Science - USA (2021-07-16)

INSIGHTS | PERSPECTIVES

278 16 JULY 2021 • VOL 373 ISSUE 6552 sciencemag.org SCIENCE

PIEZOELECTRICS

Autonomous

biocompatible

piezoelectrics

By Shlomo Berger

M

uscles provide mechanical forces

needed for the dynamic activity

of a human body. The mechanical

forces applied on various human

organs lead to different types of

physical movements (e.g., contrac-

tion and extension, rotation, and bending).

Injuries and diseases can disable muscle

functionality, which may lead to the pro-

nounced deterioration of a human activity

or even cause death. Surgical correction

is not always possible or does not always

provide optimal results. Implanted artifi-

cial biocompatible devices can replace de-

fective muscles. These should operate lo-

cally at the defective site by monitoring the

need for a specific physical movement and

then applying the correct mechanical force

to obtain it. These devices should also be

self-energized using the body’s energy re-

sources and programmed for optimal oper-

ation. On page 337 of this issue, Yang et al.

( 1 ) present a new approach for fabricating

thin piezoelectric biocompatible thin films

that actuate physical movements, demon-

strated on mice muscles, under applied

electric field.

Biocompatible piezoelectric materi-

als can provide sensing and actuating of

physical movements of organs inside a hu-

man body ( 2 ). They possess a piezoelectric

property ( 3 ) that interconverts electrical

and mechanical energy. The piezoelectric

response of materials results from a revers-

ible change in the length of neighboring

ionic bonds that create electric dipoles in

response to applied mechanical forces or

electric fields. Piezoelectric materials exist

naturally in a human body ( 4 ), such as in

bones and muscles (5, 6), and contribute

Faculty of Materials Science and Engineering, Technion–

Israel Institute of Technology (IIT), Haifa 32000, Israel.

Email: [email protected]

the transport of molecules in the vicinity of

the active site ( 6 ).

Zeolites are a diverse family of robust,

microporous aluminosilicate materials that

are widely used as catalysts in hydrocarbon

processing on an industrial scale. The ion-

exchanged metal sites in iron- and copper-

containing zeolites resemble key structural

components of the active sites in enzymes

( 2 ). Snyder et al. installed the same active

iron sites in the pores of two zeolites. Both

have similar silicon-to-aluminum ratios,

iron loadings, and cage diameters, and

might be expected to have nearly identical

activities toward methane. However, the

zeolite structures differ in the accessibility

of their iron sites. Beta zeolite (*BEA) has

a system of interconnected pores composed

of “large” 12-rings (Si 12 O 12 ), and no smaller

constrictions ( 7 ). Although chabazite (CHA)

cages have the same diameter, entry into

the cages must occur through smaller

8-ring apertures (Si 8 O 8 ) ( 7 ).

Methane hydroxylation in these zeolites

was studied as a sequence of stoichiometric

reactions at room temperature. First, the re-

duced Fe(II) sites were activated by the oxi-

dant, N 2 O, which installs reactive a-O atoms

(see the figure). When one of the activated

Fe(IV)O sites abstracts a hydrogen atom from

methane, a methyl radical and a hydroxy site,

Fe(III)OH, are simultaneously created in the

same zeolite cage. If rebound of the methyl

radical to the colocated Fe(III)OH site ensues,

the Fe(II) site is regenerated and can perform

the reaction sequence again. If the methyl

radicals diffuse away, they can be trapped by

neighboring ferryl sites, which are converted

to inert methoxy sites, Fe(III)OCH 3. Although

a C–O bond is formed in both types of reac-

tion, radical trapping at a ferryl site is un-

productive because neither of the resulting

Fe(III) sites (hydroxy or methoxy) can be re-

oxidized by N 2 O under mild conditions. The

ratio of radical escape relative to rebound is

very high in *BEA.

By contrast, diffusion of the methyl radical

in CHA zeolite is restricted by virtue of the

small size of the window through which the

radical must escape from the cage where it

is generated. This cage effect is reminiscent

of radical confinement in enzymes. Using

Mössbauer and resonance Raman spectros-

copies, Snyder et al. show that the yield of

Fe(II) (resulting from radical capture in the

same zeolite cage) under single-turnover

conditions is ~40% for CHA, whereas it is

near zero in *BEA. In CHA, the methanol

product migrates spontaneously from the

Fe(II) sites to Brønsted acid sites in the zeo-

lite pores, allowing a second reaction cycle

to take place. When the isotopic identity of

the methane was switched, from^13 CH 4 in the

initial cycle to^12 CH 4 in a subsequent cycle,

the isotopic composition of the methanol

changed, which demonstrates that Fe(II)

sites in CHA can be reactivated by N 2 O to

produce a second equivalent of methanol.

Efficient processes for converting meth-

ane to an energy-dense liquid hydrocarbon

such as methanol are important not only

for making greater use of abundant natu-

ral gas resources but also to reduce the

need to flare stranded natural gas. Rather

than generating the greenhouse gas CO 2

unproductively, conversion to methanol

would allow transport in a cost-effective

way to population centers where its en-

ergy, chemical value, or both, could be

extracted. However, using the strategy of

Snyder et al. to this effect will require the

process to become much more efficient.

For example, combining the oxidant with

methane in the zeolite should allow the re-

action sequence to proceed in a single step.

However, this approach presents a selectiv-

ity challenge. Because the relative difficulty

of activating a C–H bond in methane ver-

sus methanol is roughly constant, there is

a universal, catalyst-independent trade-off

between conversion and selectivity ( 8 ). It is

not yet clear how to achieve rapid diffusion

of methanol away from the active sites,

preventing its further oxidation, while si-

multaneously confining methyl radicals

near the active sites to form methanol and

regenerate Fe(II).

A second challenge is to replace the N 2 O

oxidant by a less expensive oxidant such as

O 2. Colocating two iron sites in a ferrierite

zeolite was recently shown to facilitate O 2

splitting ( 9 ). However, this geometry will

enhance the undesired ferryl trapping of

methyl radicals that leads to Fe(III)OH/

Fe(III)OCH 3 sites. The soluble methane

monooxygenase enzyme achieves methane

oxidation at diiron active sites linked by

bridging oxygens, Fe(IV) 2 (m-O) 2 , but it also

produces ferric sites that require an external

reductant for reactivation. A useful process

for converting stranded methane will need

to overcome both of these challenges. j

REFERENCES AND NOTES

1. N. J. Gunsalus et al., Chem. Rev. 117 , 8521 (2017).


2. K. T. Dinh et al., ACS Catal. 8 , 8306 (2018).


3. R. M. Bullock et al., Science 369 , eabc3183 (2020).


4. B. E. R. Snyder et al., Science 373 , 327 (2021).


5. X. Huang, J. T. Groves, J. Biol. Inorg. Chem. 22 , 1 85 ( 2017 ).


6. R. Breslow, Acc. Chem. Res. 28 , 146 (1995).


7. C. Baerlocher, W. M. Meier, D. H. Olson, Atlas of Zeolite
   Framework Types (Elsevier, ed. 6, 2007).


8. A. A. Latimer, A. Kakekhani, A. R. Kulkarni, J. K. Nørskov,
   ACS Catal. 8 , 6894 (2018).


9. E. Tabor et al., S c i. A d v. 6 , eaaz9776 (2020).

ACKNOWLEDGMENTS

S.L.S. acknowledges the US Department of Energy, Office

of Science, Division of Basic Energy Sciences, under the

Catalysis Science Initiative (DE-FG-02-03ER15467) for

financial support.

10.1126/science.abj4734

Implants that work

with muscles should

monitor, correct,

and be self-energized

0716Perspectives.indd 278 7/9/21 5:20 PM