Quantum dots are not dots: physicists

Dec 21, 2010
Quantum dots are solid-state "artificial atoms" that are made up of thousands of atoms (yellow spheres) embedded in a semiconductor (blue spheres). Despite this complexity, the photon emission properties of quantum dots were hitherto believed to be like traditional atoms, where a point-emitter description is sufficient. Due to their mesoscopic dimensions, however, the point-emitter description is revealed to break down by comparing photon emission from quantum dots with opposite orientations relative to a metallic mirror.

Researchers from the Quantum Photonics Group at DTU Fotonik in collaboration with the Niels Bohr Institute, University of Copenhagen surprise the scientific world with the discovery that light emission from solid-state photon emitters, the so-called quantum dots, is fundamentally different than hitherto believed. The new insight may find important applications as a way to improve efficiency of quantum information devices. Their findings are published on December 19th 2010 in Nature Physics.

Today it is possible to fabricate and tailor highly efficient light sources that emit a single photon at a time, which constitutes the fundamental unit of light. Such emitters are referred to as quantum dots and consist of thousands of atoms. Despite the expectations reflected in this terminology, quantum dots cannot be described as point sources of light, which leads to the surprising conclusion: quantum dots are not dots!

This new insight was realized by experimentally recording photon emission from quantum dots positioned close to a metallic mirror. Point sources of light have the same properties whether or not they are flipped upside down, and this was expected to be the case for quantum dots as well. However, this fundamental symmetry was found to be violated in the experiments at DTU where a very pronounced dependence of the photon emission on the orientation of the quantum dots was observed.

The experimental findings are in excellent agreement with a new theory of light-matter interaction developed by DTU-researchers in collaboration with Anders S. Sørensen from the Niels Bohr Institute. The theory takes the spatial extent of quantum dots into account.

At the metal mirror surface, highly confined optical surface modes exist; the so-called plasmons. Plasmonics is a very active and promising research field, and the strong confinement of photons, available in plasmonics, may have applications for science or solar energy harvesting. The strong confinement of plasmons also implies that from quantum dots can be strongly altered, and that quantum dots can excite plasmons with very large probability. The present work demonstrates that the excitation of plasmons can be even more efficient than previously thought. Thus the fact that quantum dots are extended over areas much larger than atomic dimensions implies that they can interact more efficiently with plasmons.

The work may pave the way for new nanophotonic devices that exploit the spatial extent of as a novel resource. The new effect is expected to be important also in other research areas than plasmonics, including photonic crystals, cavity quantum electrodynamics, and light harvesting.

Explore further: In-situ nanoindentation study of phase transformation in magnetic shape memory alloys

More information: Article in Nature Physics: dx.doi.org/10.1038/NPHYS1870

Provided by University of Copenhagen

4.7 /5 (15 votes)

Related Stories

Coupling of Single Quantum Dots to Smooth Metal Films

Jul 20, 2009

Scientists at Argonne National Laboratory's CNM Nanophotonics Group have measured how light emission from individual colloidal semiconductor nanocrystals, or quantum dots, is modified when in proximity to ...

Encouraging quantum dots to emit photons

Aug 05, 2010

(PhysOrg.com) -- One of the fields of great interest to scientists and researchers is that of using the quantum world to enhance various aspects of our lives. Advances in quantum cryptography make headlines, and scientists ...

Single quantum dot nanowire photodetectors

Dec 14, 2010

Moving a step closer toward quantum computing, a research team in the Netherlands recently fabricated a photodetector based on a single nanowire, in which the active element is a single quantum dot with a ...

Quantum dots as midinfrared emitters

Feb 23, 2009

(PhysOrg.com) -- “People are interested in the mid-infrared,” Dan Wasserman tells PhysOrg.com. Infrared light has a wavelength longer than visible light, and many molecules have numerous very strong optical resonances in the ...

Quantum Dots Could Boost Solar Cell Efficiency

Mar 11, 2009

(PhysOrg.com) -- The transition to environmentally benign energy sources is one of the most significant challenges of the 21st century. Solar power, which uses sunlight to generate electricity, is one promising source. It ...

Recommended for you

'Exotic' material is like a switch when super thin

Apr 18, 2014

(Phys.org) —Ever-shrinking electronic devices could get down to atomic dimensions with the help of transition metal oxides, a class of materials that seems to have it all: superconductivity, magnetoresistance ...

User comments : 21

Adjust slider to filter visible comments by rank

Display comments: newest first

3 / 5 (2) Dec 21, 2010
my gut feeling tells me that there sphere of influence around their base in many cases would be as big as if you where to crush the 3d quantum dot like an empty can into a flat circle, quantum dots, due to their 3d structure add more surface area to the substrate than if they where flat and hence modulate plasmons by relative phase shifting, a plasmon traveling from one end of the substrate to the other would take a longer route over the hill than on the ground, ther interesting result from this research is that you could have several closely spatiated quantum dots communicate a a single cluster dot with optmized performance through their shared base sphere
3.3 / 5 (3) Dec 21, 2010
This opens up an avenua for making logical AND/NOT gates by topological tuning of quantum dots, the single photon could be used in binary mode as well as in quantum entanglement mode for multivalue storage, also i would be interested to know if quantum valleys/depressions would give comparable interesting effects as quantum hills/dots, maybe negative diffraction or antenna like behaviour instead of photon transmitter, somebody do quantum holes please!
3.2 / 5 (6) Dec 21, 2010
I'm a bit perplexed with quantum dot craze.

I get the application in spintronics and potentially optical/plasmonic devices, but the scale they are talking about "thousands of atoms" is really huge compared to the smallest transistors that have already been made.

The 4nm transistor that Intel has promised by 2022 ( slower than Moore's Law predicts,) has only 7 atoms in a planar configuration.

Optical transistors should be useful for future 3d architectures, but the electronic transistors we have are already strictly better than these things in pure 2d applications, (except in true "quantum" device,) re. problem of scale.

A 1000 atom "dot" is 143 times the volume of the 7 atom, 4nm transistor. A 16nm transistor would have volume under 905 atoms if it thickness equal width.

the nano-lasers I've seen are much larger than the 16nm process, and have often 2 or 3 layers of metals. To be useful, they need to be down to the scale of 2 shells of atoms around 1 atom: ~50 - 60 atoms
3.3 / 5 (3) Dec 21, 2010
Do you realize that if a quantum dot has "thousands" of atoms then it's bigger than the transistors we already use?

This means fewer would fit in the same space, and because they are farther apart...because they're bigger...you have longer signal delay, defeating one of the main purpose of using them in the first place.
not rated yet Dec 21, 2010
It's a fairly new technology. Eventually they'll get smaller, but you do have a good point. Still, I like it that we are persuing many different technologies, because one may be more appropriate in different applications. Often we don't really know where research will lead us or what opportunities will become available.
not rated yet Dec 21, 2010
maybe they should make sierpinski pyramids or other fractal structures out of it to maximize cramming a lot of surface area into a small volume
1 / 5 (1) Dec 21, 2010
No no no, photondots collaborating dotphonons. And that's not really a quantum, readable on biopsychica by google Albert Marinus. Already in 37 countries ...
3.7 / 5 (3) Dec 21, 2010
maybe they should make sierpinski pyramids or other fractal structures out of it to maximize cramming a lot of surface area into a small volume

Precisely. I suggested that the Koch Snowflake in discusing how to transition to a "psuedo 3-d" architecture a few days ago on the main forum.

Intel's 3-d transistor is a step in the right direction, as are optical transistors which will become integral, but this sort of "fractal" architecture is going to require a manufacturing technology that doesn't exist yet: nano-machines.

You will require nano-machines to build a psuedo-fractal architecture beyond maybe 2 or 3 layers of complexity. Lithography cannot even remotely do this.

You need to be able to build surfaces like tubes and cubes in 3-d, like inside and outside surface of a lego block, and then send a nano-machine down into it and build the circuitry a few atoms at a time, like an automatic plasma welder in a ship yard, except on a nano-scale...
3 / 5 (2) Dec 21, 2010
With the above concept, and having optical transistors, you can then create optical buses, allowing the sending of data across the space between the 3-d structures, instead of sending data along 2-d circuits. This allows for each "cube" to ptentially be a single processor or chipset and send and recieve it's input and output optically, but it's power through conventional electronics (until they figure out how to miniaturize optical power transmission to that scale, then you don't need most of the wires at all...)
3 / 5 (2) Dec 21, 2010
But the huge advantage is that instead of having data transmitted, in some cases, almost the full length of a motherboard, if you have optical buses and 3-d architecture, then the "ends" of the motheboard coudl bee folded over, and data transmitted optically. Instead of devices being micrometers, millimeters (processor to cache), centimeters (processor to ram), even feet (processor to peripheral port) distant to one another, they might be potentially tens or hundreds of times closer together in some cases, just through one or two folds...
5 / 5 (1) Dec 21, 2010
If you are having trouble understanding or visualizing what I was describing above, do this.

Get a piece of paper or a towel and fold it in half once. then imagine the space between the halves is as small as we can make it.

Imagine that devices that are hard to miniaturize any farther are on the outside surface, while other devices are on the inside surface. Further, imagine that whenever possible, optical buses are used to "jump" the space between halfs, rather than going around the bend/fold.

Next imagine folding the paper again along the other axis, and again imagine devices that can't be further miniaturized stay on the outside edges.

Further, imagine similar "folding" having been done the circuitry for each individual chipset, RAM stick, processor, peripheral (video/sound card,) etc, and that the layers of the folds both on the motherboard and the individual devices are as close together as possible.

This is what we envision for future optical computers.
5 / 5 (3) Dec 21, 2010
SteveL: Actually, they won't get smaller; the peculiar properties of quantum dots are directly connected to their size, so if you have a QD with some desired property (emission at some desired wavelength, say), changing its size will modify or destroy the property you wanted.

Now, that said, the large size isn't necessarily an issue; as other people have noted, the really interesting application of quantum dots to computing lies in creating a quantum computer, which would be powerful enough that the larger size of the quantum dots (relative to these new transistors being discussed; incidentally, that 7-atom transistor sounds amazing) wouldn't matter too much.
5 / 5 (1) Dec 21, 2010
The optical computer's advantages in terms of flexibility fo design and spatial savings, energy use, materials savings, etc, are incredible and still bouncing around in my head.

Look at an electronic motherboard at all the electronic buses. Often the buses take up more area than the devices they connect.

In an optical computer almost NONE of that circuitry is needed, except initial power supply circuits.

At the motherboard and "expansion card" scale, all of that is space that is completely freed for more chipsets (processors, RAM, video, sound, ports, wi-fi, etc.)

While at the individual chip scale with a bit of folding and optical circuits, the chip itself shrinks both because you're folding more and more into 3 dimensions, AND because most of the electrical circuity is no longer needed at all, except initial power supply.

Because you have almost no wires to make leaks and heat waste, you probably don't even need as much cooling either.
5 / 5 (2) Dec 21, 2010
I also didn't even get to this part, which I have mentioned in previous topics.

Unlike electrical buses and ports, optical buses and ports would be able to safely cross one another in the same plane without destroying data.

With existing architecture, power circuits and buses have to go completely the long way around one another in 2 dimensions.

With optical buses this would not happen, as two beams of light can just cross one another, no problem, even if they are both "on" at the same time.

IN electronics, your wires can run left/right, top/bottom, or diagonal on one plane only.

With optical circuits, yu have no wires, and the light can pass any combination you need from among: orthogonal in all planes, orthogonal in 2 planes diagonal in 1 plane, or even diagonal in 2 planes, like the edge of a "hip" roof or a pyramid.

This gives much more flexibility in how things "line up" when you perform the "folding" described above.
not rated yet Dec 21, 2010
you might find this interesting
not rated yet Dec 21, 2010
Is there a quantum dot which emits an electron? i wonder..!
5 / 5 (1) Dec 21, 2010
Is there a quantum dot which emits an electron? i wonder..!

Huh, yes.

There is just go down some from this article.

there is a "sensor" quantum dot that can convert photons to electrons desribed in another recent article on this site. So yes, all the components for optical circuitry exist at these scales.
not rated yet Dec 22, 2010
the point emitter foton, omly one,
Emiri how much energy should be a point particle if it emits photons in all directions? or emitted in one direction only if emitted in all directions recipients receive as much energy? and as you can measure if the photon energy received if each receptorno receive a complete package according to the quantum energy ..... because if you write a full apaquete energy in each direction may then or is emitting a single photon .. ... that would be my question
not rated yet Dec 22, 2010
point sources like to bulk up on electrons, because their relative large surface area to volume ration makes them more suspectible to Hall effect in the skin and cassimir forces acting from outside,hence you find the end of your screwdriver easily magnetized, so i would expect smaller pointsorces like nanotubes and quantum dots be tunable for electron emission, think OLED
not rated yet Dec 28, 2010
"Fractal architecture is going to require a manufacturing technology that doesn't exist yet: nano-machines."

Replicator nanotechnology (molecular assemblers) is the first thing that comes to mind.
not rated yet Dec 29, 2010
anybody can manipulate quantum field?
the quantum consciousness field to produce
particle matter without going into infinite
fractal, cause no travel medium, teleportate
without medium, disappear and reappear through
particle phasing

More news stories

UAE reports 12 new cases of MERS

Health authorities in the United Arab Emirates have announced 12 new cases of infection by the MERS coronavirus, but insisted the patients would be cured within two weeks.