Hello Oscar Marin-Flores ,

One possibilitiy is electrical leakage through surface conductivity. In other words, a sample usually exhibits two types of electrical conductivity: 1) bulk conductivity through its volume, and 2) surface conductivity over its outside surface. Even if a substance is a good electrical insulator, its surface can be a source of high electrical conductivity. One way to distinguish between surface conductivity versus bulk conductivity is by using the guard electrode technique, see the attached picture.

With respect to piezoelectricity, quartz crystals are highly piezoelectric, but fused quartz is much less so.

Regards,

Tom Cuff

At 700°C, ordinary, soda-lime glasses have a fairly high electrical conductivity, which is associated with the movement of sodium ions through a viscous melt. Quartz glass does not soften at 700°C, but if there is an impurity of sodium oxide in it, then sodium ions can provide a small electrical conductivity even at such relatively low temperatures.

Hello,

As a mineralogist, I have a small note. Fused quartz is not quartz, but quartz glass. For glass, a piezo effect is not expected, resp. is not possible because it requires a strict, noncentrosymetric atomic arrangement. In addition, applying a charge to the quartz would cause an inverted piezo effect and the crystal would deform. Mutual compensation would occur on the polycrystalline aggregate.The melting point of quartz is about 1670oC.

Hello Oscar Marin-Flores ,

I agree completely with Dalibor Matýsek . It is also possible that your fused quartz has excess conductivity due to impurities as alluded to by Vadim Verlotski , but these impurities would depend on how the vitreous silica, which you are using, was manufactured. Note, the term 'fused quartz' is currently discouraged because people sometimes believe that fused quartz, which is a glass as mentioned by Dalibor Matýsek , is somehow crystalline. There are two broad categories of vitreous silica: 1) natural silica or fused quartz, made by melting natural quartz crystals; and 2) synthetic quartz, made by chemical synthesis using a gas phase reaction between O2 and SiCl4. There are four types of synthetic quartz depending on the type of flame used in the gas phase reaction, with the main impurity being hydroxyl. For further discussion, see the following book,

James E. Shelby; Introduction to Glass Science and Technology, 3rd Ed.; Royal Society of Chemistry; 2021; pp. 269-270.

I suppose the real question is, what kind of vitreous silica are you using? I mean, it is also possible that you are using some type of devitrified silica, which might be polycrystalline, for example. In a polycrystalline silica, electrical conductivity would increase at the grain boundaries.

Regards,

Tom Cuff

Thank you all for your insightful comments. As indicated by the company website, https://www.qsiquartz.com/technical-information/, the purity of the fused quartz is about 99.995%. In addition, the volume resistivity is said to be 10^19 ohm-cm at 20C and 10^7 ohm-cm at 700C. This significant decrease of 12 orders of magnitude in the specific resistance suggests that the leak might take place through the bulk of the insulator. The fact that part of the insulator is exposed at room temperature (~25C) to protect the connections appears to confirm this.

Hello Oscar Marin-Flores ,

It is not clear to me from your URL how the specific resistance was measured by the manufacturer, i.e., whether or not it was measured using a guarded electrode configuration. Re your statement that, "The fact that part of the insulator is exposed at room temperature (~25C) to protect the connections appears to confirm this." If the electrical connections penetrate all the way through the fused quartz from the 25ºC side to the 700ºC side, then the electrical connections emerge from the fused quartz on the hot side, and the electrical leakage could still occur across this surface. If the electrical connections do not penetrate all the way through the fused quartz to the hot side, then there is still the possibility that the current flows along the longer path over the surface rather than the shorter path between the connections.

Anyway, it is not clear to me that surface conductivity can be ruled out. Do you have a picture of your arrangement that you can share with us?

Regards,

Tom Cuff

Hello Thomas,

Attached is a schematics of the system. The inner electrode is simply a silver wire while the outer one is a silver coil around the quartz tube. The current obtained depends on the gas running through the tube, being always higher with Ar as compared to the one obtained with other gases such as H2 or air. I suspect the gas is being ionized, producing a current that can travel through the "conductive" fused quartz.

Hello Oscar Marin-Flores ,

Thank you for the diagram. What are the details of the T-shaped fitting at the top of the diagram? For example, is the fitting made of metal or plastic? How does the silver wire penetrate the fitting? Have you tried moving the ammeter to the negative side, and placing a second silver coil on the outside of the fused silica tube also attached to the negative side of the power supply but before the ammeter? In other words, have you tried to make a guarded current measurement? When I wake up, later, I will upload a modified drawing showing what I mean.

There seem to be a number of places in your setup that might contribute to a measurement artifact.

Regards,

Tom

Thomas, the fittings are made of teflon so there is no contact of the electrodes with metallic surfaces whatsoever. The fact that the current is only detectable at temperature above 500C leads me to believe it is not an artifact. We have also tried different tubing sizes and the current is always there.

Hello Oscar Marin-Flores ,

I must admit that I am at a loss to explain why the leakage current is greater with argon than with air, hydrogen, etc. According to one reference book, the ionization potential to remove the first outer electron are:

Ar 15.759 eV

H2 13.598 eV

N2 14.534 eV

O2 13.618 eV

which makes it look like argon is harder to ionize.

As far as your statement that the leakage current always appears at around 500°C, I have the following question. How does the wire connecting the multimeter to the silver coil penetrate through the wall of the furnace? In other words, is it a bare wire passing through a ceramic tube, or what? The reason I am asking is that I still think the leakage current might be a measurement artifact. If you disconnect wire penetrating the furnace wall from silver coil at the silver coil, do you still record a leakage current?

Have you tried a guarded current measurement to test for surface conductivity, see the attached figure?

Regards,

Tom Cuff

Thomas Cuff , there was a misrepresentation in the schematic. My bad. The electrodes never go through the quartz tube. As for the ionization, I prefer to use the electrical breakdown voltage as a reference. It turns out Argon has one of the lowest (~600V/mm) and this is why I believe the current is higher with this gas. The current is in the order of microamps which suggests the possibility of having continuous dark discharges taking place inside the capacitor.