The Water Fuel Cell

[eclipsed78]

Hi,
I had my 5mm spaced SS304 single blank light switch covers tested on a LCR meter with liquid tap water and liquid DI water....I was interested in first if the dielectric was voltage independant....also what the difference of tap water and DI water would be...so I had them tested at different voltage bias and with tap water and DI water...


-----------------------------Edit---------------------

I previously released LCR meter measurements that had calculations of DF for liquid DI water that were incorrect.....although the calculation of liquid tap water were correct, i mistakenly copied the equation down the sheet without taking into acount pF instead of nF....

I have loaded a new file that shows the correct calculations for DF.....

What this data and calculations show is that the DF of liquid water is very high and it will never resonate alone up to 50000Hz...with either liquid tap water or liquid DI water....

After further review of the data, it may seem as if capacitance increases from frequencies of lower than 5000hz....although Im not sure if this is due to such a high disapation factor or due to liquid water's dielectric constant being frequency dependent at frequencies lower than 5000hz...

-------------------------------------------------------

[kaosad]

eclipsed78,

Good experiment and keep up.

Did you measure parasitic resistance and inductance of the plates in the water? If yes, how did you measure them?

[eclipsed78]

Hi,
The measurements were done in the parallel setting..meaning the capacitor's two wires were directly connected to the LCR meter probes, and not to an external ground, which would put the capacitor in series..but im not sure because I was not the one the measured it, someone that does it professionally measured it and gave me the data.....the capacitance speadsheet shows what was measured and what was calculated....Rp and Rs are shown on the spreadsheet...but I did not have its inductance measured...as I thought this would be so low that it would be unmeasurable...

[eclipsed78]

Hi
I have had a cell tested on the same LCR meter that was coated with super corona dope

5mm spaced SS304 SBLSC with 3 to 4 coats of Super Corona Dope on both sides (non heat treated)

__________DI water:_________________________Tap water:__________________
__________Cp (nF)_______Rp (K ohms)__________Cp (nF)_______Rp (K ohms)
100hz_____24.2__________26.1_________________205.2_________4.5
1000hz____10.2__________14.0_________________44.5__________2.6
10000hz___3.8___________4.3__________________14.9__________1.0
20000hz___2.6___________2.86_________________10.5__________0.743

I have ordered a liter and am looking for a toaster oven for more coatings and to bake them to make the dielectric double in value....I am looking for higher Rp somewhere in the 5 to 10 Meg ohms...Im thinking that the plates can be in series but the capacitance needs to be at least 10 to 50 nF.....so I need to do more experimenting with the coatings and heat....I will throughly document how the plates were made..

[eclipsed78]

Hi all,

I have come to the conclusion that only one plate needs to be coated to increase the parallel resistance of the capacitor....and to allow the voltage to affect the water...

I tried dipping a plate with more super corona dope....and making the coat more thicker, and I tried baking it at 300 F (Let me warn you that you do not want to be smelling this stuff...it stinks...in fact the MSDS says that inhalation of 5000 ppm/4 hours is lethal)....but we are looking for Mega ohms not kilo ohms, and after a couple of coatings and letting the one dry..I tried baking it at 300 F...and the coating on the plate started to boil....which completely ruined the plate....

now this would be a very long, smelly, and tedious process if this was able to work, but I thought, there has to be a better way...and after searching the net, I found that plexiglas, or acrylic sheets could be used in high voltage capacitor, and the hardware store has this sheeting for relatively cheap...and I already had some so I made a cell...and I had it measured on the LCR meter today....

Multidielectric capacitors add individual capacitance of each dielectric as if they were is series....

2.032 mm plexiglas and 3.302 mm water gap and 4 nylon spacer nuts

total plate distance of 5.334 mm

DI water @ 60 hz
Cp = 178 pF
Rp = 5.4 MegOhms

Tap water @ 60 hz
Cp = 192 pF
Rp = 1.2 MegOhms

and I had it measured at 100hz, 1khz, 10khz, and 20khz, and the Cp and Rp on both DI and tap water went down as frequecy went up...

remember, the reason why the capacitance is so low is that the capacitance is added like parallel resistors, so the lowest dielectric (the plexiglas) limits the parallel capacitance.......

The point to using plexiglas is to increase the Rp, and I double checked the cell, and it was the silicon trapping water....so I plan to make a second plexiglas cell and just melt the edges with a torch to seal one of the plates in the plexiglas 3 layer sandwitch...because if any water can make a connection with the inclosed plate, the parallel resistance will go down....but if this works, the cell should be almost like a open circuit....as the LCR meter would not be able to measure a theoretical resistance value of 190 Tetra ohms, but with a Cp of at least 150 pf, which can then be put in parallel with other plexiglas cells to make a required Cp...

[eclipsed78]

Hi,

I previously made an acrylic encased SS304 SBLSC Cell but it leaked, but this cell did not leak as the melting of the acrylic material to seal the cell within the 3 layer sheeting worked very well....one side enclosed in acrylic material, and the other side open to an enviornment between the plates..I took some pictures of the cell and I had it tested on the same LCR meter.....

After having this cell tested without water leaking into the encased plate, I started to do research on the total capacitor losses and found that there are three components....all together they are known as ESR...or equivilant series resistance......I have included the paper that explains ESR below along with the measurements, and pictures....

[eclipsed78]

Hi,
It turns out that delrin is even better than teflon as it has a better dielectric constant, it has better dielectric strength, and it has a low enough DF even at 1Mhz which is much better than acrylic to allow low dielectric loss...delrin sheeting is almost as cheap as acrylic sheeting, and I will be making a third cell soon when I am able to with delrin....and I will post the LCR measurements....

I included the three material's physical properties datasheets from one distributor that I checked on pricing....they show the electrical properties of PTFE material (teflon), Acetal material (Delrin), and Acrylic material (plexiglass)....

[eclipsed78]

Hi,

I found an excellent reference to DF, dielectric strength, and dielectric constant...I dont know how accurate it is but it seems to include references...so I made sure I have the data on file before it is taken away...interestingly, it shows POM having a DF of .005 at 50 Hz....so this would be about 10 times less DF than acrylic....

http://www.pupman.com/listarchives/2003/March/msg00106.html

[eclipsed78]

Hi All,
I have made an acetal (delrin) enclosed cell...the DF of the cell was so low that most of the tap water Rp measurements were above the 100 Megaohms maximum resistance capability of the LCR meter.....whats interesting is that the DI water Rp measurements were less than tap water's....but all of my previous measurements have shown that tap water's Rp is much lower than DI water's Rp at the same frequency...

Ive included the LCR measurements, the datasheet of the specific material that encloses one of the plates, and some pictures of the acetal cell and the disasembled acrylic cell which is the same design as the acetal cell....

[eclipsed78]

Hi All,
I have been doing some calculations for multidielectric capacitors and found that acetal will not be able to focus enough of the potential field to breakdown the water with an assumed 80 V/mil dielectric breakdown, no matter how the cell is designed.....

Here are the equations that both need to be satisfied:

E1/E2 = e2/e1

where:

E = potenial field

e = permittivity

and

V = E1*d1 + E2*d2

where:

V = voltage

d = distance

Although I have not experimentally confirmed this, I will be doing experiments to verify my hypothesis...

I have researched some materials of interest:

PVDF
PEN
PI

[dankie]

So based on your calculations .

Are you saying a dielectric coating on one of the negative tube , whatever that material , will not work.

How could have Stan split the water using tap water alone , it cannot hold a charge . Was he opposing amp flow somehow with the negative coil alone ? I cant see this happening ...

[warj1990]

Eclipsed78,

The last statement indicates you are working in similar fashion to my project on the Tay Hann Hee patent. (Statement being “multidielectric capacitors”).

I will not repeat everything that has been done on that project as you can read it at your own leisure.
http://waterfuelcell.org/phpBB2/viewtopic.php?t=972

Baised on this patent:
Liquid water has a volts/mil rating of 560.
It is necessary to induce 22,400,000 volts / meter to split water.

Utilizing series cap. Formula I am unable to achieve that amount of voltage field on liquid water – similar to what you just found out.

Using TiO2 (dielectric constant of 100 and volts/mill of 101) and steam (dielectric constant of 1.00xx )
The equation can be satisfied. My experiments pertain to 2 insulated plates – not just one.

Generally when baking in the oven don’t exceed 225* F.

You have documented your work much more than I have mine – I encourage you to continue this project regardless of mine.

P.S. Normally I would send a P.M. with this info - but I think the mail server went belly up.

Warj1990

[eclipsed78]

Hi dankie,
I am only saying specifically that acetal will not be able to be designed in a way that focuses enough of the potential field on the liquid water using an assumed 80 V/mill but it will defenatly work with steam if it is possible......using the above equations, I was able to calculate that the distance ratio of acetal to liquid water would be a negative value, which would mean it was a non real answer, so acetal will not be able to be used with liquid water..

You are correct to say that water will not hold a charge becuase its dielectric loss is too high, and therefore will not resonate....as the LCR meter measurements clearly show...that fact is the reason why I am researching multidielectric capacitors...acetal has very little dielectric loss, but its dielectric strength and dielectric constant are too low to do what is needed..


Hi warj1990,

The Tay Hann Hee patent is somewhat close to the water fuel cell except it does not emply the resonant circuit...I believe that patent also used ceramic and water as the multidielectric capacitor, while the water fuel cell, I believe, used acetal and steam...which could easily be able to be designed...but I am trying to figure out if I can make a multidielectric capacitor with liquid water...

One of the early patents under that water fuel cell shows a picture of a multidielectric layer, and it is refered to as the "electron inhibitor"....shown in the picture below....

Now, 569 V/mil is an interesting measurement done by Tay Hann Hee, but im not sure if that was ever tested...or how it was specifically tested because the patent never tells you what material he uses...

I made my calculations with 80 V/mil by this reference:

http://www.rfcafe.com/references/electrical/dielectric-constants-strengths.htm

But Im not sure where this site got this information....or why I thought it was right...

I do know that sandi national labs in NM has published a paper on water dielectric breakdown.....basially, Im still trying to fully understand the wealth of information this paper conveys...Im not sure, but I think this paper shows an equation of the dielectric breakdown of water related to pulse time...or frequency if you are using a resonant circuit....I have posted it below...

Although my calculations were for 80 V/mil and I should have researched this value more, any increase in dielectric strength will further make it more imposible to use liquid water and acetal for the purpose of water dielectrical breakdown....this would be true as well as with a pure TiO2 solid ceramic layer and water as a multidielectric capacitor.....if water's dielectric breakdown is really even half of what you quoted, there may not be any material that would be able to be designed to work with liquid water...

But acetal and TiO2 will work fine with steam...I think acetal is much cheaper...but I think TiO2 has some photocatalystic properties that migh work well with the 935 nm GaAs IR LASER diodes...

I did check how much to buy a vacuum furnace (2000 C) that can melt TiO2 into a sheet and its about $20K....I also checked how much a custom job would be and they said about $5K for the high temperatures and setup using their vacuum furnace....I even found a manufaturer that has pure solid TiO2 ceramic material called "Teraglide tgk" but they are not interested in small quantity orders, nor are distributors interested in small quantity orders of pure TiO2 ceramic tube theardguides...

but the only reason to use the multidielectric capacitor is to make a resonant circuit and disassociate water by means of dielectric breakdown, so if steam is used under pressure, it may have a lower dielectric loss and steam may resonate without the need of the multidielectric capacitor...and simply allow the WFC SS304 injectors...

Now, series capacitors are not the same as multidielectric capacitors...I have made electrical schematics that show the difference below.....as series capacitors will not resonate if one capacitor's dielectric has high dielectric loss....

Lastly, I have read your experiments, but you are not using pure solid TiO2 ceramic.....so unfortunatly, that cell will be conductive and can not be considered to be a good insulator when it is conductive...What is very unfortunate is that the only way to make the TiO2 a ceramic is by sintering it at very high temperatures (1830 C)...I understand that you are trying to take the water out of the powder, but the powder is not stable enough to hold together...If you start to use anything that will hold the powder or some type of filler, then that container or filler is now part of the circuit too and has to be put within the equations....and there is where it starts to get complex...

I am thinking that if the paper does infact predict the dielectric breakdown of water in relation to frequency, then it could still be possible that some type of material could be used on liquid water if its dielectric breakdown is modeled to be at much lower at lower frequencies...

If the dielectric strength of water is in fact frequency dependant but has a lower limit of 80 V/mil at lower frequencies, then I think the three materials of interest that I listed would still possibly work with liquid water...

I think PVDF is the least expensive but the disapation factor as high as acrylic's disapation factor at 1/8 thick sheets and will not resonate....but It will resonate with 0.003'' PVDF films as its disapation factor is much lower....What I am not sure about though is that there is a paper that states PVDF's dielectric constant is frequency and voltage dependent...(It also inculdes other materials that are within the same class of materials)...if the dielectric constant drops due to higher voltage, Im hoping that there is a lower limit of at least 8.5 as this is where I calculated the 1/8 width sheeting and this would still work.....here is datashee for PVDF as a film...the paper about a variable dielectric constant of PVDF..and the calculations done at 80 V/mil dielectric breakdown of water with the PVDF film multidielectric capacitor...

[iterator]

your work here is excellent. Something I have been trying to figure out is how these materials behave with pulsed voltages rather than continuous. As most materials can withstand higher voltages if they are only exposed to them for short bursts.

Another note: If the binder you are using for your ceramics is an insulator of higher value dielectric strength, your ceramic should not become a conductor at all. This does reduce the dielectric constant of the ceramic in question however.

[eclipsed78]

Hi iterator,

Thank you for your thoughtful consideration....

I think that published scientific journal actually does indicate that water's dielectric strength is in fact frequency dependant and basically shows that as frequency decreases, water's dielectric strength decreases....Im still unsure of how their equation is modeled though with frequency as it is used with pulsed time in microseconds....

I agree with the the fact that, if a binder of high dielectric strength and constant is used, then the total strength and constant could be still be relativly high, but Im not sure how that would be calculated...basically, if dielectric breakdown of the material is possible, the multidielectric solid layer must be able to focus enough energy to the water.....if the filler somehow lowers the constant and strength by too much, then the water will not have enough of the potential field to break down..and there would be no way of knowing until TiO2 and the filler was tested with, I would think, a very little water gap....but assuming a 80 V/mil at lower frequencies, Im not sure if that could be designed in order to test this hypothesis....it would be only trial and error, which really would be time consuming...I believe once the dielectric breakdown of water is tested at 1000 Hz, then it might be possible to for later complex design of the multidielectric layer, because TiO2 does have photocatalystic properties that should be further researched with this process....

One last thing about the above calcualtion with PVDF, if the cell is not used at the maximum voltage at which the cell can withstand then the following is used as the voltage of the cell:

(x)*[(75984.1 V/mm)*(d1)]= 3149.6063 V/mm * (7.51*(d1) - d2)

where x is the percentage of the maximum voltage the cell is to be used at....

Basically, this would lower the ratio of water to PVDF, which would make sence because there would be less water to break down...and allow a safe operational voltage level of the cell....