Science - USA (2019-01-18)

INSIGHTS | PERSPECTIVES

sciencemag.org SCIENCE

GRAPHIC: A. KITTERMAN/

SCIENCE

make it a potentially attractive target metal
for use in TMCs. Iron-based TMCs would
have several advantages, such as earth-abun-
dancy, low toxicity, and strongly absorbing
charge-transfer (CT) excited states ( 4 ).
However, the photophysics of iron TMCs
have made their solar-driven applications
relatively rare ( 5 ). Molecular orbital order-
ing dictates the ability of TMC excited states
to separate charge within the complex and
maintain this separation long enough to
carry out productive chemical reactions.
Ruthenium(II) -based TMCs experience rel-
atively large octahedral splitting that drives
the eg and t2g metal-centered (MC) states

of ruthenium(II)-based TMCs far enough
apart that a ligand-centered (LC) state falls
between them ( 6 ). This molecular orbital
ordering generates a low-lying MLCT ex-
cited state that rapidly becomes populated
upon photoexcitation of a ruthenium(II)-
based TMC ( 7 ). By moving charge either
away from or onto the ligands, the result-
ing excited state maintains the potential to
drive a desired chemical reaction.
Iron(II) -based TMCs experience weak
ligand-field splitting, such that the eg state
falls below the LC state. The excited state is
a low-energy MC state that cannot be used
to carry out productive chemistry. The de-
velopment of iron(II)-based TMCs that have

long-lived CT excited states has been tackled

by inorganic chemists who have purposefully

designed ligands to overcome this issue. Dur-

ing the past several years, Wärnmark and

colleagues have developed iron TMCs con-

taining strong s-donor ligands based on N-

heterocyclic carbenes (NHCs). These ligands

destabilize the low-lying eg MC state so that

the MLCT excited state becomes the lowest-

energy transition.

Further research by the Gros and Wärn-

mark groups provided critical insight into

molecular design for iron(II)-based TMCs.

In parallel, they reported on iron(II) NHC

complexes with the longest triplet^3 MLCT

lifetimes to that point, 16.5 and 18 ps, re-

spectively ( 8 , 9 ). In 2017, Wärnmark and co-

workers reported an all-NHC coordination

iron(III) complex, [Fe(btz) 3 ]3+, where btz is

3,3 9 -dimethyl-1,1 9 - bis(p-tolyl)-4,4 9 - bis(1,2,3-

triazol-5-ylidene) , that extended the CT life-

time to 100 ps ( 10 ).

Kjær et al. build off these previous suc-

cesses using a two-pronged approach. First,

they used an exceptionally strongly s-donat-

ing ligand that is also negatively charged and

enforces a near perfect octahedral coordina-

tion sphere in order to substantially destabi-

lize the otherwise low-lying eg MC state in the

iron-containing complex ( 7 ), causing the LC

state to become the lowest unoccupied mo-

lecular orbital. The switch in orbital ordering

leads to a low-lying CT excited state that can

be used to drive chemical reactions.

Second, the newly reported iron-based

TMC contains iron(III) rather than iron(II).

The iron(III)-TMC contains an unpaired

electron in the t2g ground state, so both the

ground and ligand-to-metal charge transfer

(LMCT) excited states are doublets. The ad-

vantages of the^2 LMCT excited state are two-

fold: There is no excited-state energy loss as

a result of singlet-to-triplet conversion that

is ubiquitous in many TMCs with singlet

character, and the lower-lying MC scaven-

ger states (^4 MC and^6 MC) are less accessible

than the scavenger states (^3 MC and^5 MC) of

the^3 MLCT formed in iron(II)-based TMCs,

which reduces nonradiative losses.

For years, the goal of designing ligands for

iron-based TMCs has been to separate charge

in the excited state (forming LMCT or MLCT

states) and to reduce the nonradiative decay

through the MC scavenger states. This new

ligand design hits both marks. Kjær et al.

report that [Fe(phtmeimb) 2 ]+ (where (pht-

meimb)− is {phenyl[tris(3-methylimidazol-

1ylidene)]borate}−) has a record quantum

yield of 2% {0.3% greater than the standard,

[Ru(bpy) 3 ]2+} and a lifetime of 2 ns. The tri-

dentate, facial phtmeimb ligand provides the

key design element, namely, large s-donor

ability, which is more pronounced because of

the negative charge of the borate ligand.

The combination of time-correlated single-

photon counting and transient absorption

spectroscopy indicates that emission occurs

from the^2 LMCT state. Furthermore, the

[Fe(phtmeimb) 2 ]+ complex is a strong pho-

tooxidant [a standard reduction potential

E°(III*/II) = 1.0 V versus Fc+/0 , where Fc^ is fer-

rocenium] and photoreductant [E°(IV/III*)

= −1.9 V versus Fc+/0 ] and has been shown

to act as a facile photoredox agent in two

model reactions. With this demonstration,

iron complexes can now feasibly be used as

photosensitizers in energy-relevant small-

molecule activations, such as water oxidation

to produce hydrogen and reduction of carbon

dioxide into chemical feedstocks. j

REFERENCES

1. Y.-J. Yuan, Z.-T. Yu, D.-Q. Chen, Z.-G. Zou, Chem. Soc. Rev.
   46 , 603 (2017).


2. K. S. Kjær et al., Science 363 , 249 (2019).


3. D. W. Thompson, A. Ito, T. J. Meyer, Pure Appl. Chem. 85 ,
   1257 (2013).


4. B. Bozic-Weber, E. C. Constable, C. E. Housecroft, Coord.
   Chem. Rev. 257 , 3089 (2013).


5. Y. Liu, P. Persson, V. Sundström, K. Wärnmark, Acc. Chem.
   Res. 49 , 1477 (2016).


6. P. A. Scattergood, A. Sinopoli, P. I. P. Elliott, Coord. Chem.
   Rev. 350 , 136 (2017).


7. A. M. Brown, C. E. McCusker, J. K. McCusker, Dalton Trans.
   43 , 17635 (2014).


8. T. Duchanois et al., Eur. J. Inorg. Chem. 2015 , 2469 (2015).


9. T. C. B. Harlang et al., Nat. Chem. 7 , 883 (2015).


10. P. Chábera et al., Nature 543 , 695 (2017).

10.1126/science.aav9866

N

N

B

Fe

N

N

N

N

N

B

N

N

N N

N

Achieving photoinduced electron transfer

Photoexcited electrons must transfer between LC and MC orbitals to generate charge-transfer (CT) excited states

and survive long enough to drive subsequent chemical reactions.

Transfer from ligand to metal

In [Fe(phtmeimb) 2 ]+, the photoexcited electrons transfer from a Llled LC orbital to a MC t2g orbital and have been

shown to react with electron donors and acceptors in photoinitiated reactions.

Typical ruthenium(II) complexes

The ligand-centered (LC) orbital lies

below the eg orbitals and receives

photoexcited t2g^ electrons.

Most iron(II) complexes

The LC orbital lies above

the eg orbitals, and excited t2g

electrons transfer quickly

to the eg orbitals.

An iron(III) complex

Strong s-donor ligands drive the eg

orbital above the LC orbital. The sta-

bilized iron(III) center yields a partially

Llled t2g orbital to receive excited

electrons from a lower lying LC state.

EnergyMC t2g t2g

MC eg

LC

t2g

eg

eg

LC

LC LC LC

LC

LMCT

High-energy fuels

Long-lived CT states

Low-energy feedstocks

Photon

e–

226 18 JANUARY 2019 • VOL 363 ISSUE 6424

Iron complexes that stay excited
Most iron(II) complexes have short-lived, metal-centered (MC) photoexcited states that are unable to perform
chemical reactions. A new iron(III) complex reported by Kjær et al. overcomes these limitations.

Published by AAAS

on January 17, 2019^

http://science.sciencemag.org/

Downloaded from