Looks like a lot of necessary information is still missing:

1. You have electrodes to be connected to some pre-amp.

2. You have some optical (?) fiber to ???

I'm sure that I am not the only one not getting the "big picture" from this description.

Thanks Dreher for your kind reply.

1. Yes, the microelectrodes have to be connected to pre-amp.

2. Yes, there are optical fibers in between the microelectrodes. So we cannot print a micro-circuit through the gaps between the electrodes as it will block the light path.

Do you get the picture now?

The problem is that ours is a bio-lab and I am the only Physicist. We don't have any collaboration with labs that might have the technology to print microcircuit which is less than 1 um thin.

I still haven't got the big picture:

o This is a single optical fiber with 2 electrodes on either side? Or ???

Regarding the dimensions, such a setup would be challenging (read: "high-tech") anyway. I don't think a 1 µm substrate thickness is manageable during manufacturing.

You may want to establish contact with some lab capable of manufacturing semiconductors. Maybe the NTU's "Centre for Micro-/Nano -electronics (NOVITAS)" would be such a contact

Hi Dreher,

I hope the attached schematic will make the picture clear for you.

In the figure, the grey colored circles represent electrodes while the remaining are optical fiber cores. The green surrounding disc is the chip on which 4 pins (gathering all electrode connections) has to be fabricated to which we can connect the preamplifier.

OK. If I got it right, the current chip is some "ring".

But the dimensions do not comply with your description, where you gave 150 µm for the fiber. UIf the .pdf is right, the "fiber bundle" would be about 150 µm, leaving not more than 10 µm for the individual fibers. Anyway ...

If you change the ring to a disc (same outline) and etch "vias" for the individual optical fibers (I know that this can be done), it should be feasible to connect the electrodes ("plugged" into vias not much different from these feeding through the optical fibers) to conducting paths on the chip.

So your chip woulöd have a central area with a "hole pattern", the holes designated for electrodes with some metallization surrounding the hole. The method to establish contact will have to be found, but I think some kind of conductive glue should do the trick.

Do you get the concept ?

Thanks a lot for your help.

Yes this schematic shows an alternative design in terms of dimensions as we shall be trying different spacing and electrode width. It was just to give you a rough idea.

If by "vias" you mean as in the attached picture, then it is possible to build it only for the electrodes as the number of optical fiber cores could be thousands (don't refer to my drawing).

However owing to the number and dimensions of the individual cores and electrodes, is it practically possible to align a transparent chip perfectly on the tip of the fiber so that all the 5 um/10 um electrodes align perfectly with the viases in the chip?

Also, I don't think gluing the chip to the fiber tip with a transparent conductive glue is an option as such a glue would short-circuit all electrode signals. It could also offer high impedance and capacitance, which is something we cannot afford since we need to detect voltages of as low as 1 - 10 uV.

I think the best option is to print the circuit on the tip of the fiber itself and draw conducting lines stretching out of the 0.2 mm fiber to a large disc of over 1 cm surrounding it.

"via" is a term from PCB manufacturing - holes in the PCB (there used to implement connections between the individual layers.

So, what I suggested might be challenging, but should be doable. And the gluing would be only small droplets between the fibers - only to establish contact between the electrodes and corresponding traces on the chip. If you need mechanical strenght: use non-conducting glue for this part.

What I do not consider realistical:

1. A transparent chip

2. Transparent conductive glue. (These are usually metal-filled, thus not transparent.

BTW: I do not consider the impedance of the glue a problem - provided the input impedance of the amplifier is high. Low voltage usually comes with low current - a situation where resistance can become less important.