url: http://escribe.com/health/thesilverlist/m61358.html Re: CS> Requests For Lab Test on ULVDC CS From: Jason Eaton Date: Wed, 23 Jul 2003 00:21:00
Hi Jason, Great Post! I've trimmed to save space. > Hi Mike: > Great Post! > Unfortunately, Faraday's Equation doesn't always equate properly > with modern CS production... > I believe with low current devices, Faraday's equation is close to > accurate. There are fundamentally two problems with using > Faraday's equation to calculate PPM: > 1. A controlled-potential coulometric analysis is required in > order to obtain accurate readings.... The current must be > absolutely controlled. This can be done, I discovered quite a few > months ago, with a potentiostat... But then again, you and Bob are > the engineers, so: Engineer. According to the information below, > 100% efficiency is required with the current. > http://chem.ch.huji.ac.il/~eugeniik/instruments/electrochemical/coulometric_titrators.htm Thanks for the trip down nostalgia lane. Those are very good examples of the construction style in the 50's, and I was a teenager when those were designed. You can do the same thing now with a simple FET and opamp, and get pretty good accuracy over a large range. You are also correct, a simple resistor can provide pretty good current stabilization. It's easy to monitor the current at intervals then toss the data into WPlot. It will integrate the values and give the average current. Then put that into Mercury and you're done. I think we need to talk a little about precision and accuracy. Let me introduce you to the wonderful world of electronics. I have a frequency counter/timer on my bench with 12 digits of resolution. I bought it new in 1982, and later models have even higher resolution. I also have a dvm on the bench with 7 digits of resolution. It was designed around the same time. Newer models have higher resolution. I often have to take the thermal contact potential with dissimilar metals into account in my designs and during measurements. These are in the microvolt region - a million times lower than an ordinary 9V battery. You can buy a dvm from harbor freight for under $10.00. DVM's usually have tolerance spec's in the 0.1% to 0.01% range. In electronics, we are used to working with levels of precision that are unheard of in chemistry. For example, I see Ken's jar has the tolerance of +/-2% marked on the glass. But that is only available when the table surface is perfectly flat and horizontal. Any tilt will cause a cosine error in the measured volume, but I never see this mentioned anywhere. Temperature will change the measured volume just like in a thermometer. But I rarely see any site that is concerned enough to take that into account. There are so many error sources in chemistry, I'm amazed that anything gets produced. I really had to readjust my thinking down several orders of magnitude when I started working with cs. It took a while, but I'm starting to become comfortable with it. You often see displays or printouts with lots of digits. These give the impression the instrument is very accurate. However, many of the least significant digits may be of little value. In engineering, we are required to carry all the digits during calculations, and suppress the unusable digits in the final report. Many of the instruments made by Hewlett-Packard do this automatically, but if you collect the data over the HPIB bus, they often send all the digits and you have to figure out yourself which ones to suppress. Unfortunately, in chemistry, there is often no way to tell which digits should be suppressed. There is always noise in the readings. Averaging can be used to reduce the noise if it is random and uncorrelated. In this case, the improvement in the signal-to-noise ratio (snr) is proportional to the square root of the number of samples. The reason is the noise adds orthogonally, and the signal adds linearly. The disadvantage of using averaging is to improve the SNR is you have to double the number of samples to get a small increase. For example, averaging 100 samples will give a factor of ten improvement in standard deviation. To improve the snr by 3 dB, or a factor of 0.707, now takes 200 averages. The problem is this takes twice as long, and may not be practical due to time constraints or other factors. A really big problem occurs if the measurement drifts during the measurements. I know of no easy way to detect this in the data, but now the data is corrupted and you may not know anything has happened. You have to go through the data in blocks and calculate the mean, and see if there is a detectable drift. The dilemma is the mean change can be perfectly legtimate with random noise. Another problem with averaging is how to handle outliers. These are data samples that are far from the standard deviation. They can be legitimate, or they may be a fluke error such as a typo when transcribing the data. How do you tell? The topic of noise is a huge, ongoing problem in every field, and takes specialized knowledge of the process to begin to understand. So we really can't tell how precise a single reading is. We need to take many measurements, and get the mean and standard deviation, and prove there was no drift during the measurements. Now we that have the mean and standard deviation, we can compare the instrument to some other standard, with it's own mean and standard deviation. This will tell us how accurate our instrument is. But I think the accuracy in chemistry is not as good as may be claimed. I have no way to determine the magnitude of the error, but I suspect it can be very large in some cases. I know it is in other cases. > 2. I had a similiar reference which I cannot at the moment find > that indicated higher currents with silver could not be measured > accurately using faraday's equation... Not that faraday's law is > innacurate, it's just that Faraday didn't say that the silver > would stay in the distilled water. Even a small film on the > surface of the one of the electrodes, undetectable by the eye, can > throw the equation off by a level not acceptable... Also, readings > correlated between faraday's equation and photospectronomy showed > unexplainable anomalies in the end readings which I certainly > couldn't account for without both TEM and AAS analysis. Don't forget the salt test:) Yes, there are many anomolies we need to track down. Some may be instrument noise, some may be ordinary process noise, and some may be early signs of interesting phenomenon that we need to investigate. Basically, we can track the number of silver ions liberated during the process, then measure the ion concentration at the end. The difference is silver oxide, which we want to minimize. NIST recognises the Faraday constant in the International System of Units, and mentions that silver electrolysis was accurate enough to help detect an error in Millikan's value of the charge on the electron: "The Faraday (F) was determined by measuring the mass of material (e.g., silver) electrolytically deposited onto an electrode when a known current flowing for a known time was allowed to pass through a solution containing the material. The indirect value of the elementary charge (e) deduced in this way was (4.8021 ± 0.0009) x 10-10 esu, significantly different from the Millikan value. The major source of this disturbing discrepancy was traced in the latter part of the 1930s to the use by Millikan of an incorrect value for the viscosity of air. Millikan had taken a value that was almost entirely based on a measurement by one of his students; but it was later shown that the student had made a rather subtle experimental error. When Millikan's data were reevaluated with a correctly determined value for the viscosity of air, the value of e obtained agreed with the indirect value calculated from the Faraday and the Avogadro constant." http://physics.nist.gov/cuu/Constants/historical1.html The value for the Faraday constant is known to 7 significant digits, 9.648531E+04, so the process is probably accurate enough for our needed: http://physics.nist.gov/Pubs/SP811/appenB9.html I think we can have good confidence in the Faraday equation as long as the anode reaction is silver-related and there is no bubbling at the anode. > Never-the-less, I believe the law is a great tool, especially with > highly controlled setups and low current. Yes, I think the Faraday equation can be one of the most valuable tools for investigating the cs process. But as in any experimantal system, we have to watch carefully for errors and flukes, and handle them properly. Some may be early signs of very interesting things, and may guide us to new information. > As far as the relationship between voltage and current, perhaps > someone more versed could explain how increased voltage, even with > the same current, produces a different result. > http://www.silvermedicine.org/robertobecker.html > However, if you notice the diagram referencing the lines of > voltage and lines of current, you'll note that a FIELD is actually > established. Voltage is still present between the anode and > cathode. Don't get too hung up on this. The picture is a standard description of fields in basic electrostatics. Yes, a field exists between the electrodes. From our point of view, the only thing we can see is the voltage across the cell and the current through the wires. This looks like a simple resistor, and E = I * R. But what the picture doesn't show is the close-in field that surrounds the electrode. I have discovered a way to make this visible, and will post as soon as I have the time to finish the description. > You state that particles don't begin to form with most generators > until about 10 PPM. I don't agree with this assessment. Silver > particles begin to form very early on in the CS production > processes I've observed... I suppose if the current is extremely > low, this might be different. If you look at any of the EIS AAS > measurements out there for products below 10 PPM, you'll see that > most have at least 1% particles, but usually higher. I can see the > particles with a laser pen early on in most processes. Remember, > just because the "ion cloud" or stream is not visible to the eye, > does not mean that it does not exist. Thermal stirring, mechanical > stirring, and/or water circulation assists in reducing early > agglomeration. Yes, You Are Absolutely Right! I was very careless in that statement, and was thinking more of the process that leads to the black film on the electrodes and the side of the glass. I believe some silver oxide is produced as soon as the ion clouds surrounding the electrodes contain both silver and hydroxyl ions. The amount of silver oxide increases when the ion clouds are more concentrated. This occurs at high current density (ie, current / wetted area). It also increases when the ppm increases. Eventually, this leads to the black deposit, and misting starts as the process continues. We start noticing it around 10 ppm with the 3 nines. > In fact, if we want to get into advanced CS production, my opinion > based on speaking with every top CS producer that I have been able > to find in the world: > We'll never be able to accomplish it with batch processing. There > are too many variables involved, and the environment cannot be > completely controlled. Lab quality water purifiers must be hooked > up in-line; Argon gas pumped is used to prevent contamination with > air. I believe Ivan is currently using Argon with his setup. We > need a way to get the ions and minute particles away from the > current, and I'm not certain if it is enough to use a large > brewing container with slow water circulation, or if the desired > concentration needs to be acheived rapidly with the EIS then > pumped out immediately into a sealed container. I guess it depends on what you are trying to achieve. This might produce high ppm cs, but I'm not sure it would be stable, or even if we really want it. There is a good chance it would kill all the bacteria in the tummy, and make a person very ill. I think 20 ppm is a good compromise, and 30 ppm or more might be used in emergencies. But the basic idea is to prevent infection, or head it off early if it starts. 20 ppm seems to do that quite well. > If I could build the perfect generator the rods would not be rods > nor plates. Both are problematic ( and yes I like Ken's new > silverpuppy solution as well ). Both the anode and cathode would > be spherical: A ball within a ball. This is the only possible way > to completely control the current draw. Water circulation would be > used, counterclockwise in North America, but preferrably outside > the Earth's Magnetic field using reverse flow forms-- and > certainly with gravity nullified. I would prefer to use phase > lasers, when they are eventually invented, in-line, to prevent > agglomeration. Unfortunately, it seems, I read too much Isaac > Asimov growing up! Silver sphere-within-a-sphere. Man, that has great biblical connotations. You are going to have to keep your marketing guy on a very short leash. Even then, I would expect a significantly greater market penetration in the midwest than normal:) One tiny problem is you need to have two holes for water to circulate and to support the inner sphere. The edge effects will get you. It might be possible to flare the hole to reduce the discontinuity, just like the rounded caps on high voltage Van de Graff generators but in the opposite direction. The other problem is of great interest to me. I don't know what happened to the silver ions when Robert used parallel plates. There was apparently no black deposits until the third run. Where did the ions go? > Best Regards, > Jason Thanks, Jason. Great to talk with you - you have very stimulating ideas. But I gotta start making shorter posts. These take way too long to write, and are way too long for anyone to read:) Best Regards, Mike Monett -- The silver-list is a moderated forum for discussion of colloidal silver. Instructions for unsubscribing may be found at: http://silverlist.org To post, address your message to: [email protected] Silver-list archive: http://escribe.com/health/thesilverlist/index.html List maintainer: Mike Devour <[email protected]>
