Talk:Coulomb blockade

SET island a quantum dot?

I do not think we should call the island of the single electron transistor a quantum dot, because the use of the term "quantum dot" implies that the structure is very small on the scale of relevant length scales (for example, the de Broglie wavelength of the electron) in all three spatial dimensions. The island of the single electron transistor does not have to be small a priori (although having a small island is of course helpful for keeping the capacitance low, making Coulomb blockade easier to observe). If we decide to keep the quantum dot link, we will also have to rewrite the article about quantum dots, since it currently does not cover mesoscopic metallic structures. --DrTorstenHenning 12:00, 14 February 2006 (UTC)

You are completely right. I rephrased it, including a wikilink for self-capacitance (the capacitance you are referring to, which is the key parameter in the SET). I also removed the nonsense about non-discretely "charging" the island with the gate electrode. The gate electrode just shifts the potential of the island. --Dschwen 10:26, 3 March 2006 (UTC)

Images and more

I'm changing some pictures. The TyNiemeyerDolanTechnique.png picture is covering only a related topic. The Niemeyer-Dolan technique article should be written and the picture moved there (caption availible in the wikisource). The TySETcharacteristics.png picture is too complicated for any reader who wants introductory info about SETs. Let's leave it here for now and reinsert it if a) a descriptive caption for the pic is found, and b) the technical level of the article has advanced to match the level of the picture.

I agree that the Niemeyer-Dolan technique should have an article of its own (I just don't have the time to start it right now). But wouldn't the right time for moving the corresponding image be after the creation of that article? I don't get the reasoning behind removing that image at this time. The fabrication is not "only a related topic" as long as there is no article on fabrication. As for the image that appears "too complicated": I plainly disagree. Do you want less information, or the same information diluted over several figures, or would you prefer to have a cartoon instead of a diagram based on real data? --DrTorstenHenning 12:31, 5 March 2006 (UTC)

The picture itself is fine, this is just not the right article for it. And I don't think it helps to clutter this article with images, just to show the image. The ratio of text to images should be reasonable, and after inserting the SET figure it seemed too crammed. We should focus on the basic infos on CB and SETs first, not start with sophisticated details. Please understand that I did not delete, destroy or vaporize the pics ;-). The characteristics image is nice, but we need something simpler more basic first like here. The pics I parked on the talk page are one step ahead for the current state of the article. --Dschwen 23:23, 5 March 2006 (UTC)

The link does not seem to work, could you look into that? The images, by the way, are on Wikimedia Commons, so there is no real need of parking them, but it's nice so that others can quickly check what we are talking about. Let's see how the other graphics explain the workings of the SET. --DrTorstenHenning 08:58, 6 March 2006 (UTC)

Sorry, I corrected the link, it was chopped at the end. The images in the linked article show schematic IV curves of I(V_source_drain) and I(V_gate), the stairs and the comb curves. --Dschwen 09:09, 6 March 2006 (UTC)

The first picture in the article conveys not much information to the reader, it neither explains what tunneling really is (for example how a particle wavefunction would behave in a tunneling barrier) nor does it introduce the concept of coulomb blockade (as I'd expect from an "above-the-fold"-picture). This might have to go too.

By go I hope you mean replaced with a better one. Otherwise, agreed. --DrTorstenHenning 12:31, 5 March 2006 (UTC)

Anyway, there is still some rewriting needed. I just noticed I duplicated factoids with my last edit... --Dschwen 19:11, 3 March 2006 (UTC)

A comment on the CB images presented on the first page. The conceptual image of evenly-spaced discrete levels in an SET is not correct as presented. It's important to remember that Coulomb blockade is a classical effect, that is, it arises from the charging of a classical capacitor. Therefore the charging energy is a Fermi surface effect. So, if one "looks" below the chemical potential of the island, one will not see a ladder of discrete states spaced by the charging energy. However, in a regime sometimes referred to as "quantum coulomb blockade," in addition to the classical charging energy, another energy scale appears known as the mean level spacing. This is derived from simple particle-in-a-box physics and demonstrates the 0D nature of a quantum dot or SET. In this case, the band diagram of the SET below the chemical potential is a ladder of states spaced by the mean level spacing. In transport, we see that to access the next available state we always have to overcome the classical (electrostatic) charging energy, but in this regime must also access the next discrete quantum level. There are caveats about the influence of spin and band degeneracy, but this picture is much more accurate than that presented on the main page. 192.5.18.32 21:14, 25 July 2007 (UTC)

Niemeyer-Dolan technique

I just started that lemma. I'd appreciate if you could go over it and give some input :-). To be continued here. --Dschwen 09:57, 6 March 2006 (UTC)

Who needs two junctions?

Is there anything that indicates that Coulomb blockade does not occur in a single junction? I don't think so, so I'll be reverting that bit until we have discussed this point. --DrTorstenHenning 12:33, 5 March 2006 (UTC)

Yeah, let's discuss. To my understanding discrete levels due to a high charging energy of a low self-capacitance are nescessary for coulomb blockade. With only one tunnel contact you'll get no such self-capacitance (island electrode), or if you construct an island at least you cannot attach a drain contact, which would require a second tunnel junction. What am I missing here? How would the blockade work with one junction? Would that really be a "coulomb" blockade? --Dschwen 23:14, 5 March 2006 (UTC)

As far as I understand things, you don't need an island since the single junction itself has a capacitance that is charged by a tunnelling electron, and hence Coulomb blockade occurs in a single junction. The difficulty is to observe it, that is, to have some kind of connection (lead) to the measurement electronics, and not to destroy the effect by the capacitances introduced by these leads. That is much less difficult when you have two junctions in series, that is, the well-known single electron transistor. I have not followed the very latest developments for a while (I myself tried creating very high-ohmic low-capacitance resistors for that purpose some years ago), but it appears that for example 1D arrays of Josephson junctions can do the job. I'll have to look up things again; the group of David Haviland at the Royal Institute of Technology in Stockholm is probably still working on that problem, and their publications should be a good starting point. Anyway, googling for "Coulomb blockade" and "single junction" gives a lot of hits, some to whole books. I think that the statement about "at least one junction" in the lead is justified, but if it's disputed, we would have to sort out whether the experimental observations are conclusive enough to lift CB in the single junction above the level of original research. --DrTorstenHenning 09:19, 6 March 2006 (UTC)

I've been doing some reading myself now, and now think you right, sorry for the confusion. A reference to Averin and Linkharev should be included, and the magical resistance of $25813\Omega$ should be explained in the article. I also think that a split of the article into Coulomb blockade and Single electron transistor is justified. --Dschwen 09:40, 6 March 2006 (UTC)

I'll just add my agreement that CB phenomena only require an island coupled through a junction to a gate electrode. The more familiar SET geometry is intuitively simpler and indeed, provides a direct means of probing the discrete charge states of the island, i.e. transport.192.5.18.32 22:03, 25 July 2007 (UTC)

In the first paragraph, it says "Because of the CB, the resistances of devices are not constant at low bias voltages, but increase to infinity for zero bias (i.e. no current flows).", and "one or more". This seems to imply that the current of a single junction drops off faster than linearly with voltage. I don't think that can be true, because the rest of the device is conducing so the charge will dissipate through its resistance, allowing another electron to tunnel. David R. Ingham (talk) 01:45, 21 January 2012 (UTC)

Literature section

I wonder if the presence of Henning's phd thesis and Pekola's article is relevant in a 3-refs long literature section. In other words, we should replace it with more significant refs : Millikan (1991), Giaever and Zeller (1968), Likharev (1985), Fulton and Dolan (1991) might be a good starting point. The link pointing at the European project is debatable as well : it looks like promoting Europa and biosensing applications, disserving other countries and applications. -- Moijaiunvelo 15:55, 20 August 2007 (UTC)

In an article on the history of Coulomb blockade research, these references would be well suited. Here, I think it would be best to point to recent review articles, and maybe also to Likharev's Scientific American article. The way things work today, resources available online and free of charge are certainly preferred by most users of an online medium like Wikipedia, that's why I included my PhD thesis (available free of charge from arxiv.org) here in the hope that someone might replace it with a better link (instead of just deleting it). --DrTorstenHenning 10:00, 31 August 2007 (UTC)

Suggestions for contents in an article about the single electron transistor?

I am currently working on a paper for school about the single electron transistor, SET, which is supposed to be ten pages long, and I am trying to come up with stuff that it should probably include. That is, what sections it should contain and what things there are about the SET that might be interesting to read about. So far, I'm planning to write about the function of a SET (how it works) and how to use one, when SET:s are most applicable, perhaps some different ways to manufacture SET:s, and the different between a SET and a classical transistor in terms of capabilities, limitations and other properties that might differ – all depending on what I can find about them of course.

I may also convert the article into Wikipedia after I have got it graded, to replace (or complement) the short section about them that currently exists in this article, why I thought you could help me out a bit and suggest sections that should be in it. So, anything it should contain? Any particularly interesting fact it really needs to contain? —Kri (talk) 21:05, 2 December 2011 (UTC)

It all depends on how you define a SET, if you restrict it to CB there might not be enough content. If you do not do this restriction and define a SET as a transistor-like device with three terminals and a central island using the quantum nature of electrons, then there is definitely enough content to fill a own page. The regimes such a SET can work in are defined by basically three energy scales, the temperature $T$, the charging energy $U$ and the coupling of your island to the leads, usually given as tunneling rates $\Gamma$. The CB regime then would be $\Gamma ,T\ll U$. Even there you will have access to sequential and co-tunneling when $\Gamma \ll T\ll U$ or when you increase the coupling $T\ll \Gamma \ll U$ you can see the Kondo effect. In the limit of strong coupling $U,T\ll \Gamma$ your tunneling barriers act as semi-transparent mirrors making the SET a Fabry–Pérot interferometer. After all I vote for an own article for the Single electron transistor. — Preceding unsigned comment added by Headtrick (talk • contribs) 12:12, 22 November 2017 (UTC)

Why does a single electron transistor need the tunneling junctions?

Why does a single electron transistor need to have the two tunneling junctions? Isn't it enough with the Coulomb blockade created by the gate?

If we would remove the tunneling junctions in order to get a single wire instead of a drain, a source and an island, still with the gate capacitively coupled to it (since the island does now not exist), wouldn't we still be able to create a barrier by applying a low potential to the gate, and to control the current that way (be able to switch it on and off), just like with a single electron transistor? —Kri (talk) 04:12, 4 December 2011 (UTC)

If the is no tunnel barriers, electron cannot be confined. You end up with a simple transistor with no single charge effects. You need to define clearly the dot in which you store a finite number of electrons, otherwise they can move freely in the device. BR 18:52, 29 September 2013 (UTC) — Preceding unsigned comment added by 83.204.168.202 (talk)

Coulomb blockade, a classical effect?

"Even when Coulomb blockade can be used to demonstrate the quantization of the electric charge, it remains a classical effect and its main description does not require quantum mechanics." I agree that a quick understanding of Coulomb blockade phenomena does not require high level in quantum mechanics, however is it possible to have a full classical calculation of the renormalization by Coulomb blockade of the conductance of a device where Coulomb blockade appears? For me, such calculation involves quantum many body physics technics (see: https://link.springer.com/article/10.1007/BF00683469 and http://jetp.ac.ru/cgi-bin/dn/e_041_02_0308.pdf) and therefore we cannot reduce Coulomb blockade to classical effect, even in the simple case. I propose then to change this sentence in a more correct way "Coulomb blockade phenomena, although are simple conceptually, require quantum many body calculations for a full understanding, even in the simplest case. Furthermore, devices where Coulomb blockade appears can be used as experimental test-bed for quantum physics, for instance to investigate the quantization of the electric charge (see: https://www.nature.com/articles/nature19072)"

Viselim (talk) 10:35, 10 December 2018 (UTC)

Coulomb blockade in the orthodox model just considers that the capacitance is not important (can be calculated through some classical electrostatics means) and then the Coulomb repulsion acts as if classical. It gives a qualitative good result. The only difference with a macroscopic device is that the quantization of electric charge has to be taken into account. For many-body calculations this might not be true, but the introduction is just describing the simplest model. Maybe a line or two can be written in the description sections. --MaoGo (talk) 11:05, 10 December 2018 (UTC)