential peak Epa is observed
in the forward scan and a reduction wave at the cathodic potential peak
It is generally agreed that a less donating ligand gives rise to a more
electron poor metal center M which, in turn, is more difficult to oxidize.
For example, the complexes trans-[FeH(L)(dtfpe)2]+, with an electron-
withdrawing CF3 substituents on the aryl groups, have more positive
electrochemical potentials than the analogous dtpe complexes with CH3
substituents or depe complexes with Et groups on phosphorus (see
table 2). For a given metal and PP ligand, the E1/2 of the complexes trans-
[MH(L)(PP)2]+ and trans-MH(X)(PP)2 are generally shifted to more
positive potentials as the EL value of the Xꢀ or L ligand increases; ex-
ceptions to this ordering can be accounted for by the experimental error
E
pc is displayed in the reverse scan. The current height at Epa or Epc is
proportional to the square root of the scan rate,
ν
1/2, and the ipa/ipc ratio
is one (see Fig. 2). Both of these characteristics are expected for
reversible redox processes. However, the separation between Epa and Epc
(ΔEp) increases as the scan rate increases (see Table 5), and is not equal
to 59 mV/n (n = electron stoichiometry) as expected for a reversible
process. This is attributed to an increase in double layer capacitance
with increasing scan rate, and the uncompensated high resistance of the
solvent mixture.
It is interesting to note that at more positive potentials (potential is
increased from ꢀ 0.1 to 0.4 V vs. Fc+/Fc) a second set of Epa and Epc
peaks is observed for trans-OsH(SPh)(dppe)2 due to the d4/d5 couple
(see Table 5). The E1/2 values for the complexes trans-[MH(L)(PP)2]+
and trans-MH(X)(PP)2 are obtained as the potential half-way between
the oxidation (Epa) and the reduction (Epc) potentials.
The redox behavior of several complexes, such as trans-[FeH(L)
(dtpe)2]+ (L = H2, CO), trans-FeH(X)(dtfpe)2, trans-[FeH(L)(dtfpe)2]+,
trans-[Ru(H)(NCMe)(dtfpe)2]+ and trans-[OsH(P(OMe)3)(dppe)2]+, are
irreversible. Anodic peak potentials (Epa) have been reported in Table 3
for these complexes. A typical cyclic voltammogram of an irreversible
redox process for the iron group hydride complexes is shown in Fig. 3.
This irreversibility can generally be rationalized by investigating the
acidity of the oxidized complex, as discussed elsewhere [35], with use of
the ligand acidity constant (LAC) method. For reversible voltammo-
grams, generally the calculated pKLaAC for the oxidized complex is
Fig. 3. Cyclic voltammogram of complex trans-[OsH(P(OMe)3)(dppe)2]+ with
50 mV/s scan rate in CH2Cl2 containing 0.2 M NBu4PF6 solution.
Table 5
Measured E1/2 values (V, vs. Fc+/Fc) of the d5/d6 couple (unless noted otherwise) for the d6 complexes trans-[OsH(L)(dppe)2]+ in 0.2 M NBu4PF6 solution at the noted
scan rates.
Ligand
E1/2 obs at scan rate,
= 50
ν
(mVsꢀ 1
)
solvent
ν
ΔE, mV
ν
= 250
ΔE, mV
ν
= 500
ΔE, mV
StBuꢀ
SPhꢀ
ꢀ 0.769
ꢀ 0.646
0.262b
ꢀ 0.131
ꢀ 0.104
0.311
38 a
143
137
110
123
118
143
84
ꢀ 0.771
ꢀ 0.649
0.277b
ꢀ 0.124
ꢀ 0.116
0.311
55 a
173
158
142
128
143
158
89
THF
ꢀ 0.664
0.245b
ꢀ 0.126
ꢀ 0.101
0.311
109
72
THF
SPhꢀ
THF
Clꢀ
82
THF
Brꢀ
99
THF
MeIm
PMe3
NCMe
P(OMe)3
P(OPh)3
99
THF
0.673
108
74
0.681
0.683
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
0.584
0.584
0.587
0.97c
0.97c
>1.2 d
a Small ΔEp indicates possible 2eꢀ process; also observed: waves at ꢀ 0.52 V, ꢀ 0.14 V and an irreversible wave with Epa of 0.23 ± 0.02 V. b d4/d5 couple. c Epa value.
d The triphenylphosphite derivative is more difficult to oxidize than the dichloromethane solvent employed, and no redox potential was obtained for this complex.
7

S.D. Drouin et al.
Inorganica Chimica Acta xxx (xxxx) xxx
trans-[FeH(H2)(dtpe)2]+ (pKa = 12.6 ± 0.4) [23]. The unexpected
electrochemical behavior of trans-[FeH(H2)(dtfpe)2]+ may be due to its
acidic nature (pKa = 7.8 ± 0.3) [23]. It is plausible that this acidic
dihydrogen complex, which is in low concentration relative to the
electrolyte NBu4PF6 (0.2 M), exists as a solvent-dependent equilibrium
between the protonated and the deprotonated forms:
B + [FeH(H2)(dtfpe)2]PF6⇌Fe(H)2(dtfpe)2 + HB+PF6ꢀ
(10)
The nature of the base, B, remains uncertain but its acid form, BH+,
must have a pKa near 8. It is likely that the electrolyte, NBu4PF6, is
responsible for the deprotonation. Tilset and Parker have observed that
electrolytes commonly used for electrochemical studies are sufficiently
basic to cause the deprotonation of some acidic hydride complexes
[36,37]. For example, the acidic hydride [CpW(CO)2(PMe3)H2]+ is quite
stable in dry acetonitrile but undergoes spontaneous deprotonation in
acetonitrile/0.1 M NBu4PF6 [36,37]. The unexpectedly low, solvent-
dependent Epa value for [FeH(H2)(dtfpe)2]BF4, therefore, reflects the
presence of the dihydride species. As further support for this argument,
when Fe(H)2(dtfpe)2 was combined with an equal weight of [FeH(H2)
(dtfpe)2]BF4, the anodic wave (for the dichloromethane solution)
appeared at +0.05 V instead of +0.48 for the dihydrogen complex alone.
The oxidation potential for the dihydrogen complex is sensitive to the
presence of the dihydride species in solution as expected on the basis of
the Nernst equation.
3.3. Electrochemistry of literature-reported iron-group hydrides
A more extensive collection of literature reported cyclic voltammo-
grams (CVs) of iron-group hydrides is included in the Supporting In-
formation. A part of the tables in the SI is copied from a previous paper
[35]. The data includes both reversible and irreversible oxidations of M
(II) hydride complexes, where M = Fe, Ru, and Os. Where multiple
papers report the same data, all data is included. Some EL values are
determined here, while others are found in Lever’s paper [2]. All EL
values used are listed in Table S5.
The literature iron voltammetry data is presented in Table S1. Fig. 4a
shows the FeIII/II Lever plot for the combined data measured in this work
(Table 3) and from the literature, 43 data points in all. In order to
generate this plot the EL value of the hydride was varied within the sum
of the EL values of ligands until the optimum fit to the data was obtained.
With EL(H_) ꢀ 0.65 ± 0.02 V, the correlation was good at 0.91 The
equation of the line is the same, within error, as Lever’s equation for iron
complexes as shown in Table 2. This EL value is somewhat near to the
value calculated using the computed electronic parameter method,
which is ꢀ 0.76 [8].
The plot for just the 17 iron complexes measured in the current work
(Table 3) is shown in Figure S1. The correlation is improved to 0.94, and
the slope and intercept are also close to Lever’s equation for iron data
shown in Table 2. The same EL value of ꢀ 0.65 V is obtained for hydride
from this subset of data. Note that the two dihydrogen complexes were
omitted from this analysis.
Fig. 4. Plots of Eobs vs. ΣEL for the group 8 metal hydride complexes. The solid
blue line corresponds to the line of best fit for the data. The dashed orange line
is from Lever’s equations [2]. (a) FeIII/II couple, Eobs = 1.08ΣEL ꢀ 0.46. (b)
RuIII/II couple, Eobs = 0.77ΣEL + 0.15 and (c) OsIII/II couple, Eobs = 1.04ΣEL
–
0.44. (For interpretation of the references to color in this figure legend, the
reader is referred to the web version of this article.)
The literature ruthenium data is presented in Table S2; Fig. 4b shows
the RuIII/II Lever plot for the combined data (49 data points) from the
literature and the current work (Table 3). The correlation is modest at
0.84. The EL value for the hydride is found to be ꢀ 0.65 V which is the
same as the iron data. The equation of the line deviates somewhat from
Lever’s equation for ruthenium complexes as shown in Table 2. This
deviation is attributed to the relatively poor correlation. Note that the
correlation does not improve if irreversible CVs are excluded.
The plot for just the 15 ruthenium complexes presented in Table 3 is
shown in Figure S2. The correlation is improved to 0.87, but the slope
and intercept are still somewhat different from Lever’s equation for
ruthenium data shown in Table 2. The same EL value of ꢀ 0.65 V is ob-
tained for the hydride ligand from this subset of data.
of the E1/2 measurements. As the ligands change from
σ
and
π
-donor to
σ
σ
-donor and
-donor and
π
π
-acceptor as the value of the EL parameter increases, the
-acceptor ligands (such as N2, CO and H2) make the metal
complex more oxidizing.
The solvent influence on Eobs is probably small. For example, the
complex trans-[FeH(L)(dtpe)2]+, where L = H2, NCMe, showed similar
E
1/2 values in THF and CH2Cl2. Thus we assume that the potentials that
we are measuring in this study are independent of the solvent system.
One notable exception, however, is the dihydrogen complex trans-[FeH
(H2)(dtfpe)2]+ which exhibits a solvent dependent Epa (see Table 3,
entry 1w): Epa = ꢀ 0.03 V in THF; Epa = +0.48 V in CH2Cl2) the values of
which are surprisingly lower than observed for the more basic complex
The literature osmium data is presented in Table S3; Fig. 4c shows
8

S.D. Drouin et al.
Inorganica Chimica Acta xxx (xxxx) xxx
the OsIII/II Lever plot for the combined data (19 data points) from the
literature and this work (Table 3). The correlation is modest at 0.88, and
the EL value for the hydride is found to be ꢀ 0.37 ± 0.05 V. This EL for the
hydride is more positive for Os than for Ru and Fe. This variation is
[4] R.H. Morris, K.A. Earl, R.L. Luck, N.J. Lazarowych, A. Sella, Inorg. Chem. 26 (1987)
2674–2683, https://doi.org/10.1021/ic00263a024.
[5] C.M. Duff, G.A. Heath, Inorg. Chem. 30 (1991) 2528–2535, https://doi.org/
10.1021/ic00011a014.
[6] A.B.P. Lever, Inorg. Chem. 30 (1991) 1980–1985, https://doi.org/10.1021/
ic00009a008.
attributed to stronger M-H
σ bonding due to relativistic effects associ-
[7] H. Masui, A.B.P. Lever, Inorg. Chem. 32 (1993) 2199–2201, https://doi.org/
10.1021/ic00062a052.
ated with osmium. For the Lever plot shown, the equation of the line is
within error of Lever’s equation for osmium complexes shown in
Table 2.
[8] L. Perrin, E. Clot, O. Eisenstein, J. Loch, R.H. Crabtree, Inorg. Chem. 40 (2001)
5806–5811, https://doi.org/10.1021/ic0105258.
[9] A.J.L. Pombeiro, Eur. J. Inorg. Chem. 2007 (2007) 1473–1482, https://doi.org/
10.1002/ejic.200601095.
The plot for just the 8 osmium complexes of Table 2 is shown in
Figure S3. The correlation is improved to 0.93, but the slope and
intercept deviate from Lever’s equation for osmium complexes shown in
Table 2. The same EL value of ꢀ 0.37 V is obtained for the hydride ligand
from this subset of data.
[10] A.B.P. Lever, in: J.A. McCleverty, T.J. Meyer (Eds.) Comprehensive Coordination
Chemistry II, Pergamon, Oxford, 2003, pp. 251–268.
[11] M.R. DuBois, D.L. DuBois, in Catalysis without Precious Metals, Wiley-VCH Verlag
GmbH & Co, KGaA (2010) 165–180.
[12] S. Gonell, A.J.M. Miller, Advances in Organometallic Chemistry vol. 70 (2018)
1–69.
Overall, the literature data agrees with the data reported here,
although there is more scatter in the literature data. This scatter may be
due to the use of different referencing systems, different solvents, errors
in E1/2 values due to irreversibility, and slightly erroneous EL values.
[13] P. Zanello, Comments Inorg. Chem. 11 (1991) 339–359, https://doi.org/10.1080/
02603599108035831.
[14] C.J. Pickett, D.J. Pletcher, J. Organomet. Chem. 102 (1975) 327–333, https://doi.
org/10.1016/S0022-328X(00)95199-4.
[15] J. Chatt, C.T. Kan, G.J. Leigh, C.J. Pickett, D.R. Stanley, J. Chem. Soc. Dalton Trans.
(1980) 2032–2038, https://doi.org/10.1039/DT9800002032.
4. Conclusions
¨
[16] C.M. Habeck, C. Hoberg, G. Peters, C. Nather, F. Tuczek, Organometallics 23
(2004) 3252–3258, https://doi.org/10.1021/om030579k.
[17] W. Hussain, G.J. Leigh, H.M. Ali, C.J. Pickett, D.A. Rankin, J. Chem. Soc., Dalton
Trans. (1984) 1703–1708, https://doi.org/10.1039/DT9840001703.
[18] S.S.P.R. Almeida, M.F.C. Guedes Da Silva, A.J.L. Pombeiro, Collect. Czech. Chem.
Commun. 68 (2003) 1663–1676, https://doi.org/10.1135/cccc2003166310.1135/
cccc20031663.
The MIII/II electrochemistry of a series of group 8 metal hydride
phosphine complexes was presented. An EL value for the hydride ligand
was calculated based on the iron group phosphine complexes MHX(PP)2
and MHL(PP)+2 , and a collection of literature data . Based on the
ruthenium and iron data, an EL value of ꢀ 0.65 V was obtained. The
osmium data had an EL value of only ꢀ 0.37 V; this variation between
iron/ruthenium and osmium is attributed to stronger M-H bonding in
the 5d metal compared to the 3d and 4d metals. Ongoing work is focused
on obtaining an EL value for the hydride ligand for other metals.
[19] J. Chatt, W. Hussain, G.J. Leigh, H.M. Ali, C.J. Picket, D.A. Rankin, J. Chem. Soc.,
Dalton Trans. (1985) 1131–1136, https://doi.org/10.1039/DT9850001131.
[20] M.T. Bautista, K.A. Earl, P.A. Maltby, R.H. Morris, C.T. Schweitzer, Can. J. Chem.
72 (1994) 547–560, https://doi.org/10.1139/v94-078.
[21] F.P. Dwyer, J.W. Hogarth, Inorganic Syntheses, McGraw-Hill Book Co., Inc., New
York, N.Y., 1957.
[22] M.V. Baker, L.D. Field, J. Organomet. Chem. 354 (1988) 351–356, https://doi.org/
10.1016/0022-328X(88)80660-0.
[23] E.P. Cappellani, S.D. Drouin, G. Jia, P.A. Maltby, R.H. Morris, C.T. Schweitzer,
J. Am. Chem. Soc. 116 (1994) 3375–3388, https://doi.org/10.1021/ja00087a024.
[24] J. Chatt, R.G. Hayter, J. Chem. Soc. (1961) 2605–2611, https://doi.org/10.1039/
JR9610002605.
Credit authorship contribution statement
All of the authors contributed equally to this research.
Declaration of Competing Interest
[25] P. Pertici, W. Porzio, G. Vitulli, M. Zocchi, Inorg. Chim. Acta Lett. 37 (1979)
L521–L522, https://doi.org/10.1016/S0020-1693(00)95493-8.
[26] K.A. Earl, G. Jia, P.A. Maltby, R.H. Morris, J. Am. Chem. Soc. 113 (1991)
3027–3039, https://doi.org/10.1021/ja00008a035.
[27] G.M. Bancroft, M.J. Mays, B.E. Prater, F.P. Stefanini, J. Chem. Soc. A (1970)
2146–2149, https://doi.org/10.1039/J19700002146.
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
[28] P.A. Maltby, M. Schlaf, M. Steinbeck, A.J. Lough, R.H. Morris, W.T. Klooster, T.
F. Koetzle, R.C. Srivastava, J. Am. Chem. Soc. 118 (1996) 5396–5407, https://doi.
org/10.1021/ja9529044.
[29] S.E. Boyd, L.D. Field, T.W. Hambley, D.J. Young, Inorg. Chem. 29 (1990)
1496–1500, https://doi.org/10.1021/ic00333a011.
Acknowledgement
[30] I.E. Buys, L.D. Field, T.W. Hambley, A.E.D. McQueen, J. Chem. Soc., Chem.
Commun 557–558 (1994), https://doi.org/10.1039/C39940000557.
[31] T.Y. Bartucz, A. Golombek, A.J. Lough, P.A. Maltby, R.H. Morris,
R. Ramachandran, M. Schlaf, Inorg. Chem. 37 (1998) 1555–1562, https://doi.org/
10.1021/ic971050g.
NSERC is thanked for a Discovery grant to RHM.
Appendix A. Supplementary data
[32] J.-G. Lee, B.-S. Yoo, N.-S. Choi, K.I. Park, S.I. Cho, C. Lee, S.W. Lee, J. Organomet.
Chem. 589 (1999) 138–149, https://doi.org/10.1016/S0022-328X(99)00392-7.
[33] B. Chin, A.J. Lough, R.H. Morris, C.T. Schweitzer, C. D’Agostino, Inorg. Chem. 33
(1994) 6278–6288, https://doi.org/10.1021/ic00104a043.
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.ica.2020.120124.
[34] T.P. Fong, C.E. Forde, A.J. Lough, R.H. Morris, P. Rigo, E. Rocchini, T. Stephan,
J. Chem. Soc., Dalton Trans 4475–4486 (1999), https://doi.org/10.1039/
A906717E.
References
[35] B.E. Rennie, R. Eleftheriades, R.H. Morris, J. Am. Chem. Soc. 142 (2020)
17607–17629, https://doi.org/10.1021/jacs.0c08000.
[1] B.E. Bursten, M.R. Green, Prog. Inorg. Chem. 36 (1988) 393–485, https://doi.org/
10.1002/9780470166376.ch5.
[36] M. Tilset, V.D. Parker, J. Am. Chem. Soc. 111 (1989) 6711–6717, https://doi.org/
10.1021/ja00199a034.
[2] A.B.P. Lever, Inorg. Chem. 29 (1990) 1271–1285, https://doi.org/10.1021/
ic00331a030.
[37] M. Tilset, V.D. Parker, J. Am. Chem. Soc. 112 (1990) 2843, https://doi.org/
10.1021/ja00163a080.
[3] R.H. Morris, Inorg. Chem. 31 (1992) 1471–1478, https://doi.org/10.1021/
ic00034a030.
9