of 42
S
1
Supporting Information for:
Characterization of an Fe
N
-
NH
2
Intermediate Relevant to
Catalytic N
2
Reduction to NH
3
John S. Anderson
ζ
, George E. Cutsail
III
ζ
, Jonathan Rittle
ζ
,
Bridget A. Connor
,
William A.
Gunderson
#
,
Limei Zhang
§
*
, Brian M. Hoffman
*
, Jonas C. Peters
†*
Division of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena, California 91125,
United States
Department of Chemistry,
Northwestern University, 2145 Sheridan Road, Evanston, Illinois 60208, United States
§
Current address:
Department of Biochemistry and Redox Biology Center,
University of Nebraska
-
Lincoln,
Lincoln, NE 68588, United
States
#
Current address: Department of C
hemistry, Illinois College, 1101 West College Avenue, Jacksonville, IL, 62650,
United States
E
m
ail:
jpeters@caltech.edu, bmh@northwestern.edu,
limei.zhang@unl.edu
General Considerations
................................
................................
................................
.................
S
3
Figure S1. 10 K EPR spectrum of [(TPB)Fe
N
-
NH
2
][BAr
F
4
] (2).
................................
............
S
10
Figure S2. 77 K EPR spectrum of the addition of 1 equivalent of HBAr
F
4
·
2 Et
2
O to complex
1
.
................................
................................
................................
................................
.....................
S
12
Figure S3. XRD Structure of [(TPB)Fe
NAd][BAr
F
4
].
................................
............................
S
13
Figure S4. Mössbauer spectra of [(TPB)Fe
N
-
NH
2
][BAr
F
4
],
2
, at 80 K.
................................
.
S
14
Figure S5. Mössbauer spectra of
(TPB)Fe(N
2
), [(TPB)Fe][BAr
F
4
], and
[(TPB)Fe(N
2
)][Na(Et
2
O)
x
] at 80 K.
................................
................................
............................
S
15
Figure S6. Mössbauer spectra obtained on [(TPB)Fe(N
2
)][
Na(Et
2
O)
x
] under the listed
conditions.
................................
................................
................................
................................
...
S
16
Figure S7. Considered fits to the Mössbauer spectrum of [(TPB)Fe
N
-
NH
2
][BAr
F
4
],
2
,
at 80 K.
................................
................................
................................
................................
.....................
S
17
Figure S8.
Depiction of the ENDOR modeled structure of [(TPB)Fe
N
-
NH
2
][BAr
F
4
],
2
, and the
orientation of hyperfine c
oupling tensors with respect to g.
................................
.......................
S
18
Figure S9. Three
-
pulse ESEEM waveform and FT spectrum for [(TPB)Fe
N
-
NH
2
][BAr
F
4
],
2
.
................................
................................
................................
................................
.....................
S
19
Figure S10.
11
B ENDOR Spectrum of
[(TPB)Fe
N
-
NH
2
][BAr
F
4
],
2
.
................................
.......
S
20
Figure S11. VMT
-
PESTRE of
[(TPB)Fe
N
-
NH
2
][BAr
F
4
],
2
,
at
g
2
.
................................
..........
S
21
Figure S12.
Davies ENDOR segments from Figures 4 and 5 on an absolute RF scale to detail
31
P
couplings.
................................
................................
................................
................................
....
S
22
S
2
Figure S13.
M06L/TZVP(Fe)/SVP(P,N,B)/6
-
31G(C, H) optimized structure of
[(TPB)Fe(HNNH)]
+
.
................................
................................
................................
...................
S
23
Figure S14.
M06L/TZVP(Fe)/SVP(P,N,B)/6
-
31G(C, H) optimized structure of
[(TPB)Fe(HNNH)]
+
with spin density plotted with an isovalue of 0.006.
................................
.
S
24
Figure S15. Comparison and evolution of EPR spectra associated with addition of acid to
1
at
-
136 °C and
-
78 °C.
................................
................................
................................
......................
S
25
Table S1
. Tabulated Mössbauer parameters obtained by simulation of spectra shown in Figure
S5.
................................
................................
................................
................................
...............
S
26
Table S2. Tabulated parameters
obtained by the three simulations shown in Figure S7.
..........
S
26
Table S3. DFT optimized energies [kcal/mol] for [(TPB)Fe(N
2
H
2
)]
+
congen
ers.
....................
S
26
Table S4. Extended X
-
ray absorption fine structure (EXAFS) curve
-
fitting results for
[(TPB)Fe
N
-
NH
2
][BAr
F
4
],
2
.
................................
................................
................................
.....
S
26
Table S5.
Hyperfine coupling values (
A
) observed by ENDOR spectroscopy for
2
. All values in
units of MHz.
................................
................................
................................
..............................
S
27
Table S6. Crystal data and structure refinement for [(TPB)Fe
NAd][BAr
F
4
].
.........................
S
28
Table S7. M06L/TZVP(Fe)/SVP(P,N,B)/6
-
31G(C, H) DFT Optimized coordinates [Å] for
[(TPB)Fe
N
-
NH
2
]
+
,
2
.
................................
................................
................................
................
S
29
Table S8. M06L/TZVP(Fe)/SVP(P,N,B)/6
-
31G(C,H) DFT Optimized coordinates [Å] for
[(TPB)Fe(HNNH)]
+
.
................................
................................
................................
...................
S
32
Table S9. BP86/6
-
31G(d)(Fe,P,N,B)/6
-
31G(C,H) DFT Optimized coordinates [Å] for
[(TPB)Fe
N
-
NH
2
]
+
,
2
.
................................
................................
................................
................
S
35
Table S10. BP86/6
-
31G(d)(Fe,P,N,B)/6
-
31G(C,H) DFT Optimized coordinates [Å] for
[(TPB)FeHN=NH]
+
.
................................
................................
................................
....................
S
38
S
3
General Consideratio
ns.
Unless otherwise noted, all compounds were purchased from
commercial sources and used without further purification. [(TPB)Fe(N
2
)][Na(
12
-
crown
-
4)
2
],
1
(TPB)Fe(N
2
),
1
HBAr
F
4
·
2 Et
2
O,
2
[Cp
2
Fe][BAr
F
4
],
3
and KC
8
,
4
were prepared according to
literature procedures ([BAr
F
4
] = [
(3,5
-
(CF
3
)
2
-
C
6
H
3
)
4
B]
-
)
. All manipulations were carried out
under an N
2
atmosphere utilizing standard glovebox or Schlenk techniques. Solvents were dried
and de
-
oxygenated by an argon sparge followed by passage through an activated alumina column
purchased from S.G. Waters Company. Labeled
15
N
2
(98% purity) was obtained from
Cambridge
Isotope Laboratories.
EPR
Spectroscopy
.
EPR X
-
band spectra were obtained on a Bruker EMX spectrometer with the
aid of Bruker Win
-
EPR software suite version 3.0. The spectrometer was equipped with a
rectangular cavity which operated in the TE
102
mode. Temperature control was achieved with
the use of an Oxford continuous
-
flow helium cryostat (temperature range 3.6
300 K). All
spectra were recorded at 9.37 GHz with a microwave power of 2 mW, a modulation amplitude of
4 G, and a modulation frequ
ency of 100 kHz. Simulations were performed with the Easyspin
software suite.
5
EPR samples were thawed to
-
40 °C with a dry ice/acetonitrile slush bath for a
time period of 5 minutes or alternately thawed to room temperature for a time period of 5
minute
s.
Spin integration was performed in triplicate by preparing samples of
2
as described
below. Integration was performed by doubly integrating the signal from
2
and comparing the
resulting values to those from a solution of
1
with identical [Fe], volume, an
d scan parameters.
The three runs provided values of 90%, 78%,
and
92% for an average yield of 87(8)%.
EXAFS Measurements.
The
extended X
-
ray absorption fine structure (EXAFS) data
collection
was conducted at the Stanford Synchrotron Radiation Laboratory (SSRL) with the SPEAR 3
storage ring containing 500 mA at 3.0 GeV. Fe K
-
edge data were collected on the beamline 9
-
3
operating with a wiggler field of 2 T and employing a Si(220) do
uble
-
crystal monochromator.
Beamline 9
-
3 is equipped with a rhodium
-
coated vertical collimating mirror upstream of the
monochromator and a bent
-
cylindrical focusing mirror (also rhodium
-
coated) downstream of the
monochromator. Harmonic rejection was accomp
lished by setting the energy cutoff angle of the
mirrors to 10 keV. The incident and transmitted X
-
ray intensities were monitored using
Nitrogen
-
filled ionization chambers, and X
-
ray absorption was monitored by measuring the Fe
K
α
fluorescence intensity us
ing an array of 100 Canberra germanium detectors. During data
collection, samples were maintained at a temperature of approximately 10 K using an Oxford
instruments liquid helium flow cryostat. The energy was calibrated by reference to the absorption
of a
standard iron metal foil measured simultaneously with each scan, assuming a lowest energy
inflection point of the iron foil to be 7111.2 eV.
The EXAFS oscillations
χ
(
k
) were quantitatively analyzed by non
-
linear least square curve
-
fitting using the EXAFSP
AK suite of computer programs
.
6
Ab
-
initio theoretical phase and
S
4
amplitude functions were calculated using the program FEFF version 8.
7
No smoothing, filtering,
or related operations were performed on the data.
Computational Methods
.
Geometry optimizations
were performed using the Gaussian03 or the
Gaussian09 packages.
8
,
9
The BP86 exchange
-
correlation functional was employed with a 6
-
31G(d) basis set on Fe, P, N, and B and the 6
-
31G basis set on C and H for the Gaussian03
calculations. Alternately, the M0
6L functional with the TZVP basis set on Fe, the SVP basis set
on P, B, and N, and the 6
-
31G basis set on C and H. A full frequency calculation was
performed on each structure to establish true minima. The initial geometries used for the
calculations we
re the XRD coordinates for [(TPB)Fe
NAd][BAr
F
4
] which was modified to
feature the correct nitrogenous ligands. Computed energies were corrected for thermal energy.
Structural models were generated as .mol files and plotted in Diamond 3.2 and orbital/spin
density pictures were generated from
GaussView
03.
Synthesis of [(TPB)Fe
N
-
NH
2
][BAr
F
4
], 2
.
Complex
1
(4 mg, 0.004 mmol) was dissolved in 2
-
MeTHF (0.25 mL) and added into a quartz EPR tube. This solution was then frozen in a liquid
N
2
cooled cold well. A thawing solution of HBAr
F
4
·
2 Et
2
O (38 mg, 0.037 mmol) in 2
-
MeTHF
(0.25 mL) was then added to the EPR tube and frozen before reaching the bottom of the tube. A
long needle, which had also been cooled to 77K, was then inserted into t
he tube and used to
mechanically mix the solutions
as thawing gels
over 10 minutes. The tube was also raised
slightly out of the cold well to slightly thaw the solutions and aid in mixing. After mixing was
complete, the dark red color of
1
had disappeare
d and a brown yellow solution was obtained.
This frozen tube solution was then brought out of the glovebox and subjected to EPR analysis.
Synthesis of (TPB)Fe
N
-
Ad.
Under
an atmosphere of nitrogen, a 0.09
M solution of adamantyl
a
zide in benzene (1.7 mL,
0.148
mmol) was added to a solution of (TPB)Fe
(
N
2
)
(90 mg, 0.133
mmol) in benzene (5 mL) and stirred, giving a deep red solution. The solution was transferred to
a sealed Schlenk tube and stirred overnight at 80
C resulting in a deep green solution.
Lyoph
ilization yielded a green powder which was washed twice in THF leaving a bright green
powder ((TPB)Fe
N
-
Ad)
(58 mg, 0.073 mmol, 55
%)
. The product was suspended in benzene
and lyophilized in preparation for elemental analysis. Elemental analysis yielded the
following
results: Calculated: C 69.44, H 8.74, N 1.76; Experimental: C 69.29, H 8.47, N 1.67. NMR
and
UV
-
Vis analysis were
not obtained due to the poor solubility properties of the product.
Synthesis of [(TPB)Fe
N
-
Ad][BAr
F
4
]
, 4
.
A solution of (TPB)Fe
N
-
Ad (12.5 mg, 0.015
mmol)
in THF (3 mL) was added to a solution of [FeCp
2
][BAr
F
4
] (16.5 mg, 0.015
mmol) in THF (2
mL) giving a dark green solution which was stirred for 15 minutes. The solvent was removed,
leaving a dark green sludgy solid [(TPB)Fe
N
-
Ad][B
Ar
F
4
] which was washed three times in
pentane. Crystals
of [(TPB)Fe
N
-
Ad][BAr
F
4
] (13.5 mg,
0.008 mmol, 53%)
for XRD were
obtained by vapor diffusion of pentane into a solution of [(TPB)Fe
N
-
Ad][BAr
F
4
] in diethyl
ether.
1
H NMR (THF
-
d
8
,
δ
): 26.56 (vbr s),
10.66 (br s), 7.77 (s, BAr
F
4
-
), 7.69 (br s), 7.56 (s,
BAr
F
4
-
), 6.67 (vbr s), 6.48 (br s), 5.44 (br s), 4.96 (br s), 3.16 (br s),
-
2,85 (vbr s),
5.96 (br s),
-
S
5
6.53 (br s),
-
8.23 (vbr s),
-
13.96 (br s).
Solution magnetic moment (THF
-
d
8
): 3.9 μ
B
. UV
-
Vis
(THF)
λ
max
, nm: 655.
Anal. calcd. for C
78
H
81
B
2
F
24
FeNP
3
: C 56.48 , H 4.92, N 0.84; found: C
56.24, H 4.83, N 0.80.
Oxidation of (TPB)Fe(N
2
) with acid.
A suspension of (TPB)Fe(N
2
) (20 mg, 0.030 mmol) in
Et
2
O (3 mL) was cooled to
-
78 °C in a dry ice/acetone c
ooled cold well. A similarly cooled
solution of HBAr
F
4
·
2 Et
2
O (
30 mg, 0.030 mmol) in Et
2
O was then added to the suspension of
(TPB)Fe(N
2
) with stirring. The mixture was allowed to stir at
-
78 °C for 10 minutes before being
warmed to room temperature and
stirred for an additional 2 hours. As it warmed the mixture
became homogenous and lightened in color to provide an orange solution. The solution was then
concentrated to 1.5 mL and subjected to a pentane vapor diffusion at
-
35 °C. The product
[(TPB)Fe][BAr
F
4
] was obtained as a dark orange crystalline material (42 mg, 0.028 mmol, 93%).
The spectroscopic signatures of [(TPB)Fe][BAr
F
4
] were identical to those previously reported.
10
Preparation of ENDOR samples.
Samples of
1
were prepared by dissolving the complex (1.2
mg, 0.001 mmol) in 2
-
MeTHF (0.5 mL) and transferring this solution to an ENDOR
tube
before
freezing the solution pr
ior to shipping. Samples of
2
were prepared analogo
usly as described
above with
1
(1 mg, 0.000
9 mmol) and HBAr
F
4
·
2 Et
2
O (10 mg, 0.0099 mmol) in 2
-
MeTHF
(total volume of 0.5 mL). The only deviation from the above
described procedure was that
2
was
initially prepared in a 5 mL scintillation vial before being slightly thawed and transferred into th
e
ENDOR tube while cold. The samples were then frozen and packaged for shipment. They were
shipped to Northwestern University under liquid N
2
.
Preparation of
15
N labeled ENDOR samples.
Two stock solutions of
1
(5 mg, 0.0048 mmol)
and HBAr
F
4
·
2 Et
2
O (50
mg, 0.049 mmol) in 2
-
MeTHF (0.5 mL each) and 2
-
MeTHF (1 mL)
were each placed into 3 short test tubes. These test tubes and the ENDOR tubes were placed into
a round
-
bottom
S
chlenk flask with a glass stopcock side
-
arm. The flask was then sealed with a
rubb
er septum and brought out of the glovebox. The solution was freeze
-
pump
-
thawed 3x
before backfilling with an atmosphere of
15
N
2
. The solutions were then sparged with
15
N
2
from
the headspace of the flask with the use of a long needled syringe through the
rubber septum. The
syringe was rinsed with the 2
-
MeTHF from the third test tube to avoid cross con
tamination. The
solution of
1
was then distributed to the ENDOR tubes with the use
of the syringe. To generate
2
, the apparatus was cooled to 77 K with
liquid nitrogen. The solutions were then briefly thawed
and the solution of HBAr
F
4
·
2 Et
2
O was layere
d on top of the solutions of
1
in the ENDOR
tubes. The solutions were then mechanically mixed with the syringe
needle
as described above.
After mixing
was complete, the solutions were frozen, and the septum was removed after which
the samples were quickly dumped into liquid N
2
before being sealed for shipment. The samples
were shipped to Northwestern University under liquid N
2
.
Preparation of XAS sample
s.
The sample of
2
for XAS analysis was prepared analogously to
that reported earlier with
1
(10 mg, 0.0095 mmol) and HBAr
F
4
·
2 Et
2
O (100 mg, 0.099 mmol) in
2
-
MeTHF (0.5 mL total volume). The solution was prepared in a 5 mL scintillation vial before
S
6
bei
ng slightly thawed and transferred into the XAS sample holder with a syringe.
While care was
taken to keep the sample as cold as possible during these manipulations, the required thawing of
the solution for syringe transfer likely raised the temperature of
the solution substantially.
The
sample was then re
-
frozen before being packaged for shipping to SSRL. The sample was
shipped under liquid N
2
.
Preparation and EPR analysis of
2
at
-
78 °C.
Complex
1
(8 mg, 0.0076 mmol) was suspended
in Et
2
O (0.25 mL) and
cooled to
-
78 °C in a 5 mL scintillation vial. HBAr
F
4
·
2 Et
2
O (80 mg,
0.079 mmol) was dissolved in Et
2
O (0.25 mL) in a 20 mL scintillation vial equipped with a stir
bar and cooled to
-
78 °C as well. Once cooled, the suspension of
1
was added to the HB
Ar
F
4
·
2
Et
2
O with stirring. The resulting solution was dark yellow
-
brown and homogeneous and was
allowed to stir at
-
78 °C for 10 minutes. To this solution was then added 2
-
MeTHF (0.5 mL)
which had been similarly cooled to
-
78 °C. The resulting solutio
n was then transferred into an
EPR tube and frozen with liquid N
2
before analysis.
EPR analysis of the
reaction of complex 1
with one equivalent of HBAr
F
4
·
2 Et
2
O.
The
procedure employed was identical to tha
t used for the generation of
2
with the exception that
only one equivalent of HBAr
F
4
·
2 Et
2
O (3.8 mg, 0.0038 mmol) was used.
Preparation and Interpretation of Mössbauer Analysis of 2
.
A 500 μL aliquot of
[
(TPB)
57
Fe
(
N
2
)][Na(Et
2
O
)
x
]
(5 mM, Et
2
O) was frozen in a 5 mL scintillation vial i
nside of a cold
well chilled to 77 K. A 500 μL aliquot of chilled
HBAr
F
4
·
2 Et
2
O
(25 mM, Et
2
O) was
subsequently layered on top of the Fe
-
containing layer and frozen. The vial was elevated off of
the cold well floor with pre
-
chilled forceps and the mixture subsequently began to melt (T ~
-
110
o
C). To facilitate rapid mixing at the lowest possible
temperature, the two layers were
mechanically stirred with a similarly cooled spatula. The dark orange
-
red color of
[(TPB)
57
Fe(N
2
)][Na(Et
2
O)
x
]
rapidly bleached on mixing and was replaced with a light yellow
-
green color. This solution was subsequently chill
ed to near
-
freezing before being poured into a
M
ö
ssbauer sample cup chilled to 77 K.
The 80 K
57
Fe M
ö
ssbauer spectrum of the resulting mixture is shown in Figure
3
of the main text
and reprinted for clarity in Figure
S
4
A of the supporting information
. This
spectrum was
collected in the presence of a 50 mT magnetic field, which served to sharpen quadrupole
doublets associated with
S
= 1/2 species (
vide infra
). Six distinct features of similar intensity are
observed, suggesting the presence of multiple Fe
-
con
taining species. Similarly prepared samples
show these features in varying ratios, eliminating the possibility that the six features arise from a
single Fe species.
57
Fe M
ö
ssbauer spectra of independently prepared samples of (TPB)
57
Fe(
N
2
),
[
(TPB)
57
Fe][
Bar
F
4
]
, and
[(TPB)
57
Fe(N
2
)][Na(Et
2
O)
x
]
in frozen Et
2
O solutions are shown in
Figure
S
5
of the supporting information
and their par
ameters are tabulated in Table
S
1.
At 80
K, [
(TPB)Fe
N
-
NH
2
][
BAr
F
4
]
and
[(TPB)
57
Fe(N
2
)][Na(Et
2
O)
x
]
are slowly
-
relaxing
Kramers systems, as evidenced by the EPR signatures observed at this temperature. At
S
7
sufficiently low temperatures, this magnetic interaction may serve to broaden and/or split the
quadrupole doublet features associated with these specie
s in their M
össbauer spectra
. To
establish conditions wherein this magnetic hyperfine interaction is minimized, M
ö
ssbauer
spectra of
[(TPB)
57
Fe(N
2
)][Na(Et
2
O)
x
]
were collected under a
variety of conditions (Figu
re S6
).
At temperatures lower than 20 K, a broadened spectrum consisting of at least four features is
observed. At 80 K in the absence of an applied magnetic field, a broad, asymmetric quadrupole
doublet is observed, displaying the onset of slow magnetic relaxati
on. Application of a 50 mT
magnetic field at this temperature sharpens this quadrupole doublet, and removes most of the
asymmetry. Under these conditions (80 K, 50 mT applied field) we anticipate that
[(TPB)Fe
N
-
NH
2
][BAr
F
4
]
will also display a relatively s
harp quadrupole doublet.
To deconvolute the mixture of species obtained from low temperature protonation of
[(TPB)
57
Fe(N
2
)][Na(Et
2
O)
x
]
, we initially assumed that a finite quantity of the 1
-
electron
oxidized
congener, (TPB)Fe(
N
2
)
, would be present. (TPB)Fe
(
N
2
)
has a signature large quadrupole
splitting (3.34 mm/s) that produces featu
res at greater velocities than [
(TPB)
57
Fe][
Bar
F
4
]
or
[(TPB)
57
Fe(N
2
)][Na(Et
2
O)
x
]
. Therefore the
presence of (TPB)Fe(
N
2
)
is easily discerned and its
presence in the mixture is high
lighted by the two red lines in Figure
S4
A
. (TPB)Fe(
N
2
)
constitutes ca. 19% of the total Fe content of this sample and subtraction of its features produces
the spectrum shown in Figure
S4
B. Four distinct features remain in this spectrum. Unfortunately,
each of these features has roughly equal integrals, and three different simulations produce nearly
equivalent fits to the data. The three simulations are shown in Figure
S7
and the resulting
parameters are tabulated in Table
S
2.
Inspection of the hyperfine parameters obtained for the simulation shown in Figure
S7
C suggests
that this solution is unlikely. The isomer shift values, (
δ
=
-
0.32 and 1.38) are unreasonably
small/large for known Fe c
ompounds supported by the TPB ligand, and we hence discarded this
solution. While the hyperfine parameters obtained in the simulation shown in Figure
S7
B are not
unreasonable (
δ
= 0.19, 0.87 mm/s) we have thus far not obtained M
ö
ssbauer spectra on
(TPB)Fe
compounds that display isomer shift values higher than 0.8 mm/s. We favor the
simulation shown in Figure
S7
A, as one component displays very s
imilar hyperfine parameters
to [(TPB)Fe][
BAr
F
4
]
, a species we also believe is formed via overall 2
-
electron oxida
tion of
[(TPB)
57
Fe(N
2
)][Na(Et
2
O)
x
]
and release of H
2
.
Although this spectrum has been fit assigning this
species to [(TPB)Fe][BAr
F
4
], there are a variety of other
S
= 3/2 complexes on the TPB manifold
that have similar Mössbauer parameters and may a
lso be
reasonable assignments.
The remaining
component,
[(TPB)Fe
N
-
NH
2
][BAr
F
4
]
, displays hyperfine parameters most similar to
[(TPB)
57
Fe(N
2
)][Na(Et
2
O)
x
]
, although with a smaller isomer shift value. In the related (SiP
3
)Fe
system, functionalization of a bound N
2
ligand with silyl electrophiles was similarly found to
reduce the isomer shift value relative to the Fe
-
N
2
complex.
11
At present, we cannot rule out the
solution shown in Figure S
7
B and further studies will be needed to distinguish between these
two simula
tions. Summarizing, M
ö
ssbauer spectroscopy indicates that at least three major Fe
-
S
8
containing species are generated upon protonation of
[(TPB)
57
Fe(N
2
)][Na(Et
2
O)
x
]
at low
temperatures in Et
2
O, at least under the conditions needed to collect these M
ö
ssbauer s
pectra.
Collection and Interpretation of ENDOR
Data
.
‘Q
-
band’ ~35 GHz continuous wave (CW)
electron paramagnetic resonance (EPR) spectra were collected at 2 K with a helium immersion
dewar on a modified Varian E
-
110 by the digitization of the RC
-
smooth out
put signal under
‘rapid adiabatic’ and with 100 kHz field modulation (2.0 G modulation amplitude).
Pulsed Q
-
band EPR and electron nuclear double resonance (ENDOR) experiments were
executed on a custom
-
built instrument previously described at 2 K with data
acquisition
preformed with the SpecMan software package (specman4epr.com) in conjunction with a Spin
-
Core PulseBlaster ESR_PRO 400 MHz word generator and Agilent Technologies Acquiris
DP235 500MS/sec digitizer. ENDOR spectra were collected using either t
he Davies microwave
‘3 pulse’ sequence (
π
T
π
/2
τ
π
τ
-
echo), ‘3 pulse’ Mims (
π
/2
T
π
/2
τ
π
/2
τ
-
echo), or a ‘4 pulse’ refocused Mims, termed a ReMims, (
π
/2
τ
1
π
/2
T
π
/2
τ
2
π
τ
2
echo
), where the RF pulse is applied during time
T
and the RF frequency is randomly hopped in
each pulse sequence. A blindspot occurs at
A
τ
=
n
(where
n
= 0, 1, 2, ...) for a Mims sequence,
however, the ReMims allows for ENDOR collection of weakly coupled nucle
i with no
blindspots.
The ENDOR spectrum from a nucleus with spin of
I
=
½
(
1
H,
15
N,
31
P) and from the
m
s
= ±
½
electron
-
spin manifold exhibits a doublet at frequencies,
(1)
where
v
n
is the nuclear Larmor frequency and
A
is the hyperfine coupling. When
I
1 (
11
B,
14
N),
a nuclear quadrupole interaction (
P
) introduces further splitting of the
ν
±
manifolds.
(2)
All ENDOR spectral simulations were performed in Matlab with the EasySpin v5 (easyspin.org)
toolbox.
The absolute hyperfine coupling signs and sp
in density signs are obtained through the Pulsed
ENDOR SaTuration and REcovery protocol. Extensively described elsewhere
,
12
the PESTRE
protocol determines the absolute sign of
ρ
=
A
/
g
n
(where
g
n
is the nuclear g value for a given
nuclide) through three dif
ferent stages of repeating Davies ENDOR sequences. In the first stage,
with no RF frequency applied during the Davies sequence, a baseline (BSL) of the electron spin
echo (ESE) response is measured. Next, the ENDOR response is saturated by now applying a
n
RF pulse during the Davies sequence during the second stage. Lastly, the RF pulses are turned
off in the third stage and the ESE is allowed to relax back to the BSL to measure the DRL
response. The two distinct relaxation possibilities, a positive resp
onse of a relaxtion from above
v
±
=
A
2
±
v
n
v
±
,
m
I
=
v
n
±
3
P
(
2
m
I
1
)
2
S
9
to down to the BSL, or a negative response from below up to the BSL is the
δ
DRL response
(
δ
DRL = DRL
-
BSL). The positive or negative sign of the
δ
DRL determines the spin density
sign and therefore the absolute hyperfine sig
n.
VMT
-
PESTRE
Based on
the
original PESTRE protocol, stages of Davies microwave pulse sequences are
employed with and without an fixed applied RF (ENDOR) frequency to observed saturation and
relaxation behavior (termed
δ
DRL) of a given ENDOR response. VMT
-
PESTRE compares the
relaxation
behavior
of the ENDOR
response
as the ‘mixing time’ is lengthened, eventually
inducing a ‘flip’ of the relaxation
response
, confirming the initial, sometimes weak, relaxation
observation o
f ‘fast’ mixing times.
ESEEM
Electron spin echo envelope modulation (ESEEM) spectra were collected using a 3 microwave
pulse sequence (
π
/2
-
τ
-
π
/2
-
Δ
T
-
π
/2
-
τ
-
echo
) with phase cycling. Simulations of the ESEEM
results were performed using the MA
TLAB based OPTESIM software package.
13
The spin
Hamiltonian for interaction of the Fe(I) electron spin (
S
=1/2) with the remote
14
N nuclear spin (
I
= 1) is formulated with a nuclear Zeeman term, a hyperfine (hf,
A
) and a nuclear quadrupole
(nqi) term, as fol
lows:
In this expression,
g
n
is the nuclear
g
-
value,
β
n
is the nuclear magneton,
S
is the electron spin
operator,
I
is the nuclear spin operator,
A
is the hf coupling tensor, and
Q
is the nqi tensor. The
hf tensor has the principal components,
A
= [
A
xx
A
yy
A
zz
], and is composed of an isotropic part,
, and a dipolar part,
T
dip
= A
A
iso
. The nqi tensor has the principal components,
Q
= [
Q
xx
Q
yy
Q
zz
], and is defined by the nuclear quadrupole coupling constant,
e
2
qQ
/
h
, and the
electric field gradient as
ymmetry parameter,
η
. In its principal axis system (PAS),
Q
is related to
e
2
qQ
/
h
and
η
by the following expressions:
The orientation between the nqi tensor PAS and the hf tensor PAS is defined by the Euler angles,
[
α
Q
,
β
Q
,
γ
Q
]. For a system with more
than one coupled nucleus, the orientation of the hf PAS of
each nucleus relative to the g
-
tensor reference frame is defined by the Euler angles, [
α
A
,
β
A
,
γ
A
].
The OPTESIM optimization uses the Nelder
-
Mead simplex method.
H
=
g
n
β
n
S
!
B
"
!
+
h
S
!
A
I
!
+
I
!
Q
I
!
a
i
s
o
=
1
3
A
i
i
i
Q
z
z
=
e
2
qQ
2
I
(
2
I
1
)
h
η
=
Q
x
x
Q
y
y
Q
z
z
S
10
14/15
N ENDOR
: Full interpretatio
n of the ENDOR data is convoluted by the
11
B
ν
-
feature and
harmonics described in the caption of Figure 5 of the main text. The incomplete tensor of
2
-
(
14
N
)
obtained from ENDOR spectroscopy was better estimated from ‘Q
-
band’ ESEEM spectroscopy
(Figure
S9
)
. The single
-
field three microwave pulse ESEEM spectrum of
2
-
(
14
N)
is well
simulated with a single
14
N tensor of A = [4.34, 7.18, 6.22] MHz and quadrupole parameters of
e
2
qQ
/
h
= 1.76 and
η
= 0.64. The hyperfine tensor is within reasonable agreement with
the two
principle hyperfine values obtained by ENDOR spectroscopy. As there is no orientation
assignment of the ESEEM hyperfine tensor
A
with respect to
g
from the single field position, it
is not known whether the maximum hyperfine coupling from the ESEE
M simulation is not
observed by ENDOR spectroscopy or is a slight overestimation as an artifact of high
-
frequency
ESEEM
.
However, the
e
2
qQ
/
h
= 1.76 quadrupole parameter is equivalent to an ENDOR
quadrupole splitting of
P
max
= 2[
e
2
qQ
/(4I(2I
-
1)] = 0.88 MHz, in excellent agreement with the
observed
P
2
= 0.90 MHz. Additionally, the ESEEM simulation estimates
P
rotated away from
A
as described an Euler angle,
α
= 60°.
11
B ENDOR
:
An unusually large coupling of 21 MHz is observed for
1
whi
ch decreases to 9
MHz upon protonation to form
2
coupled with a decrease in 2s
11
B unpaired orbital spin density
from 0.40% to 0.17% (Figure S10
). This observation is also consistent with the assignment of
2
as a hydrazido(2
-
) species, as an increased Fe
-
N bond order would be accompanied by an
increased Fe
-
B distance, as observed in other (TPB)Fe
NR species such as [(TPB)Fe
N
-
Ad][BAr
F
4
]. The decrease in coupling constant, although not explicitly required by an increase in
the Fe
-
B distance, is fully consis
tent with a longer Fe
-
B interaction.
Based on
the
original
PESTRE protocol,
12
stages of Davies microwave pulse sequences are employed with and without
an fixed applied RF (ENDOR) frequency to observed saturation and relaxation behavior (termed
δ
DRL) of a given ENDOR re
sponse. VMT
-
PESTRE compares the relaxation
behavior
of the
ENDOR
response
as the ‘mixing time’ is lengthened, eventually inducing a ‘flip’ of the
relaxation
response
, confirming the initial, sometimes weak, relaxation observation of ‘fast’
mixing times.
As shown in Figure S
1
1
, the
ν
+
responses of the
11
B nuclei switches from a
positive
δ
DRL to a negative
δ
DRL response, indicated by an downward pointing arrow for the
Δ
VMT
-
δ
DRL (‘fast’ to ‘long’ mixing time). This is typical of a nuclei with negative spin
density
and a negative A as
ρ
A
/
g
n
(
g
n
> for
11
B,
1
H,
g
n
<
14
N). The
ν
yields
the anticipated
opposite VMT
-
PESTRE response (negative
δ
DRL to positive
δ
DRL).
Figure
S
1.
10 K EPR spectrum of [(TPB)Fe
N
-
NH
2
][BAr
F
4
] (
2
)
.
S
11
Conditions:
2
-
MeTHF, 10 K, microwave frequency 9.4 GHz,
microwave power, 20.313 mW.
0
500
1000
1500
2000
2500
3000
3500
4000
Magnetic Field (Gauss)
S
12
Figure
S
2
.
77 K EPR spectrum of the addition of 1 equivalent of HBAr
F
4
·
2 Et
2
O to complex
1
.
2800
3000
3200
3400
3600
Magnetic Field (Gauss)
Conditions: 2
-
MeTHF, 77 K, microwave frequency 9.4 GHz,
microwave power, 20.313 mW.
S
13
Figure
S3
.
XRD Structure of [(TPB)Fe
NAd][BAr
F
4
]
.
Note that hydrogens have been removed for clarity. Selected bond lengths (Å) and angles (°):
Fe(1)
-
N(1) = 1.660(2),
Fe(1)
-
P(1) = 2.3098(9), Fe(1)
-
P(3) = 2.3805(10), Fe(1)
-
P(2) = 2.3893(9),
P(1)
-
Fe(1)
-
P(3) = 103.85(3), P(1)
-
Fe(1)
-
P(2) = 110.10(4), P(3)
-
Fe(1)
-
P(2) = 113.50(3).
S
14
Figure
S4
.
M
ö
ssbauer spectra of [(TPB)Fe
N
-
NH
2
][BAr
F
4
],
2
,
at 80 K
.
(A) 80 K
57
Fe
Mössbauer spectrum obtained
by low temperature protonation of
[
(TPB)Fe
(
N
2
)][Na(Et
2
O
)
x
]
. The presence of (TPB)
Fe(
N
2
)
is highlighted with the red notches.
Th
e red trace shows a 19% (TPB)Fe(
N
2
)
component. (B) Spectrum obtained by subtracting out
19% (TPB)Fe
(
N
2
)
from the spectrum shown in (A).
S
15
Figure
S5
.
M
ö
ssbauer spectra of (TPB)Fe(N
2
), [(TPB)Fe][BAr
F
4
], and
[(TPB)Fe(N
2
)][
Na(Et
2
O)
x
]
at 80 K
.
57
Fe Mössbauer spectra of (A)
(TPB)Fe(
N
2
), (B) [(TPB)Fe][
BAr
F
4
], and (C) [(TPB)Fe(
N
2
)][
Na(Et
2
O)
x
]
. Spectra were
collected on dilute frozen solutions (5 mM, Et
2
O) in the presence of a
50 mT applied field applied parallel to the propagation of the gamma beam. The spectrum of
[(TPB)Fe(N
2
)][
Na(Et
2
O)
x
]
con
tains a 10% impurity of (TPB)Fe(
N
2
)
that was likely generated
during the brief exposure of the sample to air/water prior to immersion in liquid nitrogen.