14 NATURE MATERIALS | VOL 11 | JANUARY 2012 | www.nature.com/naturematerials

news & views

Fluorescence intermittency, or, more humbly, blinking, refers to the discontinuous and random emission

of light from single fluorescent sources, and is ubiquitously observed in emitting

dye molecules, polymers, biomolecules and nanoparticles1. In spite of 15 years of intense investigations, its microscopic origin for nanoparticles has eluded detailed understanding and thus continually

hampered efficient utilization of these objects. Now, in a Letter recently published in Nature, Christophe Galland and co-workers report exciting observations that demonstrate an unexpected diversity

QUANTUM DOTS

A charge for blinkingNo accepted description of luminescent blinking in quantum dots is currently available. Now, experiments probing the connection between charge and fluorescence intensity fluctuations unveil an unexpected source of blinking, significantly advancing our fundamental understanding of this baffling phenomenon.

Todd D. Krauss and Jeffrey J. Peterson

MATERIA

L WITN

ESS

The current exhibition of the artworks of Leonardo da Vinci at London’s National Gallery offers an unprecedented view of his oeuvre, including nine of the 15 paintings attributed to him. The negotiations behind the loans are said to have been hugely delicate and fraught; in general these depend on a borrower’s ability to demonstrate that it can comply with international regulations on the preservation of art materials. These typically specify, for example, that works must be displayed at temperatures no greater than about 21 °C and a humidity of no more than 50 per cent, with barely any fluctuations.

But many of these rules and guidelines were laid down decades ago, sometimes to safeguard works relocated during wartime. Today we know much more about the responses of materials to environmental conditions, and have more sensitive and versatile means of monitoring changes. Yet this new understanding does not necessarily feed into art-conservation standards, which are bound by past principles that have hardened into an accepted, almost inviolable tradition rather than anything rooted in science.

That point has been making headlines following the recent call by the director of the UK’s Tate galleries, Nicholas Serota, for these rules to

be reconsidered. Serota’s suggestion that excessive air conditioning and heating in art museums is contributing unnecessarily to global warming could easily seem faddish, given how negligible the contribution of galleries to total carbon emissions is. But on the contrary this call for a reconsideration of the rules makes a valid point that touches on the role of science in the preservation of cultural heritage.

This isn’t a new story. Some conservators have been protesting for years that normal practices are largely divorced from a modern scientific appreciation of how materials are aged and affected by their environment. Conservation tends towards conservatism — ‘this is how we’ve always done it’ — rather than seeking to pose and answer scientific questions. Given the nature of the objects under consideration, this attitude is understandable. But it can end up making art museums unnecessarily energy-hungry and dimly lit, reducing the enjoyment of visitors while raising costs in a time of diminishing funds.

Serota has been part of the debate for some time. In 2008, he and Mark Jones, former director of London’s Victoria and Albert Museum, convened a meeting of conservators to review the preservation guidelines of British museums. “Different objects have different requirements”, Jones

said. Ceramics, for example, are very tolerant, whereas metals need to be kept reasonably dry. “It is time”, Jones concluded, “for museums and funders to stop imposing standard environmental conditions” — and also for more research to be done on how materials respond.

The UK National Museum Directors’ Conference agreed, stating that conservation standards need to be made more intelligent and flexible, geared towards reducing carbon footprints and to acknowledging the value of smart building design, such as the use of passive air circulation and localized microclimate control in different spaces. These ideas are now being discussed at an international level, for example at a meeting called Rethinking the Museum Climate held in Boston in 2010. There’s still a way to go, but the conservation of art is gradually becoming more of a science. ❐

KEEPING ART ALIVE

PHILIP BALL

© 2012 Macmillan Publishers Limited. All rights reserved

NATURE MATERIALS | VOL 11 | JANUARY 2012 | www.nature.com/naturematerials 15

news & views

in the fluorescence blinking behaviour of colloidal semiconductor nanocrystals2.

During blinking, a random switching takes place between bright fluorescence periods (ON state) and dark, non-emissive ones (OFF state)3. Colloidal semiconductor nanocrystals, also known as quantum dots (QDs; ref. 4), are well suited for mechanistic studies of blinking because of their superior stability under light irradiation compared with traditional fluorescent molecules. However, QD blinking has proved incredibly difficult to study: the blinking frequency varies by over six orders of magnitude, and the photon statistics used to describe these events are insensitive to nearly all experimental variables (for example, temperature, excitation intensity and energy, QD size and chemical composition)5–6.

Quantum dot blinking was traditionally believed to be the result of a light-induced charging process7. When an electron is photoexcited it leaves behind a positively charged localized region (hole), which in a QD forms a stable electrostatic pair (or exciton) that efficiently emits light (ON state; Fig. 1a). However, multiple photon absorption forms multiple excitons, which, owing to their strong interaction energy, can ultimately lead to the ejection of either an electron or a hole yielding a charged QD. At this point, if a new exciton is formed, it will decay through fast, non-radiative charge recombination (OFF state), a mechanism known as Auger recombination (Fig. 1b). On neutralization of the charged QD, normal fluorescence cycling is restored. Because Auger recombination usually occurs at a much faster rate than radiative recombination, a decrease in emission intensity is expected to be accompanied by a decrease in the fluorescence lifetime in the charging–blinking model.

Galland et al. tested the charging–blinking theory by intentionally adding electrons to individual core–shell CdSe–CdS QDs by applying a negative external electrochemical potential2. They then recorded the variations in fluorescence dynamics this applied potential caused using a time-resolved, single QD detection system, a technique known as spectroelectrochemistry. Of particular note is that the QDs studied by the authors were designed to have a relatively bright OFF state, so that fluorescence lifetimes could be measured even when the dot was nominally OFF. Unexpectedly, they observed two distinct types of fluorescence blinking. The first type (termed A-type) is consistent with the charging–blinking model described above: electrochemical

addition of an electron to the QD — which simulates the charged state commonly thought responsible for blinking via the Auger mechanism — indeed quenches QD emission with an accompanying decreased fluorescence lifetime of the OFF state (Fig. 1b). In the second type of blinking (termed B-type), the OFF state is not associated with a decreased fluorescence lifetime; rather, the fluorescence lifetime remains constant even when the QD intensity ‘blinks OFF’.

Fluorescence intensity and lifetime are normally correlated in fluorescing optical transition. The lack of correlation, here, means that B-type blinking does not involve fluorescing energy states. To explain their results, Galland et al. hypothesize that B-type blinking arises from the activation and deactivation of non-radiative recombination centres that efficiently capture photoexcited ‘hot electrons’ (that is, electrons excited well above the lowest energy conduction band; Fig. 1c). Hot-electron capture efficiency — which competes with internal deactivation to the lowest excited energy band — can be modulated by applying an external electrochemical potential. For instance, at high negative potentials, the non-radiative recombination centres

become populated with electrons, thus rendering them ineffective in capturing additional negative charges from the core of the dot. As a result, B-type blinking is suppressed (Fig. 1d). The authors suggest that such non-radiative recombination centres reside on the surface of the dot and describe several treatments that support this hypothesis. For instance, blinking suppression is observed by increasing the distance between the core and the surface by increasing the number of layers of the semiconductor shell.

The insight that fluorescence blinking is actually a combination of two mechanisms is a landmark observation, and may help reconcile disparate models of blinking that have lately appeared in the literature to explain experimental observations that had started challenging the traditional charging–blinking model8–10. This new understanding of blinking may also lead to more advanced QD biological fluorescent labels designed to mitigate blinking entirely. However, this report also stimulates a number of important questions. For example, how general is B-type blinking among different QD materials? Are the photon statistics describing A-type and B-type sensitive to experimental variables? The nature of the OFF state in

ON

ON

Excitedstate

Externalelectrons

Hot electronHot electron

RadNon-rad

Non-radRad

Blinking suppression

Ener

gyEn

ergy

e–

h+

Rad

Ener

gyGroundstate

a

b

c d

OFF

Ener

gy

A-typeblinking

B-typeblinking Trap

state

Figure 1 | Simplified energy-level diagrams for ON and OFF states in A- and B-type blinking events. a, ON states contain a photoexcited electron (e–) and hole (h+) and emit efficiently. b, In an A-type blinking event, the OFF state contains an additional electron in an excited state and is dark due to an increased non-radiative relaxation rate. c, In a B-type event, ‘hot electrons’ are trapped by surface states immediately following photoexcitation, and combine non-radiatively with the remaining hole. d, Addition of external charges in the trap states deactivates the trapping pathway, and causes B-type blinking suppression.

© 2012 Macmillan Publishers Limited. All rights reserved

16 NATURE MATERIALS | VOL 11 | JANUARY 2012 | www.nature.com/naturematerials

news & views

Soft materials surrounded by a semipermeable membrane — such as vesicles or cells — can experience an

imbalance in osmotic pressure as a result of a difference in solute concentration between the two sides of the membrane. If the pressure imbalance is large — which can occur in hypotonic solutions — stress build-up can ultimately lead to membrane rupture. Yet such an osmotic shock is not exclusive to cells. Reporting in Nature Materials, Zavala-Rivera and colleagues describe an ingenious method for triggering multiple

osmotic shocks in a coordinated fashion within ordered block-copolymer (BCP) films to produce a well-defined pore geometry1. The researchers show that the perforated nanostructures behave as one-dimensional photonic crystals, and that they can be used as ultrafiltration membranes and as electrodes in light-emitting devices.

Linking together two or more incompatible homopolymer chains usually results in phase separation on the nanometre scale. For instance, BCPs tend to self-assemble into well-ordered

bulk morphologies — such as spheres or cylinders of one phase in a matrix of another, as well as gyroids, lamellae or honeycombs. These structures have been transformed into ordered nanoporous materials, for instance by removing the minority phase with a selective etching process2. Although only a small number of labile blocks can be removed by etching, the use of a photocleavable linker at the junction of the BCP enables the use of practically any BCP in the formation of nanoporous materials3. However, in these approaches the dimensions and geometry of the porous structures are fixed. This is because the morphology of self-assembled BCPs primarily depends on the size ratio of the polymer blocks, which cannot be easily altered by synthetic means. To achieve pores with larger sizes, the volume of the removable block needs to be expanded. This has recently been achieved in two ways: by introducing small molecules through hydrogen bonding to the block4, and by inducing swelling of the block with a selective solvent, a method that can also cause a change in morphology5. As well as the expansion of the minority component, the latter approach also involves simultaneous plasticization of the matrix phase. For example, hot-ethanol treatment of polystyrene–block–poly(2-vinyl pyridine) (PS–b–P2VP) not only swells the P2VP phase, but also plasticizes the PS matrix, thereby enabling stress dissipation during the expansion process6. Another example is the use of supercritical CO2 — which selectively swells fluorinated domains of BCPs — to induce morphological changes

B-type events is also intriguing. For the fluorescence lifetime to remain constant and the emission intensity to decrease, the decrease in radiative recombination rate must be exactly balanced by the increase in the non-radiative recombination rate to the surface trap states. This suggests these processes are not independent and thus may have important implications for a more general understanding of QD energy states. Guided by the report of

Galland and colleagues, future efforts will undoubtedly make this area of research highly charged indeed. ❐

Todd D. Krauss is in the Department of Chemistry and the Institute of Optics, University of Rochester, Rochester, New York 14627, USA. Jeffrey J. Peterson is in the Department of Chemistry, State University of New York at Geneseo, New York 14454, USA. e-mail: krauss@chem.rochester.edu; petersonj@geneseo.edu

References1. Moerner, W. E. & Orrit, M. Science 283, 1670–1676 (1999).2. Galland, C. et al. Nature 479, 203–207 (2011).3. Nirmal, M. et al. Nature 383, 802–804 (1996).4. Alivisatos, A. P. Science 271, 933–937 (1996).5. Kuno, M., Fromm, D. P., Hamann, H. F., Gallagher, A. &

Nesbitt, D. J. J. Chem. Phys. 115, 1028–1040 (2001).6. Frantsuzov, P., Kuno, M., Jankó, B. & Marcus, R. A. Nature Phys.

4, 519–522 (2008).7. Efros, A. L. & Rosen, M. Phys. Rev. Lett 78, 1110–1113 (1997).8. Zhao, J., Nair, G., Fisher, B. R. & Bawendi, M. G. Phys. Rev. Lett.

104, 157403 (2010).9. Jha, P. P. & Guyot-Sionnest, P. ACS Nano 3, 1011–1015 (2009).10. Tang, J. & Marcus, R. A. J. Chem. Phys. 123, 1–12 (2005).

Figure 1 | Perforated multilayers with look-alike structures. a, Cross-sectional SEM image of a PS-b-PMMA film ruptured as a result of coordinated osmotic shocks. Reproduced from ref. 1. b, A layered rock formation on Jeju Island, Korea.

a b

200 nm 10 cm

P. T

HEA

TO, U

NIV

. HA

MBU

RG /

SN

U

NANOPOROUS ORDERED MATERIALS

Osmotically shockedCoordinated osmotic shocks within ordered materials lead to nanoperforated multilayer structures that may find application in photonics, optoelectronics and ultrafiltration.

Patrick Theato and Goran Ungar

© 2012 Macmillan Publishers Limited. All rights reserved