6 MARCH 2015 • VOL 347 ISSUE 6226 1073 SCIENCE sciencemag.org
Once this had been accomplished, the
race was on to image single proteins with
the NV center. The first major milestone on
this path was the detection of small volumes
of nuclear spins, close to those volumes pre-
viously achieved by MRFM, except at room
temperature. This was done by two groups
simultaneously (7, 8). In the latter work, the
technique was pushed a step farther by inter-
rogating the nuclear spin ensemble to give
information on their local environment. By
applying these techniques to single proteins,
low-resolution images should be possible.
However, ultimately atomic-scale resolution
image of proteins are needed, at least in the
reaction centers. This requires the ability to
locate individual nuclei.
In the past year, detection and imaging
of near-single nuclear spins outside the diamond lattice was reported (9). Here, four
silicon nuclei were observed on a specially
designed glass-like coating placed on the
diamond surface. This work was important
because it proved that the magnetic noise
produced by unpaired electrons at or near
the diamond surface could be overcome, at
least in principle. Until then, this magnetic
noise was a roadblock that made many question whether a single external nuclear spin
could ever be detected by a NV center.
Although the earlier studies proved that
it is possible to greatly reduce the numbers
of troublesome diamond surface spins, it
has so far not been possible to eliminate
them completely. However, very recently it
has been shown that these few remaining
surface spins can often be stable enough
to use as a resource—namely, making them
into “reporter” spins to image single nu-
clei in their immediate environment (10); a
single proton bound to the diamond surface
However, what is ultimately needed is a
way to deploy reporter spins that are not
limited to the diamond surface but that can
be detected at long range and selectively
placed at critical locations in a protein where
they can sense conformation changes. It is
exactly this lofty goal that has now been accomplished by Shi et al. (see the figures). A
In the future, it will be necessary to monitor far more than
one nuclear spin in the protein,
and so better reporter spin labels
are needed. Specifically, they
must have longer spin coherence times to achieve longer-range nuclear
spin imaging, and they must be photostable
to the light used to interrogate the NV center. Finally, to image proteins in living cells,
the bulk diamond used in these experiments
must be replaced by a nanodiamond with
similar NV magnetic sensitivity. ■
1. J. S. Valastyan, S. Lindquist, Dis. Model. Mech. 7, 9 (2014).
2. F. Shi et al ., Science 347, 1135 (2015).
3. C. L. Degen, M. Poggio, H. J. Mamin, C. T. Rettner, D. Rugar,
Proc. Natl. Acad. Sci. U.S.A. 106, 1313 (2009).
4. G. Balasubramanian et al ., Nature 455, 648 (2008).
5. J. R. Maze et al ., Nature 455, 644 (2008).
6. B. Grotz et al ., New J. Phys. 13, 055004 (2011).
7. T. Staudacher et al ., Science 339, 561 (2013).
8. H.J.Mamin etal.,Science 339,557(2013).
9. C. Müller et al ., Nat. Commun.5, 4703 (2014).
10. A.Ajoy etal., Phys.Rev.X 5,011001(2015).
See it move. Dynamics of the spin label attached to the MAD2 protein.
B0 is the external magnetic field, ZM is the spin label quantization axis,
and ∆Φ is the angular range of motion.
Metallic conductivity is a familiar phenomenon, but metals can dis- play surprising and exotic behavior, such as superconductivity and the quantum Hall effect. Advances in anotechnology have led to new
questions about how metallic conductivity
might change in structures with dimensions
approaching a few atoms. On page 1129 of
this issue, Kolkowitz et al. (1) provide us with
another tool for answering these questions:
a magnetic sensor based on atomic defects
Conductivity measurements have historically used sensors similar to an electrical
multimeter, in which electrical contacts
are placed across a device to measure its
properties. The approach is akin to many
people’s first exploratory investigations of a
new toy—pick it up and touch it. However,
like delicate toys that can be damaged by
even the most careful handling, miniaturized electrical devices are now at the point
where the measuring contacts disrupt the
material characteristics. For this reason,
“look but don’t touch” methods such as
superconducting quantum interference
devices (SQUIDs) and cold atomic gases
have been developed. Instead of measuring
resistance and conductivity via electrical
contact to the sample, the magnetic field
produced by electron motion is observed
from a distance.
One criticism of these techniques has
been their inability to measure nanometer-sized devices, because most are limited to
the micrometer or millimeter scale. Recently, single dopants in diamond were
proposed for experiments in nanoscale
magnetic sensing (2, 3). Optically active dopants often carry ground-state spin, and for
the nitrogen-vacancy (NV) color center in
particular, the spin state can be coupled to
photons and achieve ultrasensitive optical
readout of the spin levels (4, 5). As spin en-
Nitrogen-vacancy defects in
diamond can probe metallic
conductivity at a distance
By Liam P. McGuinness1,2
and Fedor Jelezko1,3