Diffraction-unlimited optical microscopy

Diffraction-unlimited optical microscopy Optical microscopy, and fluorescence microscopy in particular, has emerged as one of the most powerful and convenient microscopic tools available today. This power does come at a price, however, in terms of a limited spatial resolution: traditionally fluorescence microscopy has been limited by diffraction to a resolution of a few hundred nanometers, far too large to discern nanostructuring in biological or material samples. Recent conceptual advances have emerged that challenge this once-thought ‘unbreakable’ barrier, and fluorescence microscopy with nanometer resolution is now within reach. In this review we highlight some of the approaches that have made this paradigm shift possible. Peter Dedecker, Johan Hofkens, and Jun-ichi Hotta Department of Chemistry, Katholieke Universiteit Leuven, Celestijnenlaan 200F, B-3001 Heverlee, Belgium E-mail: [email protected]

One of the great paradoxes of life is that in order to truly

strengths and weaknesses, but one of the most convenient is optical

understand the everyday issues and concepts that affect and

microscopy, which, in an ideal case, requires only the illumination of

fascinate us, one invariably has to focus on the microscopic world

the sample. Because of this, optical microscopy tends to be relatively

around us. This idea, while no longer particularly surprising,

noninvasive, nondestructive, and convenient to use. Fluorescence

was realized as soon as the first lenses and microscopes were

microscopy in particular has proven to be one of the most powerful

developed, and has only become stronger as our ability to observe

microscopy tools available today, by combining the advantages of

smaller and smaller details improved. While this idea is usually

optical microscopy with the selectivity and extreme sensitivity of

brought up in the context of the life sciences, it is no different for

fluorescence emission. Indeed, fluorescence microscopy allows for

the materials of today and tomorrow. Designing new materials

measurements down to the level of individual molecules1–3, and

tailored to specific needs and properties requires both study

has been successfully applied in the study of, for example, polymer

and manipulation of their micro- and nanoscale structuring, and

dynamics4–9, catalysts10–15, dendrimers16.

microscopes are invaluable tools in modern-day research. Not all approaches to visualize small details are equally useful for

12

Nevertheless, revealing nanoscale structuring and ordering remains one of the great challenges of science. Indeed, one aspect that is true

a given sample or problem, however, and a wide range of microscopy

for every microscopy technique is that there is a lower limit to the

techniques have been developed. Each of these has its share of

sizes of the details that can discerned, that is, the spatial resolution is

MICROSCOPY SPECIAL ISSUE

ISSN:1369 7021 © Elsevier Ltd 2008

Diffraction-unlimited optical microscopy

REVIEW

Fig. 1 Point spreading: after imaging, an object of negligible size (such as a single molecule) will show up as a distribution of much bigger dimensions. The centroids of these distributions are indicated by the green and blue markers. If the emitters are spaced close together, as is the case for the molecules in the blue circle, then they cannot be distinguished and the structural information is lost. (Adapted from 83. © Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.)

limited. Intuitively, if the sample structure becomes too small, then

microscopes and objectives have advanced to the point where the level

it cannot be revealed. There are several possible reasons why this is

of aberrations is so low that their contribution is negligible. Therefore

so, ranging from technical aspects, such as instrument imperfections,

little gain in resolution can be expected from further improvements

vibrations, and measurement noise, to limitations imposed by the

in the quality of the microscope optics. While it is possible to modify

fundamental laws of nature. While the former limitations are, to some

the optical system in such a way that the point spreading is strongly

extent, specific to a particular instrument or experiment, the latter

reduced, e.g. by including a second objective20,21, these approaches

present limiting barriers that are hard to overcome.

tend to be limited by the fact that the instrument design and

In the case of optical microscopy, this fundamental limitation

alignment become increasingly complex.

stems directly from the wave-like character of light: no matter what

It thus appears that the spatial resolution in optical microscopy

kind of optical components we use, we cannot focus the light waves

is limited to a few hundred nanometers, and that details at smaller

to an infinitely small point, only to a finite region. This might seem

distances are fundamentally out of reach. Fortunately it turns out that

surprising at first: when we collect the fluorescence emission from a

it is possible to obtain an imaging resolution that is fundamentally

single, isolated molecule and image it through a lens system (such

unlimited by diffraction, and a variety of techniques that aim to

as a microscope), the resulting image will not be a point but rather

achieve this have recently appeared. Intuitively these techniques

a three-dimensional shape of much larger dimensions (hundreds of

work by making fluorescence emission into a rare event: if neither the

nanometers; see Fig. 1). The same holds true for the excitation light:

excitation illumination nor the fluorescence emission can be confined

even if the light source is infinitely small, we can only focus the light to

further, then we have to make sure that only a small fraction of the

a diffraction-limited region and cannot confine it further. This is known

molecules emits fluorescence even though all of them may experience

as point spreading, and is an immutable law of nature.

a similar intensity excitation. Alternatively, if we process the images

Point spreading is problematical if we want to image a sample

mathematically, then we can invert the problem, and try to minimize

that contains fluorescent molecules in close proximity, such as in a

the number of molecules that are not emitting beyond what is

finely structured sample. In this case the emission distributions from

normally allowed by diffraction. It turns out that the answer to this

closely spaced fluorophores overlap in the resulting fluorescence

problem lies in the properties of the ﬂuorescent molecules themselves,

image, making it impossible to distinguish between molecules that are

and in the fact that ﬂuorescence is a dynamic phenomenon22–24.

close together (Fig. 1). Hence point spreading imposes a fundamental limitation on the spatial resolution that can be attained. (We will not

Nonlinearity

go into further detail here; more extended, yet accessible, discussions

When we think about fluorescence emission, we tend to think about

have been published

elsewhere17–19.)

While imperfections in the imaging system worsen the point spreading beyond the theoretical minimum, the optics in today’s

it as a fairly well-behaved, constant phenomenon. If we take a fluorescent molecule and irradiate it with excitation light, we expect to see a continuous, and predictable, fluorescence emission. Moreover,


13

REVIEW


Fig. 2 The emission of a single molecule of a fluorescent protein, clearly showing complex fluorescence dynamics, and the resulting nonlinearity in the fluorescence emission. The inset shows an expansion of the emission trace.

if we double the excitation intensity then we expect to see a twofold

makes use of this fact in some way. It is possible to roughly divide

increase in fluorescence emission, and so on. This seems intuitive,

these approaches based on whether they make use of nonlinearities

because it agrees with the observation of emission from a dye-loaded

in space (that is, try to exploit the nonlinearity to increase the spatial

solution in a fluorescence cuvette.

confinement of the fluorescence emission) or time (that is, try to

This apparent proportionality, however, only holds at the ensemble

separate the emission of the different labels in time). In what follows

(averaged) level. One of the most fascinating findings from single-

we will discuss each of these approaches, and mention some possible

molecule experiments is that fluorescence emission is in fact a very

advantages and disadvantages of each.

dynamic phenomenon: individual molecules are seen to rapidly cycle between emissive and nonemissive states, displaying a highly

Nonlinearity in space

complex and fascinating dynamic behavior25–29 (Fig. 2). The origin

For a more detailed example of how nonlinearities can be put to

of this on–off ‘blinking’ behavior is not always clear, though it can

work, let us look at the process of fluorescence emission itself. The

sometimes be attributed to, for example, the formation of triplet

initial absorption of an excitation photon causes the molecule to

states, photoisomerization, or electron transfer. It is also known to

enter an electronically excited state, where it remains for (usually)

be virtually ubiquitous, being observed in wide variety of systems,

a few nanoseconds, before returning to the ground state through

including organic molecules25–27,29, semiconductor nanocrystals28, and

the emission of a fluorescence photon. The central issue here is

fluorescent proteins30.

that there is always some nonzero delay between the absorption

Apart from transient nonfluorescent or ‘dark state’ formation,

increase up to the point where the molecule is constantly in the

well, including saturation (the fact that there is an upper limit for the

excited state, when it reaches a plateau. At this point, increasing the

emission rate due to the finite excited-state lifetime), and phenomena

excitation intensity does not increase the emission rate any further,

such as photoswitching or

photoactivation31–41.

The net result is that the fluorescence emission is no longer strictly proportional to the excitation intensity. For a given sample and

and the fluorescence emission is said to be saturated (and strongly nonlinear). Fig. 3 shows an example of this. In this figure a cross-section of the

experimental setup the emission can depend on time, on the excitation

fluorescence emission rate along the focus of a confocal fluorescence

intensity or intensities, or, by applying a spatial intensity distribution,

microscope is shown, at different excitation powers. The shape of the

on the position of the molecule with respect to the microscope’s focus.

diffraction-limited excitation intensity distribution is shown by the red

Every approach to achieve a diffraction-unlimited resolution

14

and emission events. In other words, the emission rate can only

interruptions in the fluorescence emission can arise for other reasons as


curve in the figure, and does not change throughout the experiment.


REVIEW

Fig. 3 The effect of saturation: he relative emission rate at the focus of a confocal fluorescence microscope and different excitation powers. The effect of saturation of the fluorescence emission is clearly visible. The relative emission powers, from red to purple, are 1, 5, 20, 100, and 1000, respectively.

However, as the excitation power is increased, the saturation of the

input laser beam. Especially interesting are ‘donut modes’, which are

emission drastically alters the resulting emission distribution.

characterized by a sharp intensity zero at the maximum, surrounded

The most crucial aspect of Fig. 3 is that it demonstrates that saturation

by a region of intense irradiation42–44 (Fig. 4). Only at the exact center

enables us to create emission distributions with sharp edges and

of the microscope focus is the intensity truly – mathematically – zero,

strongly confined nonemissive regions, even though these were not

while there is some irradiation intensity everywhere else in the focal

allowed in linear systems by diffraction.

volume.

This simple example shows how using fluorescence nonlinearities can allow us to bypass the limitations of diffraction. Saturation of the emission rate is not the only possibility, however,

Let us now assume that we have shaped the output from a laser into a donut mode, and that the wavelength and duration of the light is chosen so that it can deactivate the dye molecules. This

as any transition to a state with different emissive properties can

deactivation could be because the light induces the formation of a

be used as long as it is saturable (that is, its light-induced formation

nonfluorescent state (e.g. triplet or charge-transfer states, fluorescence

is much faster than its recovery) [22]. Approaches that make use of

photoswitching), or because it induces stimulated emission. Moreover,

this principle to create ‘sharp’ distributions in emission are known as

due to the intensity distribution of the donut shape, this deactivation

RESOLFT microscopy, standing for ‘reversible, saturable optically linear

occurs everywhere in the focal volume except at the center of the

fluorescence transitions’ [23].

focus. This approach becomes even more interesting when it is

By itself, the example shown in Fig. 3 is not directly useful

combined with saturation (Fig. 5). In this case, it is clear that the

for attaining a higher imaging resolution, since the fluorescence

confinement of the molecules in the fluorescent state can be made

confinement is not improved but, rather, worsened compared to the

arbitrarily small, limited only by the power and duration of the donut-

‘standard’ diffraction-limited case. Hence we must either modify the

mode beam.

imaging to use the saturation in a different way or process the images

At present the most successful technique that uses this approach

mathematically. A number of techniques have been developed that

is ‘stimulated-emission depletion’ (STED) microscopy45–50. The

attempt both approaches, and we will expand on these in what follows.

principle underlying this approach is shown in Fig. 6, and operates as follows: in the first step of the acquisition, the sample molecules in

STED microscopy

the diffraction-limited spot are excited by a normal mode (Airy disk)

While the focusing of the excitation light will always lead to some

laser pulse with a duration of picoseconds or less. This is followed

spatial distribution, the exact shape of this distribution can be modified

immediately (within a few hundred picoseconds) by a second donut-

into different ‘modes’ by manipulating the relative phases of the

mode laser pulse irradiating the same region, which induces stimulated


15

REVIEW


Fig. 4 Different intensity distributions along the focal plane of a confocal microscope: a ‘standard’ Airy disk shape and a ‘donut mode’.

emission in the majority of the dye molecules, except those at the

state can be used, including triplet-state formation51 and fluorescence

center of the focus (due to the intensity distribution of the donut

photoswitching52–55. the sequence of the pulses likely has to be

mode).

interchanged, and an additional laser pulse may be necessary.

The sample molecules that are still in the excited state are then

The most attractive features of donut-based microscopy are

left to fluoresce normally, which is detected. Because of the depletion,

that the effective resolution increase is completely dictated by the

however, the fluorescence collection is effectively confined to a spot

experimental setup and the irradiation powers used. Moreover, the

with sub-diffraction dimensions (Fig. 5b), leading to an increased

imaging is altered directly, and obtaining a subdiffraction resolution

resolution. This measurement cycle is then repeated at the next point

does not require any additional processing. In addition, if stimulated

on the sample, until a complete image is obtained. In this way, a spatial

emission is used, then the image acquisition times can be as fast as any

resolution down to a few tens of nanometers has been

reported48,49.

As mentioned above, any saturable transition to a nonfluorescent

laser-scanning microscope56. As indicated, the effective resolution increase that can be obtained

Fig. 5 The effect of saturation using a donut-mode beam. (a) By saturating the transition to a nonfluorescent state, the spot of molecules that remain in the bright state can be made arbitrarily small (the relative excitation powers are 1, 5, 20, 100, and 1000, respectively). (b) The resulting apparent spot size.

16



REVIEW

Fig. 6 A diagram of an experimental setup for STED microscopy, showing both the activation and depletion pulse, and the resulting reduction of the fluorescent spot size. (Reproduced with permission from 18.)

depends on the number of irradiation photons in the dump beam:

structure. One way to do this is by nonlinear structured illumination

more irradiation leads to a greater depletion and tighter confinement,

microscopy60.

and hence an improvement in resolution. While this is a conceptual

Structured illumination relies on projecting the excitation light

advantage, as the diffraction-limit is effectively broken, it also means

as a standing wave along the focal plane of the microscope, with

that rather high levels of irradiation are required, especially when

a sufficiently high intensity to saturate the emission (though other

the depletion has to occur within a very short time window (as is

saturable transitions can also be used). A cross-section of the

the case when using stimulated emission). This can be mitigated

fluorescence emission along the standing wave pattern looks similar

somewhat by making use of longer-lived nonfluorescent

states51–55,

to Fig. 5a, with ‘fluorescence deactivation’ replaced by ‘fluorescence

but photobleaching or photodestruction is likely to remain a limiting

emission’. By recording a set of images of the same sample, but with

factor in techniques that rely on saturation to create sharp emission

shifted orientations of the pattern, a diffraction-unlimited picture

distributions. Nevertheless, a wide range of fluorescent dyes, including

can be reconstructed (Fig. 7). Moreover, it is possible to to extend

fluorescent

proteins57,58

and ‘atto’

dyes58,

have been found to be stable

the underlying principle to three dimensions, as has recently been

enough to produce impressive results49,59.

demonstrated61,62.

Nonlinear structured illumination

microscopy in that the measurement time is theoretically independent

As already discussed, saturation of the emission rate can be used

of the labeling density and the photostability of the sample determines

directly to achieve a diffraction-unlimited resolution, though it

the performance. They differ, however, in that STED is based on

requires processing of the raw images to discern the underlying sample

confocal microscopy instead of widefield microscopy, and that image

Nonlinear structured illumination is fairly similar to STED


17

REVIEW


Fig. 7 Examples of STED and nonlinear structured illumination. (Left) A comparison between a standard confocal image and a STED image of 20 nm fluorescent beads. (Right) A widefield image of 50 nm fluorescent beads. (a, b) Conventional and filtered image; (c) linear structured illumination; (d) nonlinear structured illumination. (Structured illumination data reproduced with permission from60. © National Academy of Sciences, USA.)

processing is essential for structured illumination but not required for

make use of steady-state saturation: the high- resolution information is

STED.

only obtained after the system has reached saturated conditions. While

The exact mechanism by which the resolution is increased, and

this makes for an intuitively clear picture, it is not essential. We will

the required imaging processing is usually expressed in terms of

not consider ‘dynamic saturation optical microscopy’ in detail here, but

Fourier components and spatial frequencies. A detailed discussion

two possible approaches have been discussed elsewhere63,64.

of this is beyond the scope of this paper, but has been published

Nonlinearity in time

elsewhere60. In a sense, the two approaches that we have introduced here both

The problem of limited spatial resolution can be approached in a

Fig. 8 The principle behind PALM imaging: by repeatedly activating, localizing, and bleaching, individual emitters a high-resolution picture of the sample can be reconstructed. (a) A simulated conventional image of the sample. (b) The PALM imaging. (c) The reconstructed picture.

18



REVIEW

different way: let us assume that we somehow know that only a single emitting molecule is present in a given diffraction-limited spot. We can easily pinpoint the exact position of this molecule with nanometer precision, simply by looking at the centroid of the emission distribution65,66 (indicated by the markers in Fig. 1). If the density of the labeling molecules is sufficiently low, then this approximation is likely valid, though in that case it would be impossible to obtain any detailed structural information on the sample. If we are interested in subdiffraction microscopy, then our sample should contain many fluorophores within nanometers of each other. Hence, the only way to apply this method would be to separate the emission of the molecules in time, so that at any given moment only a single molecule within a given diffraction-limited spot would emit. Then, by locating and mapping out the positions of all of the molecules in time, a high-resolution image can be constructed, effectively bypassing the diffraction limit. Such a separation would be possible if one could precisely control the distribution of the molecules between the fluorescent and nonfluorescent states over an extended period. Until recently

Fig. 9 An example of localization microscopy. (a) A conventional image of clathrin-coated pits in a fixed cell; (b) a STORM/PALM image of the same region. (Reprinted with permission from 71. © AAAS.)

this task would have been considered virtually impossible, especially for large numbers of dye molecules. Now, however, a numbe rof

a new set of molecules is activated, and the cycle is repeated until

new fluorescent dyes have become available that display highly

the position of (nearly) all molecules has been determined. Because

efficient fluorescence photoswitching or photoactivation31–41.

the activation of the fluorescence is a stochastic process, different

Photoactivation means that the molecule starts out in a thermally

molecules are activated in each cycle, and the probability that two

stable nonfluorescent state but can be converted to a fluorescent

activated molecules will be present within a given diffraction-limited

state through irradiation with light of an appropriate wavelength,

spot can be neglected. Hence, by mapping out all the localized

where it remains until photobleaching. Photoswitching, on the other

positions, a high-resolution image of the sample can be reconstructed

hand, means that the molecule can exist in both a thermally stable

(Fig. 9).

fluorescent state and a nonfluorescent state, and can reversibly

PALM is a fascinating technique in that it does not attempt to

interconvert between these states through irradiation with light

modify the optical capabilities of the instrument in any way. Instead,

of an appropriate wavelength. By using these properties, precise

it sidesteps the issue completely by ensuring that situations where

control over emission of individual fluorescent labels is now possible.

there are two emitters within the same diffraction-limited spot are

However, other processes can be used as well, such as diffusion and

avoided. By this technique, localization precision to within a few tens

docking67.

of nanometers has been demonstrated68,71. The advantages of this

Several related techniques have appeared that make use of

technique are that it is relatively simple, as no special equipment

photoswitching or photoactivation in combination with localization,

is required, and fairly gentle, in the sense that only low irradiation

and are commonly known as ‘photoactivation localization microscopy’

intensities are required. Also, as some of the most promising

(PALM)68,

photoswitchable and photoactivable fluorescent markers are genetically

‘fluorescence photoactivation localization microscopy’

(FPALM)69, or ‘stochastic optical reconstruction microscopy’

encodable fluorescent proteins, this approach is especially useful for

(STORM)70. For convenience, we will refer to these techniques

biological samples.

as PALM microscopy. Briefly, the principle behind PALM, shown

However, the technique is not without its disadvantages: at present,

in Fig. 8, is as follows: a given sample is densely labeled with

the resolution along the optical axis is still fairly limited, and the

photoactivatable or photoswitchable dyes, which are initially in the

requirement for a large number of activation–localization–bleaching

nonfluorescent state. The sample is then weakly irradiated with a

cycles means that a fairly long measurement time (several minutes

small amount of activation light, so that only a very small fraction

or more) is required. Nevertheless, attempts to extend the imaging

of the molecules are activated. These activated molecules are then

to three dimensions71,72 and to use different spectrally shifted

imaged until they deactivate again, either through photoswitching or

labels simultaneously73,74, as well as to increase the speed of the

photobleaching, and their locations are precisely determined. Next

acquisition75–77, are currently underway.


19

REVIEW


Outlook and conclusions It does not happen often that a barrier that has stood for over a hundred years is suddenly broken, but that is precisely what is happening today with fluorescence microscopy. Over the course of just a few years, several techniques have appeared that tackle or avoid this limitation in different ways, thereby allowing fluorescence microscopy to cross a familiar frontier into a previously inaccessible world. While the field of diffraction-unlimited fluorescence microscopy is still very young, it has evolved and continues to evolve at a very rapid pace. The validity of the different approaches presented here has been demonstrated, and the experimental focus has started to shift away from the ‘proof-of-concept’ stage to that of a valid and useful research tool. That is not to say, however, that there is a lack of challenges for subdiffraction imaging. For example, until recently, most experiments focused entirely on the imaging resolution along the focal plane of the microscope, using very thin samples, while effectively neglecting the resolution along the optical axis for the most part. By itself, this is not unreasonable, especially considering the recent nature of these approaches. However, in many experiments the ability to image ‘thick’ samples or to construct optical sections is essential. While considerable progress is being made in this direction71,72,78,79, no conclusive approach has emerged so far. A second complication is the time resolution of the acquisition,

This discussion leads to another issue. While diffraction-unlimited

though this is currently mostly a limitation for the PALM-based

fluorescence microscopy has, to some extent, been made possible

approaches. As we have tried to highlight in the paper, the approaches

through technical advances, a perhaps even more crucial aspect has

that aim to exploit nonlinearities to increase the confinement

been the fluorescent label itself. As we have emphasized here, all of the

(RESOLFT, nonlinear structured illumination) have the advantage that

techniques available today rely on special properties of the fluorescent

the duration to acquire an image is always the same, irrespective of the

label. This ranges from exceptional photostability to the presence

labeling density of the sample, and that it is possible to collect fewer

of long-lived nonfluorescent states, or ‘special’ properties such as

photons per molecule, as they do not need to be individually localized.

photoactivation or photoswitching. These processes are not yet fully

In contrast, in PALM-based microscopy, the rate at which molecules

understood, and the dyes that are available today all have their own

can be localized is essentially fixed, and is limited by the requirement

drawbacks, directly limiting the imaging. As such, there is a strong

that each activated molecule should emit enough photons to be

requirement for new dyes with improved properties80.

accurately localized. As such, it remains to be seen whether localization

As for the applicability of these techniques to practical

approaches will prove capable of resolving fast dynamics (minutes or

problems, in the experiments demonstrated so far the focus has

less) on meaningful samples.

mostly been on biological samples, and comparatively little has

But what exactly is a ‘meaningful sample’? To answer this question

been done with respect to the material sciences81,82 (Fig. 10). In

we have to look in closer detail at the imaging itself. Fluorescence

principle, however, these techniques can be applied to any sample

microscopy is interesting in that it never observes the sample structure

that can be studied by conventional fluorescence microscopy. One

directly, but observes the labeling molecules that have been introduced

possible limitation here might be the requirement for special probe

artificially. The ‘structure’ that is observed thus depends on two factors:

molecules, though, with careful experimentation and the knowledge

(i) which part of the sample has been labeled; and (ii) the density of

and labels available today, we feel that it should be possible to

the labeling. Intuitively: if we have an optical resolution of 50 nm but

obtain a significant improvement in resolution for most kinds of

the average distance between the labels is 200 nm, then the resulting

samples.

fluorescence image contains no meaningful information on features

20

Fig. 10 Applications in material science: a composite image showing the three-dimensional colloidal packing of 200 nm fluorescent latex spheres, acquired using STED microscopy. Both well organized areas and defects are clearly visible. (Reprinted with permission from 82. © 2007 American Chemical Society.)

In particular, research areas where these techniques are set

smaller than 200 nm. This is an aspect that has become very relevant

to have a significant impact include nanometer-sized particles or

with the development of high-resolution imaging.

structures such as plasmonic particles, phase transitions in colloidal



REVIEW

systems, single active-site catalysis, laser-induced lithography and

the techniques that we have presented here have their own strengths

micropolymerization. Similar possibilities exist in polymer dynamics,

and weaknesses, and hence will be more or less suitable for a particular

where the material properties depend strongly on the conformation

experiment. At the moment it is impossible to say which, if any, of

of individual polymer chains, which is hard to obtain directly using

these techniques will eventually turn out to be superior, or if the way

diffraction-limited techniques. With time it should be possible to

forward will be to use them in a complimentary fashion. Perhaps an as-

extend these techniques to most fields where nanoscopic particles or

yet undiscovered approach will take over – it is too early to tell. What

features are important.

can be said, however, is that subdiffraction fluorescence microscopy

There is no doubt that diffraction-unlimited fluorescence microscopy is here to stay, and will continue to evolve rapidly. All of

has opened up an entirely new playing field, and that many untold possibilities await.

REFERENCES

42. Klar, T. A., et al., Phys. Rev. E (2001) 64, 066613

1. Zander, C., et al. (eds.) Single-Molecule Detection in Solution. Wiley-VCH (2002)

43. Török, P., and Munro, P. R. T., Opt. Express. (2004) 12, 3605

2. Kulzer, F., and Orrit, M., Annu. Rev. Phys. Chem. (2004) 55, 585

44. Dedecker, P., et al., Opt. Express (2007) 15, 3372

3. Tinnefeld, P., and Sauer, M., Angew. Chem. Int. Ed. Engl. (2005) 44, 2642

45. Hell, S. W., and Wichmann, J., Opt. Lett. (1994) 19, 780

4. Barbara, P. F., et al., Acc. Chem. Res. (2005) 38, 602

46. Klar, T. A., and Hell, S. W., Opt. Lett. (1999) 24, 954

5. Becker, K., and Lupton, J. M., J. Am. Chem. Soc. (2006) 128, 6468

47. Klar, T. A., et al., Proc. Natl. Acad. Sci. USA (2000) 97, 8206

6. Uji-I, H., et al., Polymer (2006) 47, 2511

48. Westphal, V., et al., Appl. Phys. B (2003) 77, 377

7. Zondervan, R., et al., Proc. Natl. Acad. Sci. USA (2007) 104, 12628

49. Donnert, G., et al., Proc. Natl. Acad. Sci. USA (2006) 103, 11440

8. Zondervan, R., et al., Proc. Natl. Acad. Sci. USA (2008) 105, 4993

50. Auksorius, E., et al., Opt. Lett. (2008) 33, 113

9. Wöll, D., et al., Angew. Chem. Int. Ed. (2008) 47, 783

51. Bretschneider, S., et al., Phys. Rev. Lett. (2007) 98, 218103

10. Roeffaers, M. B. J., et al., Nature (2006) 439, 572

52. Hofmann, M., et al., Proc. Natl. Acad. Sci. USA (2005) 102, 17565

11. Riuz, A. Z., et al., Angew. Chem. Int. Ed., (2006) 45, 5282

53. Bossi, M., et al., New J. Phys. (2006) 8, 275

12. Hashimoto, S., et al., Micropor. Mesopor. Mat. (2007) 101, 10–18

54. Schwentker, M. A., et al., Microsc. Res. Tech. (2007) 70, 269

13. Roeffaers, M. B. J., et al., Angew. Chem. Int. Ed. (2007) 46, 1706

55. Dedecker, P., et al., J. Am. Chem. Soc. (2007) 129, 16132

14. Cox, M. H. F., et al., Chem. Eur. J. (2008) 14, 1718

56. Westphal, V., et al., Science (2008) 320, 246

15. De Cremer, G., et al., Angew. Chem. Int. Ed. (2008) 47, 2813

57. Willig, K. I., et al., Nat. Methods (2006) 3, 721

16. De Schryver, F. C., et al., Acc. Chem. Res. (2005) 38, 514

58. Rittweger, E., et al., Chem. Phys. Lett. (2007) 442, 483

17. Stelzer, E. H. K., J. Microsc.( Oxford) (1998) 189, 15–24

59. Willig, K. I., et al., Nature (2006) 440, 935

18. Hell, S. W., Nat. Biotechnol. (2003) 21, 1347

60. Gustafsson, M. G. L., Proc. Natl. Acad. Sci. USA (2005) 102, 13081

19. Heintzmann, R., and Ficz, G., Brief. Funct. Genomic. Proteomic. (2006) 5, 289

61. Gustafsson, M. G. L., et al., Biophys. J. (2008) 94, 4957

20. Hell, S. W., and Stelzer, E. H. K., Opt. Commun. (1992) 93, 277

62. Schermelleh, L., et al., Science (2008) 320, 1332

21. Gustafsson, M. G. L., et al., J . Microsc. (Oxford) (1999) 195, 10

63. Enderlein, J., Appl. Phys. Lett. (2005) 87, 094105

22. Hell, S. W., et al., Appl. Phys. A (2003) 77, 859

64. Fujita, K., et al., Phys. Rev. Lett. (2007) 99, 228105

23. Hell, S. W., Science (2007) 316, 1153

65. Gelles, J., et al., Nature (1988) 331, 450

24. Moerner, W. E., Proc. Natl. Acad. Sci. USA (2007) 104, 12596

66. van Oijen, A. M., et al., Chem. Phys. Lett. (1998) 292, 183

25. Kohn, F., et al., J. Phys. Chem. A (2002) 106, 4808

67. Sharonov, A., et al., Natl. Acad. Sci. USA (2006) 103, 18911

26. Yeow, E. K. L., et al., J. Phys. Chem. A (2006) 110, 1726

68. Betzig, E., et al., Science (2006) 313, 1642

27. Hoogenboom, J. P., et al., ChemPhysChem (2007) 8, 823

69. Hess, S. T., et al., Biophys. J. (2007) 91, 4258

28. Cichos, F., et al., Curr. Opin. Colloid Interface Sci. (2007), 12, 272

70. Rust, M. J. R., et al., Nat. Methods (2006) 3, 793

29. Clifford, J., et al., J. Phys. Chem. B (2007) 111, 6987

71. Huang, B., et al., Science (2008) 319, 810

30. Dickson, R. M., et al., Nature (1997) 388, 355

72. Fölling, J., et al., Angew. Chem. Int. Ed. (2007) 46, 6266

31. Cinneli, R. A. G., et al., Appl. Phys. Lett. (2001) 79, 3353

73. Bock, H., et al., Appl. Phys. B (2007) 88, 161

32. Patterson, G. H., and Lippincott-Schwartz, J., Science (2002) 297, 1873

74. Bates, M., et al., Science (2007) 317, 1749

33. Ando, R., et al., Proc. Natl. Acad. Sci. USA (2002) 99, 12651

75. Geisler, C., et al., Appl. Phys. A (2007) 88, 223

34. Ando, R., et al., Science (2004) 306, 1370

76. Flors, C., et al., J. Am. Chem. Soc. (2007) 129, 13970

35. Fukaminato, T., et al., J. Am. Chem. Soc. (2004) 126, 14843

77. Shroff, H., et al., Nat. Methods (2008) 5, 417

36. Habuchi, S., et al., Proc. Natl. Acad. Sci. USA (2005) 102, 9511

78. Schmidt, R., et al., Nat. Methods (2008) 5, 539

37. Lukyanov, K. A., et al., Nat. Rev. Mol. Cell. Biol. (2005) 6, 885

79. Juette, M. F., et al., Nat. Methods (2008) 5, 527

38. Heilemann, M., et al., J. Am. Chem. Soc. (2005) 127, 3801

80. Shaner, N. C., et al., Nat. Methods (2008) 5, 545

39. Bates, M., et al., Phys. Rev. Lett. (2005) 94, 108101

81. Westphal, V., et al., J. Phys. B, At. Mol. Opt. Phys. (2005) 38, S695

40. Bossi, M., et al., Angew. Chem. Int. Ed. (2006) 45, 7462

82. Harke, B., et al., Nano Lett. (2008) 8, 1309

41. Fukaminato, T., et al., J. Am. Chem. Soc. (2007) 129, 5932

83. Dedecker, P., et al., Angew. Chem. Int. Ed. (2007) 46, 8330


21