Science a GoGo's Home Page Forums General Discussion General Science Discussion Forum diode array nanotech abundant clean energy

I don't understand what you mean by detail. The top face of the N type InSb substrate is sprinkled with C60 or C84 buckyballs at a density of 10^11 / cm2. Each buckyball is a diode anode. There is an ohmic contact metal layer on the back face of the substrate and a metal layer contacting the buckyballs on the side of them away from the substrate. There are still details like processing techniques which is what the $50K will be used to determine.

This is not a thermocouple. It can start from a mesoscale uniform temperature (temperature is not uniform on the nanometer scale without high heat conductivity). It will get cold as electrical energy is tapped off.

Aloha, Charlie

OK, so you have an InSb-C60 diode.Fine.Now explain the phenmenology,i.e. how does this thing work. How does it take heat from the ambient, and after it takes it, what happens such that you deliver a DC current?

And speaking of technical details, what thickness will the C60 layer be? Because you might have dominant tunneling from the InSb surface to the metal electrode above rather than "conduction" through fullerrenes. If I recall correctly (there are more than 10 years since I worked with fullerenes), the band gap for C60 is ~3.2-3.6 eV, and you do not have classical conduction of current through electron/hole flow but rather something called hopping (specific mostly to amorphous materials) on states at the edges of the band gap. And the hopping rate might in your case be smaller than tunneling from InSb to the metal electrode covering the C60.
Not to mention that you will need much more than 50K to produce such a device, because if you want a monolayer of C60 (you calim that each C60 will be an anode and I deduce you would like a monolayer), you will have to construct it by nanomanipulation. 50K won't even cover the cost of the instrumentation necessary (which BTW is not instrumentation that is appropriate for industrial processes). If you plan to deposit it by some technique or another, beware that C60 has a tendency to coalesce, to form clumps, hierarchically. Smaller clumps coalesce in larger clumps, which coalesce in even larger clumps, until you reach clumps of the order of half a micron in a layer of a few microns.

Pasti:

Thanks for the details of your concerns. I read from W point contact / insulator /Ni base diode papers that ~0.7 nm was the outer limit for tunneling so the C60 1.2 nm D spacing would be adiquate but close. The C60 is a shottkey diode anode where a high work function material will accept electrons.

Yes, a monolayer of widely dispersed C60s is needed.

Each diode as an electronic component has kTB watts of random power at its disposal. Such a diode is a nonlinear resistor with a direct associaton between conductivity and the instantanious distribution of the electrons in the semiconductor which set up a depletion zone of varying size and well defined stratification in juxtaposition to a conductive zone. When the electrons move towards the anode, conductivity increases: when they withdraw, resistivity increases. The insides of the diodes can change even if the outer extremities of all the diodes are tied together in massive parallel. I have
provided mental images of these features and processes elsewhere.


dcarnahan@nano-lab.com at www.nano-lab.com was investigating prototyping the diode array. They went so far as to borrow the use of spincasting machines. They were working on the clumping problem. I think that they didn't like how excited I was getting about the diode array comming out.

Clumping should be easy to solve in a 1/10% solution.

Aloha, Charlie

CB:"I read from W point contact / insulator /Ni base diode papers that ~0.7 nm was the outer limit for tunneling so the C60 1.2 nm D spacing would be adiquate but close. The C60 is a shottkey diode anode where a high work function material will accept electrons."

First of all, it is a Shottky diode, not shotkey. And those arguments you read for the W/D/Ni do not necessarily apply in the case of your configuration. Do not confuse a layer of insulator wit a monolayer of C60. Monolayers behave very differently than layers. In the cse of the monolayer, for example, you will not have a gap in the C60 (you needa C60 layer for the existence of such a gap), you will only have states introduced by the C60 in the BV/BC of the InSb substrate, and a slight modification of the band structure in C60. So traditional diode theory on your device is lost, it won't work that way.

CB:"Yes, a monolayer of widely dispersed C60s is needed".

Well, then you must apply for a grant that is at the very least one order of magnitude larger than what you have in mind.

CB:"Each diode as an electronic component has kTB watts of random power at its disposal. Such a diode is a nonlinear resistor with a direct associaton between conductivity and the instantanious distribution of the electrons in the semiconductor which set up a depletion zone of varying size and well defined stratification in juxtaposition to a conductive zone. When the electrons move towards the anode, conductivity increases: when they withdraw, resistivity increases. The insides of the diodes can change even if the outer extremities of all the diodes are tied together in massive parallel. I have
provided mental images of these features and processes elsewhere."

As I said, spare me the mental images, WMD on a pizza slice and such. Try to explain the physical phenomenology.

Above you have only described the equilibrium of a standard MOS device, which is nothing new. This (fluctutions that is))does not yield you any net current, regardles of the impedance of the diode. So, how does one of your diodes work to transform surrounding heat into current?


CB:"dcarnahan@nano-lab.com at www.nano-lab.com was investigating prototyping the diode array."

I found no useful info on that site, sorry.

CB:"They went so far as to borrow the use of spincasting machines. They were working on the clumping problem."

Spincasting of C60? you must be kidding. And those fellows definitely have not too much of an idea about C60 deposition.

First of all, C60 is a rather insoluble matrial.

Second of all, even if you dissolve it in something (usually a polar solvent), it behaves like a colloid and will automatically coalesce in solution. I could have told you that before even attempting to spincast films.

Third of all, good luck with obtaining a monolayer of C60 by spinning. That will be a stupid waste of research money.

Fourth of all, the clumping I was talking about is present in the layer, after the deposition, and not in the gel.

Fifth of all, even if you make a monolayer of C60, the metal deposition for the upper electrode will short out your C60 monolayer.Think about how you can tightly pack speres on s asurface, and how metal atoms will get in the deposition process into the empty spaces resulting from the packing od the C60 spheres. You might want to look at more than one monolayer of C60.

CB:"I think that they didn't like how excited I was getting about the diode array comming out."

Optimism is good, enthusiasm is good. Pestering though, can become annoying. it would have pissed me off too.

CB: "Clumping should be easy to solve in a 1/10% solution."

Good luck with that!

No offense, but I suggest you read more about the physics of a diode/MOS device, C60, deposition techniques and similar. Up to this moment, you have disregarded elementary phenomenology in a device like yours, and applied to it theory that just doesn't work that way in your particular case. You would have an extremely hard time convincing me to invest in your ideea.

Pasti: Thank you for being a now-hear-this person not a yes man.


The buckyballs are to be directly in contact with the substrate.
Part of the research will be to find out if the semiconductor under the buckyballs has to be annealed.
The embeddant surrounds the buckyballs. The paper on tunneling insulators assures me that 1.2 nm is thick enough to be an electrical barrier.
Researchers have produced monolayers of uniform 12 nm Co spheres . C60 may also be dispersible though it is much smaller.
A diode array based on ~10 nm metal spheres could be made but should be done only if it doesn't deflect too much money from the goal of maximum diode packing density.
Johnson Noise is disequilibrium power in resistors.
Perhaps buckyballs could be dispersed dry electrostatically under a very thin cover slip.
Buckyballs as particles can be suspended in any liquid; perhaps some liquids would impart an electric charge on the buckyballs which would disperse them.
Mike of the nanofab network gave me the initial $50K quote. If serious money begins to emerge, the quote can be refined. During the refining everyone can make sure that adequate assumptions are made.

Aloha, Charlie

CB:"The buckyballs are to be directly in contact with the substrate."

OK.

CB:"Part of the research will be to find out if the semiconductor under the buckyballs has to be annealed."

?!? And you need the annealing why? You already have the semiconductor structure present, you don't need anymore annealing.

CB: "The embeddant surrounds the buckyballs."

If you use an embeddant, for the most common embeddants, your conduction will go through the embeddant, not throught he fullerenes.

CB: "The paper on tunneling insulators assures me that 1.2 nm is thick enough to be an electrical barrier."

Well, that might work for some layers of insulators, but does not work for monolayers. It is elementary solid state physics.

CB: "Researchers have produced monolayers of uniform 12 nm Co spheres.C60 may also be dispersible though it is much smaller."

Yes, and they have produced also monolayers of polymers through dipping, but how is this relevant to what you want to do? The fact that Co can be depositied in monolayers has no bearing on the deposition of C60. Each element/compund has its own physics and chemistry, and you cannot extrapolate one's behavior to the other. According to your reasoning, just because Co can be radioactive, so should be C60. And this is not whty C has radioactive isotopes.

CB:"Johnson Noise is disequilibrium power in resistors."

No, it's not. It is due to the scattering of the electrons on the phonons.This does not mean non-equilibrium (instead of disequilibrium). You should also learn the technical lingo if you want to be able to convince investors....

CB: "Perhaps buckyballs could be dispersed dry electrostatically under a very thin cover slip."

Right! That is why they come in soot-like form, because they can easily be diespersed electrically! I don't think so.

CB: "Buckyballs as particles can be suspended in any liquid;..."

This is gibberish. Big time. Either browse the literature in the mid-nineties, or just try to disolve C60 in water and benzene. You will see exactly how "easy" it is to suspend C60 in any liquid.

CB:"...perhaps some liquids would impart an electric charge on the buckyballs which would disperse them..."

Riight!Why don't you actually browse the literature on fullerenes, and see waht a colloid is, and how a colloidal suspension id formed.

CB: "Mike of the nanofab network gave me the initial $50K quote. If serious money begins to emerge, the quote can be refined. During the refining everyone can make sure that adequate assumptions are made."

Well, what can I say, if Mike from nanofab said it, how can I contradict him...!Charlie, as much as I appreciate your enthusiasm, in this way you won't get a dime from investors. You are not proficient about the current status in the fields that are important to your project, you lack basic knowledge regarding the elementary processes involved in your device, and even worse, we've been corresponding for some time now and you still haven't explained to my satisafaction how your device works, how it takes heat from the ambient and converts it into DC current.

The most important question on this topic is if Johnson Noise, thermally excited thermal radio frequency power, can be rectified and then the rectified output of many diodes in consistent physical orientation parallel can be aggregated for an external electrical load with the load imposing only the proper burden on the supply.

May I hear a variety of responses to this question?

Aloha, Charlie

CB:"The most important question on this topic is if Johnson Noise, thermally excited thermal radio frequency power, can be rectified and then the rectified output of many diodes in consistent physical orientation parallel can be aggregated for an external electrical load with the load imposing only the proper burden on the supply."

Aha, so this is what you assume your current generating mechanism is for your diodes.

Well, there is only one (fundamental) problem with this. The "Johnson noise mechanism" is not a current generationg mechanism in a junction. It is an effect of the resistance/impedance on the current flow. ALREADY EXISTING current flow. And it produces fluctuations in the already existinc DC?AC current.

As I said before, the Johnson noise appears because the electrons "hit" the ions in the wires, resistances, etc, and instead of flowing straight, their flow is perturbes by these "colisions" with the ions. It is similar to the way an athlete has to go through a bunch of swinging tires or sandbags. Instead of running straight the distance, he is hit ny the swinging bags, and runcs much slower, and with much more energy consumption. Furthermore, this noise is random, so it has no DC component.

There are only a few mechanisms generating carriers in a diode: photoeffect, biasing, thermoelectronic (Richardson) emission. And as far as I know, all the applications for these effects have been developed almost half a century ago.

If indeed you want to harness heat from the environment, the best bet would be photodiodesdiodes for the optical and far IR spectrum, so that they can take the heat and transform it into electric current. But then these are avenues already already explored by the solar power "people".

So no, with this mechanism your diodes won't do anything aong the lines you say they would. It ain't working that way.

Many noise generators produce noise greater than kTB by sending current driven by a voltage, the two factors of D.C. Power, through a noisy medium like a vacuum or gas tube, or semiconductor. The noise power is less than the D.C. Power. Such a beginning inevitably produces losses. Johnson Noise in a resistor does not require energizing D.C. electrical power. Mobile electrons in a resistor move at random to produce random radio frequency power. At a rectifying junction between N type InSb and a C60 Shottkey anode, more electrons can be conducted forward from the semiconductor to the anode than in the reverse direction. In a massively parallel system with all the anodes on a branch supplying electrons to a load and all the cathodes on the electron return branch, net current can be aggregated by a plurality of diodes at an equilibrium voltage supplying a higher power load.

Aloha, Charlie

Aha, this becomes intresting. You are actually talking about thermal noise in a resistor (the thermal component of the Johnson noise). OK, that would make some sense.

But now I am even more confused.Why do you need a didode array? To incorporate the rectifier in the device? The reasoning for the rectifying junction as you mention it does not apply for your diode, since the diode behavior does not apply to random phenomena of this sort. If it did, you would obtain infinitely increasing voltage between the electrodes, which does not happen.

But if this is your ideea, this is trivially easy to test. Forget the rectification for a moment, and go to Radio Shack to by a bunch of resistors of a few megohms. Then connect them in whatever configuration you like (parallel seems to be your preference, such that you "gain" current). Then measure the thermal noise at the electrodes of your system (for this you will need either a ratherexpensive DVM, or alternatively an even more expensive frequency analizer or scope)for a varying number of resistors connected since you are about to measure microvolt rms voltages). It might help if you had a friend, aquaintance in a university lab or in an electronics lab. See if your total Johnson current actually increases!
Then heat the resistors with a hair dryer, and see what happens. If it works with the resistors, then you can think about rectification. But nevertheless, you would have a tangible (and easily reproducible) proof that your ideea works in practice. Which might be very persuasive for your (potential)inventors.

Sorry for the too many typos in the last message, I really am in a hurry.

its ok ...
but i opened the message to see whether any conclusion has been reached...but i guess not yet...
It is theoretically possible to generate light from bio-degradable substances..
their are insects and fishes which do it... we need to understand their technology and then we will be able to get the eco friendly light without coal and reactors...
Essentially the light is generated using a chemical reaction in which photons play a vital role... energy level change is triggered with the absobtion of low frequency photons and light is visible with the emission of high frequency photons..
Certain Biological chemical substance have meta stable state which is more likely to be reached when a fall of energy level is triggered.
Lev1
Lev2
Lev3
From lev2 electrons reach to level1(using infra red photon absorbtion) and then it falls straight to lev3 generating visible light of high frequency...
Overall energy loss is covered up by agitaing the atoms in a particular fashion by the generated heat of the chemical reaction triggered by the 'food'eaten by the insect.
Cheers

Hmmm, wouldn't the internal resistance of the voltmeter have to be much larger than that of the resistors? The Nyquist noise inside the voltmeter would be dominant, unless you use a cryogenic voltmeter. But lowering the temperature of the voltmeter would defeat the purpose of the experiment.

You're right, it should be larger, but since he wants a parallel configuration, his resistance will drop to 100's of kohms, and any respectable DVM should be able to handle that. But in case he wants a series configuration, he should probably use an op-amp buffer (which should have it's own problems, but never mind that). But the ideea was that there is a much simpler circuit he can use to test his concept, without the C60/inSb malarkey.
But something sounds fishy to me to this thing as a source of current. I don't exactly know what it is, but if you apply a classical reasoning, the resistors should heat in cascade up to breakdown. If I have some time, I will take a look.

I believe the written sources that tend to establish that several resistors in a group revert to a single equivalent resistor.

I believe that the parallel diodes will gather a reverse open circuit voltage up to the infra red photon energy or a voltage where reverse avalanche breakdown current returns the power.

I do expect each diode to rectify the Johnson noise at the depletion region with its asymmetrical tunneling making the forward resistance lower. The interior of each diode is a separate kTB element.

I vaguely remember from Scientific American a chemical system using a Cr compound that collected IR photons and released an occasional visible light photon.

Aloha, Charlie

CB:"I believe the written sources that tend to establish that several resistors in a group revert to a single equivalent resistor."

Well said that you believe the above. Now, it is the time that you actually start looking at the details. First of all, your noise voltage is not proportional to the resistance, it is proportional to the square root of the resistance in the case of thermal noise. Second of all, the current have a random distribution, which means that you cannot exacly add their rms values to get the total rms value. So it maters that you have several resistors in series/parallel, when it comes to testing your theory.

CB:"I believe that the parallel diodes will gather a reverse open circuit voltage up to the infra red photon energy or a voltage where reverse avalanche breakdown current returns the power."

Ok, OK, you belive that too. The problem is that your "beliefs" above do not make sense. A voltage cannot be compared with photon energy, so you most likely wanted to state something else. The sentence " voltage where reverse avalanche breakdown current returns the power" really does not make any sense, except maybe to you. In general, once the avalanche process starts, you cannot stop it, and the device gets shot.

CB: "I do expect each diode to rectify the Johnson noise at the depletion region with its asymmetrical tunneling making the forward resistance lower. The interior of each diode is a separate kTB element."

First of all,unless you use tunnel diodes, there is no tunneling involved. In MIS diodes, you do not have (in general) tunneling. What you have is called conduction, and the mechanism for charge flow is well known and it's not tunneling.
Second of all, do you have a model of your diode (you know, resistances, capacitances, idela diodes, noise sources) that justifies your claim?

CB: "I vaguely remember from Scientific American a chemical system using a Cr compound that collected IR photons and released an occasional visible light photon."

So? What relevance do your memories have regarding your device? I remember in Nature an article stating that you can achieve cold fusion in a glass, or that some morons form Samsung or so exceeded the speed of light. See any relevance with respect to your device?

I agree that the square root of the voltage is proportional to the resistance for constant power. This means that the reciprocal of the square root of the current is proportional to the resistance for constant power. These relationships hold whether the resistance is in one resistor or many. For example, 4 resistors of 1 ohm each in series is the same as a single 4 ohm resistor. If the constant power is 1 watt , 2 volts is involved when the resistance is 4 ohms (either 1+1+1+1 ohms or 1 x 4 ohms) and 1 volt is involved when the resistance is 1 ohm.

Tunneling is a colorful word for leakage conductivity through a barrier. There is some leakage through the depletion zone in standard diodes. A diode has asymmetrical conductivity involving amounts of current useful in a circuit through a narrow depletion zone under forward voltage conditions and a useful blockage of current through a wide depletion zone under reverse voltage conditions.

Constant leakage current created by hole / electron pair generation will eventually keep the reverse voltage from going to infinity. Experiment suggests that the open circuit voltage is 0.1 V. If a well designed array operates at a higher voltage, I will drop my provisional hypothesis of an association to IR photon energy. Electromagnetic energy does involve photon energies measured in electron volts where an electron volt is the energy of an electron accelerated by an electromotive force of one volt.

Aloha, Charlie

OOPS I made another time wasting mistake:

The numerical example of the resistors is correct but the description should be that constant power implies that the voltage squared is proportional to the resistance. Constant power also implies that the reciprocal current squared is proportional to the resistance.

Aloha, Charlie

Charlie, just assume that this works. Then, unless some ghosts are making it work, you could simulate this process on a computer. Since the total entropy decreases in this process, the simulation cannot always map different initial states to different final states. I.e. there must exist states x and x', with x different from x' such that after time t both are mapped to the same final state.


But that would mean that the system is not invariant under time reversion. Also you can see that you cannot (always) calculate the initial state from the final state. So, the simulation must be throwing away information (which contradicts the laws of physics).

Information about random states is uninportant to collect, hold, or use. Maxwell's demon was burdended with too much work. He had to figuratively put his palm out for a tip, inspect for smuggled towels, remember everyone's name and call the car valet. Capturing energy as the guests walk over pressure bladders won't pay the wages here. If molecules push through flaps because they are there and going in the preferred direction and the event is not recorded in a high signal to noise memory the energy from returning the compressed molecues can run a turbine producing net energy.

Aloha, Charlie