Sciencemadness Discussion Board - Spin locking of electrolytic reactions

Spin locking of electrolytic reactions

merrlin - 4-4-2009 at 22:36

This topic is a continuation of my earlier posts on the "reactions that emit radiation" thread. I am developing a microwave-pumped electrolytic cell that will be used to investigate spin locking of electrolytic reactions. In contrast to conventional microwave pumping by irradiation (e.g. in ESR), I am using a current flowing in the cathode of the electrolytic cell to produce an oscillating magnetic field at the surface of the cathode. The cathode is part of a transmission line structure driven by a solid-state microwave current source. RF Magnetic fields with intensities greater than 10 Gauss may be achieved at the cathode surface with currents on the order of one ampere at low voltage. In combination with an applied DC magnetic field, The cell provides for pumping of unpaired electrons at the cathode surface and in the adjacent electrolyte. It is my assumption that an electron donated by the cathode and a cationic species may be considered as a "quasi-radical pair." In operating the cell, a DC reduction potential is applied to an electrolyte flowing through the cell to obtain a current associated with a reduction reaction. A DC magnetic field is applied to provide Zeeman splitting, followed by the application of microwave current to provide magnetic pumping at the cathode surface.

My Goal is to achieve pumping of the free electrons contributed by the cathode in combination with isotope selective pumping of a cationic species so that the quasi-radical pairs containing a particular isotope at the cathode surface are locked in a triplet state (T+, T-). During the sweep of the DC magnetic field/RF current frequency, resonance and locking would be detected by a drop in the observed reduction current.

In 2001 Buchachenko wrote " Besides the many factors controlling nuclear spin selectivity, there are two outstanding and highly promising but not yet properly exploited, microwave induced MIE and dimensionality." The system I am developing addresses both of these factors. I am interested in isotope separation and selective organic synthesis, and I believe that proper selection of electrode and electrolyte materials can greatly reduce the relaxation effects that work against efficient spin locking at a cathode surface. Although I will be starting with H2O and D2O based electrolytes, I would like to hear from anybody who has experience with aprotic solvents or room temperature ionic liquids, and transition metal complexes.

bquirky - 4-4-2009 at 23:36

!

sounds interesting whould an electrolyte simalar to a lithium battery be suitible ie propylene carbonate & Dimethoxyethane ?

Thats just a wild guess that sounds like a spectacularly complex experiment !

How do you measure the electron spin or know that its working as you expected ?

hissingnoise - 5-4-2009 at 06:32

Most here merrlin are interested in the applications of electrochemistry rather than the cutting-edge nitty-gritty. . .
You would seem to have more to teach than to learn.
Pictures of cell(s) you're working on would be welcome. . .

Mr. Wizard - 5-4-2009 at 07:06

It sounds like a bunch of bull to me. Is it some kind of contest to stack as many eclectic exotic phrases together and see how many people will swallow it? I was checking the posting date, looking for April 1

hissingnoise - 5-4-2009 at 08:11

If it is just an exercise in urine-extraction, it's a fairly sophisticated one!
I'm waiting for the shots!

merrlin - 5-4-2009 at 10:43

Quote: Originally posted by Mr. Wizard

It sounds like a bunch of bull to me. Is it some kind of contest to stack as many eclectic exotic phrases together and see how many people will swallow it? I was checking the posting date, looking for April 1

Have you read Buchachenko's book or any of the papers I posted?

merrlin - 5-4-2009 at 13:15

Attached are the following files:

system_diagram.pdf--This file shows an overview of the system. RF components can be obtained from Mini-Circuits, data acquisition modules from Measurement Computing, circuit boards produced by AP circuits, and the ICs and passives can be obtained from Digi-Key. Not including the PC, The system comes in at around $5000.

60w__1di-1g_pfa_280Mhz_400-30.asc--This file is the LTSpice simulation file for the electrolytic cell and microwave driver. If you download LTSpice you can tweak it and observe the results. The transmission line section is .4 inches long, and uses the distributed parameters obtained from the E-M model. LTSpice is an excellent version of PSpice that can be downloaded from Linear Technology. Their 1084 series voltage regulators and LT1210 op amp will be used to control the DC magnetic field sweep and provide a stable 5V for instrumentation power.

I have been working on this for a couple of years and a considerable part of my time has been spent trying to find out if anybody has done this before or whether there is a reason it wouldn't work. The brief correspondence I have had with academic experts leads me to believe that it is novel and cannot be rejected out of hand on a theoretical basis. I have been advised to take a combinatorial approach. There are literally millions of electrolyte compositions that can be run through the system, and I am trying to learn which would be the best to start with.

I have tried to attach some graphics files (.png), but they are being ignored.

Attachment: system_diagram.pdf (28kB)
This file has been downloaded 1412 times

Attachment: 60w__1di-1g_pfa_280Mhz_400-30.asc (19kB)
This file has been downloaded 948 times

merrlin - 5-4-2009 at 13:30

I've been able to attach graphics files by changing the .png extension to ".pngfile." Hopefully they will be downloadable and visible after editing the extension back to ".png."

E_cell_2D_E-M_model_structure.pngfile--This file shows a 2D cross section of the cell. The model employs symmetry to reduce node count and computational load. Supporting structure is FR4 glass-epoxy. Sea water is the electrolyte. A .001"T x .060"W (.030"half width shown) cathode is shown above .001" x .300" ground plane, separated by .001" PFA dielectric film and .001" electrolyte gap.

E_cell_2D_E-M_model_flux-density_overall.pngfile--This file shows the overall magnetic flux density plotted for a cathode excitation current of 2 amperes at 280MHz. The structure model has the following distributed characteristics: 80n H/m, 4.78 W/m (2.39 ohms/m), and 755 pF/m.

E_cell_2D_E-M_model_flux-density_closeup.pngfile--The magnetic flux in the gap is about 1.4 millitesla (14 Gauss).

LTSpice_plot.pngfile--This plot shows the input current I(R1) and the magnetic pumping current I(R63).

Attachment: E_cell_2D_E-M_model_structure.pngfile (7kB)
This file has been downloaded 932 times

Attachment: E_cell_2D_E-M_model_flux-density_overall.pngfile (8kB)
This file has been downloaded 946 times

Attachment: E_cell_2D_E-M_model_flux-density_closeup.pngfile (8kB)
This file has been downloaded 913 times

Attachment: LTSpice_plot.pngfile (21kB)
This file has been downloaded 923 times

merrlin - 5-4-2009 at 14:27

Quote: Originally posted by bquirky

Propylene carbonate would be a good candidate due to its voltage window. Since the cell electrolyte gap is on the order of a few mils, do you know if viscosity would be a concern? Since the electrolyte is pumped through the cell, much of the apparent conductivity of the electrolyte solution will be due to bulk transport as opposed to diffusion.

After the electrolyte is flowing in the cell and the applied DC potential produces an observable current associated with a reduction reaction, the RF current is applied to produce the high frequency magnetic field. The DC magnetic field is then swept over a range of values. If resonance occurs and spin locking is achieved, a drop in the reduction current would be expected. The reduction current for a .06" by .400" electrode is expected to be on the order of .1 to 1 milliampere. Spins would not have to be measured, and a sweep could be done in less than a minute for a particular RF frequency and amplitude. Once resonant conditions have been established for a particular cation (e.g., iron) the cell could be run to collect the reduction reaction product. Mass spectrometry of the product or electrolyte solution would then be used to determine which isotope(s) had been locked.

Two electrons with the same spin cannot share the same orbital. If an unpaired electron and an electron donated by the cathode have the same spin, reduction of the cation cannot be completed. The object is to pump the electrons under resonant conditions so that they have the same spin. The good news is that very little energy is required to move an electron between magnetic energy levels. The bad news is that the small fields generated by the movement of species in the electrolyte and hyperfine coupling with other spins lead to spin-lattice and spin-spin relaxation. The goal is to provide pumping at sufficient amplitude and frequency to overcome relaxation effects. Relaxation effects can also be reduced by electrolyte component selection.

merrlin - 5-4-2009 at 14:58

Quote: Originally posted by hissingnoise

Most here merrlin are interested in the applications of electrochemistry rather than the cutting-edge nitty-gritty. . .
You would seem to have more to teach than to learn.
Pictures of cell(s) you're working on would be welcome. . .

The applications I am pursuing are isotope separation and organic synthesis. An industrially useful system would probably require thousands of cells, but I think that in those quantities the cells could be constructed for less than $100 each.

Having worked on various missile programs, I am familiar with microwave circuits and specifications for electroplating (my background is materials science), but I have very little hands-on experience with electrochemistry. My experience is pretty much limited to electropolishing and a little bit of copper plating.

I hope you are able download the .png files. As stated in an earlier post, I am just finishing the design phase. The electrolytic cell is basically a shorted transmission line that is masked so that part of the top conductor facing the ground plane is exposed to a flowing electrolyte. The concept is straightforward but there is a lot involved in optimization. In keeping with the philosophy of "measure twice, cut once," I have spent a lot of time on design and analysis. I considered designing and building the RF amplifier, but decided to buy the components instead. The microwave world is focused on delivering maximum power to a matched load. In this system, the goal is to develop a current with minimum power input. A good EE/technician could probably cannibalize a cell phone to produce a driver for a smaller version of this cell.

Mr. Wizard - 5-4-2009 at 16:14

The more I read, the more it sounds like horse feathers.

"A good EE/technician could probably cannibalize a cell phone to produce a driver for a smaller version of this cell." Would you hook the cellphone to some solar powered yard lights?

Bafflegab Bullshit.

FrankRizzo - 5-4-2009 at 16:50

Mr.Wizard,

You *do* realize that cell phones use microwave frequency transmission circuits, right?

merrlin - 5-4-2009 at 17:22

Quote: Originally posted by Mr. Wizard

The more I read, the more it sounds like horse feathers.

"A good EE/technician could probably cannibalize a cell phone to produce a driver for a smaller version of this cell." Would you hook the cellphone to some solar powered yard lights?

Bafflegab Bullshit.

It isn't how much you read, it's how much you comprehend. So I'll rephrase the question:

How much of Buchachenko's book and the journal articles I posted do you understand? Do you understand the circuit I posted?

Mr. Wizard - 5-4-2009 at 18:26

Quote: Originally posted by FrankRizzo

Mr.Wizard,

You *do* realize that cell phones use microwave frequency transmission circuits, right?

No, why don't you explain it for me? Why would somebody bother with a cell phone when you can get real watts out of a microwave oven? Oh wait, I know; you can set the transmission frequency by punching in the secret numbers available at certain sites on the internet, and you can pack the phone in dry ice to keep it from overheating when the SWR and impedance mismatchs between the dilithium crystals and the beer can wave guide you made. Is there some special benefit to the 1.8 GHZ compared to 2.4 GHz or are you using a 900 MHz one?

If only I was smart enough to pick up on the secrets that are being scattered in front of me, and read all the very complicated papers that are shown in all the links.

Where did I put the link for the mono-atomic magnetic spin stabilized hydrogen rocket fuel with the nickel 1/4 wave matrix stabilizer and doppler damper?

Seriously, I think somebody is blowing smoke here. Am I the only one not biting on this?

Sedit - 5-4-2009 at 19:34

Nope not the only one not hungry for the bait, Im on the edge though. It seems as though he knows what hes talking about but that can with ease make someone that dont know WTF hes talking about say.."um... OK sure..". Tell us a bit about your self, your company..merits ect..

If you could dumb it down for us as simple as possible that could quite easly help matters. To me it basicly sounds as though your trying to do basic electrolysis with AC current overlay such as used in some cells to improve yeilds or cause reactions that would happen otherwise. Just in this case it seems ultrahigh frequencys are being used. Although if thats the goal it all seems a little over the top.

Dumb it down and explain to us laymen terms because if one isn't able to do that either there full of shit or they succeed and no one in the general public cares because they see no use in the overly complicated equipment. Just a thought.

FrankRizzo - 5-4-2009 at 19:56

Quote: Originally posted by merrlin

Quote: Originally posted by Mr. Wizard

He doesn't need to read or comprehend the articles or schematic because somehow he's already made up his mind that you're a fool. I guess he doesn't realize that people rarely take the time to make a post unless they think their ideas have merit, or they're trolling You don't seem to be doing the later.

How did you first come across the effect in question? Is this a thought experiment come to the physical experiment state, or did you notice some anomalous effect in another project? I'm interested, please continue posting.

merrlin - 5-4-2009 at 23:12

Quote: Originally posted by Sedit

Nope not the only one not hungry for the bait, Im on the edge though. It seems as though he knows what hes talking about but that can with ease make someone that dont know WTF hes talking about say.."um... OK sure..". Tell us a bit about your self, your company..merits ect..

If you could dumb it down for us as simple as possible that could quite easly help matters. To me it basicly sounds as though your trying to do basic electrolysis with AC current overlay such as used in some cells to improve yeilds or cause reactions that would happen otherwise. Just in this case it seems ultrahigh frequencys are being used. Although if thats the goal it all seems a little over the top.

Dumb it down and explain to us laymen terms because if one isn't able to do that either there full of shit or they succeed and no one in the general public cares because they see no use in the overly complicated equipment. Just a thought.

I'll give it a shot, but I am going to have to ask you to read the section 2. "The Radical Pair Mechanism" in the file "A study in Spin Chemistry.pdf." I've been studying this stuff for 2 1/2 years and it isn't easy. The key here is that reaction between radicals in a radical pair in a triplet state is forbidden. One of the mechanisms that can enhance conversion of the triplet state to a reactive singlet state is hyperfine coupling between an unpaired electron and a nucleus with a magnetic moment. Most of the experimental work that has been done has involved organic compounds containing both carbon-12 and carbon-13. Since carbon-13 has a magnetic moment and carbon-12 does not, carbon-13 containing radical pairs will have a slightly higher probability of recombining without diffusing apart. This provides a basis for isotope selective reactions.

Consider two identical chemical ions or radicals that differ only in their isotopic composition. One has a magnetic nucleus (e.g. carbon-13) and the other does not (e.g. carbon-12). The hyperfine coupling between the magnetic nucleus and the surrounding electrons changes the available energy levels and thus the energy differences between levels. The differences provide the possibility of selective pumping. The differences between energy levels correspond to resonant frequencies. Nuclear Magnetic Resonance and Electron Spin Resonance both use applied DC magnetic fields to split energy levels, followed by the applications of high frequency radiation to determine resonant frequencies. NMR operates at lower frequencies than ESR because the differences in energy levels are smaller. For an electron, the resonant frequency has a proportionality of about 28MHz per mT of applied DC field. Most ESR work is done at X-Band or other bands where the equipment is readily available, not because there is anything special about a particular frequency. Although ESR at higher frequencies can provide more detailed information, ESR has also been used at around 250MHz. I've chosen to start at 280MHz and a DC field of about 10millitesla. Lower frequencies are easier in terms of dealing with skin depth losses and standing waves. If Mr. Wizard had bothered to look at the circuit model he would have noticed that the current in the resistor at the shorted end of the transmission line is only slightly greater than the current in the resistor closest to the source. The electrical length of the transmission line is much shorter than a 1/4 wavelength and I don't anticipate standing wave problems until operating at about 1.5GHz. At these frequencies the cell doesn't have to be treated as a distributed circuit. Since the transmission line is shorted and doesn't have to be connected to the anode, there is only a very small high frequency current flowing through the electrolyte. The high frequency current is simply there to provide the magnetic field that is traditionally supplied by electromagnetic radiation.

The attached paper "Control of a Chemical Reaction by Spin Manipulation of the Transient Radical Pair.pdf" describes the use of microwave radiation to control a photolytic reaction. Notice that they are using a high power tube (1kW) and mention a maximum microwave magnetic field of about 5mT. By running a microwave current in the transmission line geometry I have posted , a field of 1.4mT can be achieved with a power dissipation of less than a watt in the transmssion line, albeit at a much lower frequency.

Since electrolytic reactions occur at or very near the electrode surface, a magnetic field can be confined to a small volume adjacent to the cathode surface. If the magnetic field is produced by a current flowing in the cathode, it will have its greatest intensity were it is most desired. Since no metal is a perfect conductor, the magnetic field will actually penetrate a finite distance into the electrode surface.

The process of electroplating involves a free electron donated by the cathode combining with a cation that has unpaired electrons. At some point in the process, the donated electron and an unpaired electron will have to share an orbital. I consider the combination of two to be a "quasi-radical pair." It has been well demonstrated that radical pairs are subject to spin manipulation by the application of microwave magnetic fields. The question I seek to answer is whether electrolytic reactions can be spin manipulated. I believe they can, if you put them in a strong enough DC magnetic field and apply an oscillating field that strong enough and fast enough to put the unpaired electrons next to the cathode into the same spin state.

I know that the structure I am building can produce locally intense microwave fields with far less power than traditional approaches. I worked on the AMRAAM and HARM missile programs in the 1980's so I know something about microwave integrated circuits. (Mr. Wizard, I even worked with tungsten/copper alloys!) I have also discussed my project with EE friends who are still in the microwave business and they can't see a problem with the basic physics of my approach. The chemistry professors (two of them spin experts) have told me the approach is novel and interesting. One of them told me not worry about theory at this point, but to try some experiments. The point is, this may not work, but it wasn't obvious to the experts why it wouldn't. I got involved in this when I learned how expensive some isotope separation techniques are. When I learned a little about spin chemistry and the DC and AC field required for spin manipulation, I knew that I could build a small structure that could produce fields more efficiently and also be combined with an electrolytic cell. Ultimately, the primary source of power loss is going to be whatever switching device is used to generate the high frequency input.

One of the attached papers that helped convince me to pursue this project has nothing to do with spin chemistry. "The isotopic effects of electron transfer-An explanation for Fe isotope fractionation in nature.pdf" describes isotope separation during electrodeposition of iron without any externally applied magnetic fields. It is not clear to me that the technique is independent of hydrogen generation at the cathode, but the results are interesting. If a simple high-current electrodeposition process can produce isotope fractionation, what might happen if you give it some help?

Attachment: The isotopic effects of electron transfer-An explanation for Fe isotope fractionation in nature.pdf (355kB)
This file has been downloaded 738 times

Attachment: Control of a Chemical Reaction by Spin Manipulation of the Transient Radical Pair.pdf (353kB)
This file has been downloaded 831 times

Attachment: A study of spin chemistry.pdf (406kB)
This file has been downloaded 816 times

not_important - 6-4-2009 at 07:26

Think it might have gone better had you made your first post more like this most recent one, including those background papers. Explain the concept, then move on to how you propose to exploit it.

A lot of this is pretty recent, along with the related studies on cryptochromes, it looks as if there is active debate over it. I'll delve deeper into those papers tomorrow.

watson.fawkes - 6-4-2009 at 09:05

Quote: Originally posted by merrlin

It is my assumption that an electron donated by the cathode and a cationic species may be considered as a "quasi-radical pair."

What spin state are you trying to get the cathode-donated electron into?

If you're looking at free electrons, for example, they follow the Pauli exclusion principle with 1s electrons. For example, free electrons create atom-sized bubbles in liquid helium. If you're looking at cathode-bound electrons, the available electron states are entirely dependent upon the electrode material.

These kinds of basic material choices and reaction goals drive the rest of the apparatus. It's difficult to offer advice without knowing this.

watson.fawkes - 6-4-2009 at 09:08

Quote: Originally posted by Mr. Wizard

Seriously, I think somebody is blowing smoke here. Am I the only one not biting on this?

It's real. On the other hand, if you haven't had undergraduate quantum mechanics, it's not going to make much sense. Learn about the principal quantum numbers of the atomic central force solutions to Schrodinger's equation. Learn how angular momentum coupling works. Understand Zeeman splitting. That's a start. There's more after that.

merrlin - 6-4-2009 at 12:19

Quote: Originally posted by watson.fawkes

Quote: Originally posted by merrlin

It is my assumption that an electron donated by the cathode and a cationic species may be considered as a "quasi-radical pair."

I have spent some time studying electrodics (Bockris' Modern Electrochemistry 2A) and I have only learned a little bit, but I've learned that an electrode surface and the adjacent interphase is a complex place. I'm sure some of you are better mathematicians than I am, so I am going to try and present my qualitative perspective rather than getting in the way of your assimilation of the references I provide.

As you have pointed out, a lot depends on the electrode material. Due to the desirability of low losses and the problems associated with skin depth and proximity effects in the transmission line (flat single turn solenoid), I am going to neglect everything except highly conductive metals for the time being. Whether one looks at electrons as particles sitting in a conduction band or as a wave function, there is a physical distance between the electrode surface and cation that must be crossed in order for the electron to take up its place in an orbital of the cation being reduced, say cupric (2+) ion to cuprous (1+). In the intervening distance there are typically adsorbed species on the electrode surface and electrolyte itself. I believe that (neglecting tunneling for the moment), as an electron nears the subject cation its spin behavior is governed less by the electrode surface and more the adsorbed species and electrolyte constituents. The collective influences on the spin of an electron are often referred to as its spin hamiltonian. (For you mathematicians, "Spins in Chemistry" can be purchased for $11.01 at Amazon. I bought it for the title and price, but unfortunately most of it is beyond me. ) The spin hamiltonian of an electron that is a finite difference from the electrode surface will be dominated by the contributions of the surrounding chemical species. For example, a water molecule or a linear organic molecule bound to a gold surface by a thiol group could be interposed between the electrode surface and the cation. At some small distance from the electrode surface I believe that the electron is associated with a normally neutral species in the vicinity of the cation, thus effectively converting the neutral species to a transient radical. This transient radical and cation I consider as a "quasi-radical pair."

Since the two electrons involved have different spin hamiltonians, they will typically have different spins. However, if the elements of the two spin hamiltionians are small, and a common dominant hamiltonian element is introduced by the application of external magnetic fields, the two electrons would have essentially the same spin. Coupling by magnetic nuclei is one of the hamiltonian elements that can be modified. By substituting D2O for H2O and using sulfates composed mostly of 0-16 and S-32 unwanted hyperfine coupling could be reduced. Relaxation effects can be reduced by increased viscosity and higher molecular weights. From what I have read, spin-lattice relaxation in solutions is in the microsecond to nanosecond range, offering hope for pumping in the 100MHz to 1GHz regime. In general, I am looking for materials that will minimize unwanted hyperfine coupling and relaxation. For example, a thin gold surface might be preferable since gold atoms have a smaller nuclear magnetic moment than copper atoms.

I can't make any predictions about the results of my investigation, but I do believe that a small flat solenoid can be used to produce oscillating magnetic fields that are of the same magnitude or greater than those that have been shown to produce effects in chemical reactions. I also believe that a portion of the small flat solenoid can be masked to serve as a cathode in an electrolytic cell. An exploded view of the cell I am constructing (first parts shipped today) shows from the bottom up:

.062" FR4 base
.001" copper ground plane, could be integrated as copper-clad base
.001 PFA film--bought from McMaster-Carr with special etchant/epoxy kit
.062"cathode support with electrolyte ducts--being machined by AP Circuits
.062"chamber top

The shims for establishing the gap under the cathode and tubing connections are not shown. Anodes can be placed upstream and/or downstream. The assembled height is less than a 1/4" so that it can be fit into the air gap of a low power homemade electromagnet. The wide area of the ground plane can be used to support an additional dielectric sheet and input conductor trace along with a few capacitors and inductors for tuning. I hope this gives the general idea.

Characterization and tuning of the structure and real-time current measurement will be time consuming, but a friend has loaned me a 6GHz oscilloscope and it is amazing what you can buy off-the-shelf these days. When I told an EE friend what I was doing a couple of years ago, he told me that the cell phone industry had what I needed. I was amazed when I saw that a 6GHz amplifier (see attached) could be had for $2.00 these days. For the system I will be using a 20 watt amp with limited drive that can handle open and short loads since I am likely to have a few mistakes.

Operationally the idea is to initiate an electrolyte flow in the cell, apply sufficient voltage to observe a small reduction current, apply a high frequency current to the solenoid loop, and ramp the electromagnet current. I plan to ramp the current so that the magnetic field reaches at least 150% of that required for resonance of a free electron at the selected RF frequency. If spin locking occurs, I expect to see momentary drop in the reduction current during the scan. During the scan Zeeman splitting and resonance will occur for a lot of unpaired electrons. The question is: can the two unpaired electrons of interest be converted to the same spin at the same time? Since conductive losses in a short copper loop are smaller than heating losses due to irradiation of an electrolyte, I believe that this approach has something to offer as a research tool. If an industrial use can be found, I think that the cell is cheap enough to be used in a large-scale array.

Attachment: cell_chamber_exploded.pdf (9kB)
This file has been downloaded 587 times

Attachment: GALI-84+_mini_amplifier.pdf (135kB)
This file has been downloaded 937 times

watson.fawkes - 7-4-2009 at 08:21

@merrlin: You have two basic kinds of questions here. One is about equipment and another is about reactants. These are not independent. Given an apparatus, it will only be appropriate for certain reactants. So you can either start with an apparatus and figure out what you can do with it, or start with a selection of reactants and figure out how to build an appropriate apparatus. Given that your apparatus is in the design stage, you haven't fully committed to which of these paths you're on. Certainly the path of least resistance, though, is going to leave you with an apparatus.

Personally, I'm more able to address the reaction. Others are better able to address the apparatus. Now you named a bunch of potential reactants but no reaction. Please write down some candidate reactions that we can analyze specifically.

As for the reactants, I might recommend that you consider using fluoride ions in the electrolyte. Fluorine has only a single isotope in nature (incidentally, that's one of the reasons it's used in uranium separation). It has an unpaired 2p electron, meaning it will form electronic triplets readily. It has paired neutrons and an unpaired proton, so the nuclear spin has doublet degeneracy. In my current state of understanding, the doublets won't interfere with the effect you're looking for much.

There are some other effects you should be aware of. In the class of non-metallic electrodes, ionic conduction is an important phenomenon. Proton (or in your case, deuteron) conduction is the most common charge carrier (obviously), but oxygen ion conduction is also common. Proton conduction is particularly important for fuel cells. Now fuel cell configuration is not the electrochemical configuration that you've been looking at, but you might consider that preferential reductions of protons over deuterons in a fuel cell might be a pretty good way of making heavy water.

Another matter you hinted at but didn't address directly is that some electrochemical reactions happen with the adsorbed species themselves. In this case the electron configuration in the metal crystal electrode is a Fermi distribution, so you're not going to see triplet behavior. What you left unsaid is that you've got to suppress electrochemical reactions of the adsorbed species with the cathode, or least make sure that they're not energetically favored.

You've been assuming aqueous solutions, apparently. Have you considered other solvents? You might be able to manage the adsorption layer more readily with a different solvent. It's not even clear to me what the effect of polar vs. non-polar solvents are. Electrostatically speaking, polar species are preferred for adsorption, but that may not be true for any particular electrode-electrolyte combination.

It seems you'll also need to consider electron mobility in your adsorbed species. If you don't have high internal mobility, you'll be much more into the tunneling regime and your whole experiment will be harder. My first thought is that this speaks to consider aromatic solvents, since electron mobility in the ring is quite high.

merrlin - 7-4-2009 at 12:46

@watson.fawkes: I will be starting with a copper cathode or a gold plated copper cathode. The reduction reaction may be an electrodeposition of zinc, copper, iron or other metal from an aqueous solution. Film adhesion or continuity isn't required since the goal is to observe a change in reduction current associated with reduction. It's possible that film formation could even be a problem since there would be a transition from a heterogeneous deposition to a homogeneous deposition if the substrate and cation are not the same material. Another reaction that may be used is the partial reduction and subsequent precipitation of a cation from the exit stream by an available anion such as fluoride.

In reviewing my previous post, I realize that my statement:

"However, if the elements of the two spin hamiltonians are small, and a common dominant hamiltonian element is introduced by the application of external magnetic fields, the two electrons would have essentially the same spin."

is ambiguous. In a quantized system at any given instant, the spins are either the same or they are not. I have attached a lecture series by Professor Nicholas Turro of Columbia University that has contributed to my understanding of spin. Table 1. on page 42 shows a vector diagram representation of singlet, doublet and triplet states for electrons in an applied magnetic field. A pair of coupled electrons will be in a singlet state or a triplet state, depending upon their total spin. The singlet state and the T0 state have the same energy, with the electrons having antiparallel spins (one up, one down). The difference between the singlet state and the T0 state is a matter of phase between the two spins as they precess about the magnetic axis provided by the applied DC magnetic field. Frequency of mixing between the T0 state and the singlet state is a function of the difference in their precession frequencies. S0-T0 mixing depends upon the applied DC magnetic field and can be very fast (e.g., nanosecond timescale). In an applied DC magnetic field, the T+ and T- states are separated from the T0 state by an energy gap and a change in system energy is required for conversion to T+ or T- from T0.

Ignoring everything except the coupled electron pair and the applied DC magnetic field, there are four available states for the electron pair: S0, T0, T+, and T-. The two electrons cannot share the same orbital if they are in the T+ or T- states. If the coupled electrons are the unpaired electrons of a radical pair, the radical pair cannot react until spin conversion occurs. In an applied DC magnetic field, conversion between S0 and T0 states is typically a rapid ongoing process, and conversion between T0 and T+,T- can be induced by microwave pumping. Under ideal microwave pumping conditions, S0 -> T0 followed by induced T0 -> T+ and T0 -> T- will halt the reaction entirely. This is why Buchachenko repeatedly refers to "purification" as the theoretical upper limit on the separation factor for the microwave induced magnetic isotope effect (MIMIE). Isotope separation by centrifuge or diffusion rely on the mass isotope effect that is limited by the difference in mass of the isotopes or the compounds they are a part of and thus require a cascade system to produce a useful amount of separation. With MIMIE, a single reactor could in theory exhibit a much high separation factor.

At what point in an electrolytic reduction reaction can the cathode donated electron and an unpaired cation electron be considered a coupled electron pair? I don't know the answer to this, but I believe that if the applied DC magnetic field and the microwave pumping intensity are large enough and the other magnetic influences are minimized, that spin correlation could be achieved between an electron at the cathode surface and a cation that is some distance away. Ideally, one would want to eliminate all magnetic nuclei from the electrolyte (except perhaps the cation of interest) and have an electrode surface composed of atoms with nonmagnetic nuclei. If one uses a polar solvent that does not dissociate into species with unpaired electrons and does not contain magnetic nuclei it would further reduce undesirable influences. I'm not sure, but I think this might be why Malcolm Forbes' group at the University of North Carolina is working with supercritical carbon dioxide as a solvent. With regard to fluorine, the important thing is that notwithstanding the nuclear spin state of the fluorine atom, the nucleus has a magnetic moment that can influence the spin of nearby electrons. Broadly put, anything that has or can generate a magnetic field could influence the spin state of the coupled electron pair. Spin-orbit coupling, in which the magnetic moment due to an electron's orbital motion couples with its intrinsic spin, may also have an influence. Even for polar molecules that do not dissociate, rotational motion of the dipole can contribute to relaxation effects.

Pretty soon I am going to have take my best guess and simply try it. At this point I plan to use metal sulfates that can be dissolved to form an electrolyte from which the metal can be reduced. Zinc, iron, and copper are leading candidates. Zinc and iron are attractive since fractionation by electrodeposition has already been reported for them, and copper is attractive since it should be relatively easy to plate onto a copper cathode. I wonder if isotopically pure copper would have enough of an improvement in electrical conductivity to make the preparation worth the effort for use in integrated circuits.

Attachment: Turro_spin_lectures.pdf (604kB)
This file has been downloaded 1233 times

[Edited on 8-4-2009 by merrlin]

merrlin - 7-4-2009 at 16:27

The first actual part arrived. The top view shows the copper clad FR4 with a .100" wide cathode trace. The board on the left shows the as received board with solder coating that is standard with their basic service. The right hand board shows the solder removed. The bottom shows a .060" wide cathode trace with and without solder. One trace will be selected for use and the other removed when the cell is assembled.

Attachment: cell_board.jpg_file (99kB)
This file has been downloaded 885 times

[Edited on 8-4-2009 by merrlin]

watson.fawkes - 8-4-2009 at 05:43

Quote: Originally posted by merrlin

With regard to fluorine, the important thing is that notwithstanding the nuclear spin state of the fluorine atom, the nucleus has a magnetic moment that can influence the spin of nearby electrons.

I don't know the relevant magnitude of the energies involved. That does matter. I think I was assuming it was small, but I'm likely wrong there.

On the other hand, in a magnetic field fluorine nuclei form a spin glass that can be pumped up, eliminating most of the doublet effects. It would require a second oscillator, but since the gyromagnetic ratio of the nucleus is fixed, and since you know what your magnetic field is, the resonant frequency for pumping is determined.

watson.fawkes - 8-4-2009 at 06:53

Quote: Originally posted by merrlin

At what point in an electrolytic reduction reaction can the cathode donated electron and an unpaired cation electron be considered a coupled electron pair? I don't know the answer to this, but I believe that if the applied DC magnetic field and the microwave pumping intensity are large enough and the other magnetic influences are minimized, that spin correlation could be achieved between an electron at the cathode surface and a cation that is some distance away.

My guess is that it's when they're within reaction distance, not farther. The whole point is to suppress reaction by putting the electrons in a state that forbids reaction. So if they're not close enough to react, the reaction is forbidden because of distance and spin coupling is irrelevant.

The problem you've got is that you need the cathode-donated electron to be bound into an orbital state with angular momentum (that is, not an 's' state) so that it's got the possibility of forming triplets. I may be wrong on this, but I don't see how an unbound electron form a triplet. Electrons inside the metallic electrode are bound, but they're not in orbital states (at least not the conduction electrons) because of the crystal lattice. So you need to rely upon whatever is adsorbed onto the electrode surface to bind the electron. Ideally, that adsorbed material should completely coat the cathode to prevent direct reduction of the metal ions.

In your proposed reaction, you've got sulfate, metal ions, a metallic cathode, and water. The negatively charge species, sulfate and OH(-) will flee away from the cathode because of the electric field. The only adsorbed species are ions, H3O(+), and neutral water. In this setup, I don't see a situation that will lead to forbidden triplets.

On the other hand, you report this:

Quote:

Zinc and iron are attractive since fractionation by electrodeposition has already been reported for them

Could you post these papers? I'd be interested in seeing the details of their apparatus and electrolyte. Either I'm missing something, got something wrong, or don't know some details about what the situation is.

merrlin - 8-4-2009 at 11:20

Quote: Originally posted by watson.fawkes

Quote: Originally posted by merrlin

With regard to fluorine, the important thing is that notwithstanding the nuclear spin state of the fluorine atom, the nucleus has a magnetic moment that can influence the spin of nearby electrons.

I don't know the relevant magnitude of the energies involved. That does matter. I think I was assuming it was small, but I'm likely wrong there.

It would require a second oscillator, but since the gyromagnetic ratio of the nucleus is fixed, and since you know what your magnetic field is, the resonant frequency for pumping is determined.

You've anticipated one of the approaches that are available for pumping. For efficiency's sake, I am planning to tune the physical circuit structure to the resonant frequency being used for magnetic pumping. A single frequency and two entirely different resonant phenomena! However, a wideband structure could be simultaneously pumped at more than one frequency. Buchachenko has referred to the desirability of simultaneously pumping magnetic and non-magnetic species, assuming that their resonance frequencies are sufficiently separated.

I contacted an expert in spin chemistry last June and he said in brief that it was interesting and would require more thought. Since he was out of the country and also working on a textbook about spin chemistry, I had not heard from him since August, until today. He has doubts about overcoming the relaxation effects I've referred to in my previous posts:

"The idea you propose is interesting but I think you will always be fighting some rather efficient electron spin relaxation processes that can eliminate the coherence that you need on a very fast time scale, perhaps as fast as 100 ps."

The encouraging thing is that he has not rejected it out of hand, nor has he cited a fundamental flaw that dooms the approach. He has also indicated that in spite of his busy schedule he is open to further discussion. The upper bound he cites for relaxation processes (100 ps) equates to 10 GHz and is quite daunting, but by periodic cathode masking to avoid standing wave nodes and/or replacing the transmission line short by a matched termination, I think the operating frequency could be extended to several GHz for research purposes. Solute and solvent selection offer a lot of potential for shifting the relaxation times towards the microsecond end of the relaxation timescale.

I stumbled on to this approach in part because of my employment background involving microwave missile components and research and development in electrochemical energy storage devices (ultracapacitors) for implantable defibrillators. I've come up with an apparently novel approach, but there are a lot of people more capable than I am of carrying it forward. I am also sure that there are people on this board that possess knowledge and abilities that I do not. Electrochemically, this is a vast unexplored territory that I believe is within the means of the individual explorer. I used to do sideline research on high pressures and temperatures using a 10kV 200 microfarad capacitor that sits in my garage. This is a lot less dangerous.

merrlin - 8-4-2009 at 15:02

Quote:

Originally posted by watson.fawkes
Could you post these papers? I'd be interested in seeing the details of their apparatus and electrolyte.

The two papers concerning iron and zinc fractionation during electrodeposition are below. From what little I know about spintronics, it seems that the spin of free electrons in solid conductors can be manipulated by external fields, so I guess that the task at hand is to induce the same spin state in the free electron and unpaired electron. In simplest (too simple?) terms it would seem to be a matter of keeping them both "up" or both "down" by appropriate magnetic pumping in an applied DC field.

Attachment: Redox-driven stable isotope fractionation in transition metals--Application to Zn electroplating.pdf (375kB)
This file has been downloaded 640 times

Attachment: The isotopic effects of electron transfer-An explanation for Fe isotope fractionation in nature.pdf (355kB)
This file has been downloaded 626 times

[Edited on 8-4-2009 by merrlin]

watson.fawkes - 8-4-2009 at 17:34

Quote: Originally posted by merrlin

The upper bound he cites for relaxation processes (100 ps) equates to 10 GHz and is quite daunting [...] Solute and solvent selection offer a lot of potential for shifting the relaxation times towards the microsecond end of the relaxation timescale.

That 100 ps time shouldn't be taken as corresponding specific frequency. You're looking at population dynamics. Essentially, you've got two competing rates. One is a decay rate going from a target state to a non-target one; this is the relaxation rate. The other is going from non-target to target; this is your pumping rate. Since these are species conversions, the underlying math starts off similar to radioactive decay. In this case, however, you've got one species decaying to the other, and vice-versa. It's pretty straightforward to compute, given the two figures for half-life, what the steady-state populations are.

I'd recommend doing this computation, because it will give you some good intuition about the effect of these manipulating these rates. I would not assume, though, that these rates have the same mathematical form. The pumping rate can be assumed to be random, so that one's an exponential. I can't say offhand, though, whether the decay process looks like a similar exponential or whether it can be assumed to be a roughly constant-time process (or something else).

The principal means that you can decrease the pumping half-life is to increase the intensity of the pumping radiation. After that, I don't see immediately how to increase the absorption cross-section, although that might be possible. radiation. Perhaps some modulation of the magnetic field would make the species to be pumped appear "blacker".

merrlin - 8-4-2009 at 18:30

@watson.fawkes: You're right about looking at this as a problem in statistics. In spin chemistry there two relaxation times T1 and T2 that are associated with spin-spin relaxation and spin-lattice relaxation respectively. Quoting Buchachenko's book:

"It is worth remembering that the transitions between electron spin states (and, consequently, between Zeeman sublevels) are responsible for electron spin resonance (ESR), while the transitions between the nuclear spin states for the basis for nuclear magnetic resonance (NMR). These transitions are induced by a microwave magnetic field at the frequency of electron or nuclear precession and are accompanied by spin projection changes. The magnetic field may be applied externally, as in ESR or NMR, or be produced by the lattice stochastic molecular motion, which produces the fluctuating, noisy magnetic fields of different frequencies and amplitudes. The spectral component of this noise at the frequency of electron or nuclear precession induces the spin-lattice or spin-spin relaxation (i.e., transitions between spin states)."

Increasing the intensity of the magnetic field is definitely an advantage. This is one of the reasons that I believe radiationless pumping will be an improvement. When working with a solution containing charged particles or dipoles, interactions with the electric field component of an electromagnetic wave is going to increase the noise level without contributing to the magnetic pumping. Heating alone can prevent CW operation. By using the magnetic field induced by a microwave current in a single-turn solenoid or short transmission line segment, a locally intense magnetic field can be produced without an appreciable electric field being produced in the electrolyte. Because of skin depth losses at higher frequencies, there is probably some kind of trade-off between high frequency/low intensity pumping and low frequency/high intensity pumping.