[ScientificSorcerer (OP)]

Hello readers. I haven't posted much in a while, but today I have a very important question regarding an invention of mine. I won't go into details about the invention but it involves the use of transistors and diodes.

I was looking into the subject of "printed transistors" (which do exist) and I was wondering if it was possible to print my own transistors on paper using some kind of silicon or germanium inks/paints. Basically my idea is to make a silicon semiconductor pen and dope a coating onto it, to make flat P and N junctions.

imagine this for example, I want to make a simple diode so I draw a rectangle with the silicon ink then when the ink drys I coat half of the rectangle with a doping agent like Boron to make a P junction.  Next I coat the other half of the rectangle with another doping agent like Phosphorus to make an N junction. Then walah! you just printed a diode on paper.



or you could dope the ink before you draw it, and make 2 different pens a P+ pen and N-pen. If you wanted to make a diode then all you need to do is draw 2 squares next to each-other to make a diode.  in the same sense you could draw 3 squares next to each-other in-order to make PNP or NPN junction transistors



Transistors are based on the same concept as diodes (basically) so if I can print a diode on paper then I can print a transistor on paper too.
The transistors and diodes don't necessarily have to be small for my application, just flat on regular old paper.

Is what I'm imagining possible? If so how do I go about making printed transistors and diodes with just pen and ink?

I was looking through Google patents and found this
http://www.google.com/patents/US20110109688

It is basically what I am looking for, but I want to put it into a pen. From what I have found on Google it looks like "nanocrystal inks" are what I'm looking for but I dont know enough about them to be sure. But the bottom line remains the same. Is It Possible To Make What I'm Talking About? (From a physics stand point)

Also If you have any information on this subject which might point me in the right direction please post a reply or comment below.

[evan_au]

Paper is a challenging material on which to create transistors.

It may look flat to the human eye, but it is constructed of a compressed random mat of wood fibers, sometimes bleached and sometimes with embedded clay. (See Fig 2b at http://cool.conservation-us.org/coolaic/sg/bpg/annual/v15/bp15-17.html)

Conventional semiconductor manufacturing tries very hard to make the surface flat, and single-crystal, with the right atomic spacing for Silicon. Paper fits none of these criteria.

Conventional silicon processing needs to embed the dopants in the silicon, with a ratio of perhaps 1 dopant atom for every 10 thousand to 10 million silicon atoms. These are often diffused into the silicon at quite high temperatures, which would burn up normal paper (and most plastics). It is not good enough to just print the dopant atoms on top of the silicon.

There are research teams working on printing semiconductors on plastic surfaces, using low-temperature processes. The material of your pen may be a better surface on which to build transistors than paper.

[alancalverd]

You might do better with a field-effect transistor, but a smooth surface would still be a lot more likely to yield a working device.

[ScientificSorcerer (OP)]

Indeed I know the complications with drawing transistors on paper, "conventional" transistor manufacturing requires quite a lot of processes, most of which involve heat treating which would destroy the paper on which the crystal is printed.

That's why I'm trying to think outside the box. What could be capable of being a semiconductor ink like silicon ink.

http://ceramics.org/ceramic-tech-today/japanese-group-prints-amorphous-silicon-photovoltaic-cell-created-using-silicon-inks

They already have conductive ink and dielectric/resistive ink why not silicon ink?



The way I see it is I need some kind of "silicon liquid crystal" and it would have to be pre-doped. I know that 2 potential candidates for something like this

But lets keep things simple for the moment, instead of trying to make a full blown transistor on paper lets instead think about a simple diode, and how it can be made on paper, diodes work on almost the same principle as transistors but are simple in comparison, more easy to visualize. If you can make a diode on paper then you can make a transistor on one too. take a look at this diagram.



it may look a little complicated but in reality it's much simpler then it looks. There are basically 4 layers, 2 electrically conductive layers and one P area and one N area.

Imagine some conductive ink on paper, then a layer of N ink, P ink then finally a top layer of conductive ink then alikazam! printed diode easy.

[alancalverd]

Multilayer printing or deposition of mesoscopic electronic components works fine on glass, ceramic or even plastic substrates, but the inherent surface inhomogeneity, porosity, and areal instablity of paper make it a very poor choice.

You can cheat a bit (indeed you have to!) by starting with a metal layer but you still have the problem of reliably connecting your finished gizmo to the outside world: a rigid substrate, or one that can flex without cracking, is more likely to yield a working product.   

[ScientificSorcerer (OP)]

Ok I think I see what you mean, Paper is bad for doing things like this. But what could replace it? Some thing like paper. I was looking around the internet looking for alternatives to paper, I found this stuff called parchment, hemp and cotton papers.

would one of these other kinds of paper work? If so which one? If not what do you think would be the best material for drawing/painting circuits?

I ordered conductive ink to see if that would work on paper, it's being shipped. In-order to make, say, a diode I would have to put the silicon ink on-top of the conductive ink/paint when it dries put a clear conductive layer over the top. that's 4 layers of paint layered on-top of each other, sounds like it wouldn't work but it's been done before. with electroluminescent ink.

look at these ink links
https://d3s5r33r268y59.cloudfront.net/69821/products/thumbs/2014-03-06T01:13:43.179Z-50W-2.jpg.2560x2560_q85.jpg
electroluminescent ink

https://d3s5r33r268y59.cloudfront.net/69821/products/thumbs/2014-03-06T05:02:27.666Z-52-2.jpg.2560x2560_q85.jpg
clear conductive ink

https://d3s5r33r268y59.cloudfront.net/69821/products/thumbs/2014-03-06T04:51:43.008Z-51-2.jpg.2560x2560_q85.jpg
insulating ink/dielectric ink (resistive ink)

In order for the electroluminesent ink to work you kave to layer it with 4 layers, first a conductive base then the electroluminesent ink, clear conductive ink then finally an insulating clear coat. all that for one kind of ink to work.


If you can layer inks for that to work then I bet it's basically the same concept for making layered transistors and diodes.

[evan_au]

Some other inks are somewhat easy to make:

• To make insulating ink, ensure that insulating particles are adjacent to other insulating particles (in some matrix)
  
• To make conducting ink, ensure that conducting particles are adjacent to other conducting particles (in some matrix)
  
• To make capacitive electroluminescent ink, embed electroluminescent particles in some transparent matrix
  

To make a diode, a clean interface between the p and n-type silicon is essential (and also between the silicon and metal contacts)

Printing P-type particles on top of N-type particles (in some matrix), mounted between two layers of conductive ink will result in some contacts between P & N-type particles, but also many contacts between N-N, P-P, N-Metal, P-Metal and even more between Si and the matrix. These other contacts will dominate the behaviour of the material. The Si-Metal contacts can themselves form a Schottky diode.

Where P & N-type particles come into contact, continuity of the silicon crystal structure across the junction is important in ensuring that you can form a wide depletion region which allows a good diode effect, and allows a high current to flow across the bulk of the material, rather than through a small contact area between the particles.

Silicon has such a high melting point that joining particles to make an integrated whole requires high temperatures, which is difficult for paper.

There has been some work in turning organic chemicals or carbon nanotubes into electrical circuits. Graphene is another material that is a candidate for a different type of transistor, a field-effect transistor, which uses majority carriers; anything with minority carriers (like a bipolar transistor) requires careful control over geometry and concentration.

The contact between two Si microcrystals effectively forms a Point-Contact Diode, something that was used in the very early days of wireless. To get them to work required very careful positioning of a sharp metal point on the correct part of a crystal - but once positioned, it could pick up radio stations (until you bumped it, or the temperature changed and it moved...).

One challenge for printed diodes and transistors is that you end up with a random network of parallel and series point-contact and Schottky diodes, without the possibility of adjusting them for best operation. Repeatable manufacturing becomes a nightmare.

[alancalverd]

Layering inks to produce a macroscopic effect is no problem, though we still prefer a rigid and/or smooth substrate, but there's a lot of difference between a bulk property like electroluminescence or piezoresistivity, and the depletion layer properties of a junction transistor. Hence the suggestion that you might get an insulated-gate field effect transistor to work on a poorly-characterised substrate as the interface is less critical.

That said, old fogeys like me can remember trying to make point-contact transistors with single germanium crystals and bits of razor blade, just as Bardeen and Brattain did when the world was young. The trick was to "form" the device by applying a bias voltage that diffused a few Gilette atoms into the crystal, wthout melting the whole thing.

Have a look at www.youtube.com/watch?v=vmotkjMSKnI   

[aengel]

Similarly Rapisarda etal teach :


"three different operating regimes of the device:


– At low |Vds|, the channel and the Schottky diodes at both
source and drain behave as gate voltage dependent
resistors. The partition between channel resistance
and contact resistance depends upon the gate bias, with
the channel resistance controlling the voltage drop at
low |Vgs|. When the contact resistance becomes comparable
to the channel resistance (|Vgs| > 35 V in the present
devices), an appreciable fraction of the applied drain
voltage drops on the contact resistance.


– As the |Vds| increases, the reverse biased diode at the
source starts to absorb most of the applied potential.
The device can be schematized as the series of the
reverse biased diode with the transistor having a floating
source potential (Vc). Vc-values can be evaluated
directly from the potential cut-lines and perfectly agree
with the values obtained following the method proposed
in Ref. [13]. The onset of this regime, V1, can be
estimated by the peak in the derivative of the Vcs-curve.


– By further increasing |Vds| the Vc-potential reach a certain
value at which |Vgc| is sufficiently low to drive the
channel into a more resistive condition. The additional
increase in |Vds| produces the pinch-off of the channel
at the drain end. The onset of this regime can be determined
by considering the condition for the pinch-off:
Vgc–VT0 = Vdc. For |Vds| > |V2|, any further increase in
the drain bias is now absorbed by the pinch-off region
and the current is controlled by the saturation current
of the transistor."



Analysis of contact effects in fully printed p-channel organic thin
film transistors
M. Rapisarda a, A. Valletta a, A. Daami b, S. Jacob b, M. Benwadih b, R. Coppard b, G. Fortunato a,
L. Mariucci a,⇑
a CNR – IMM, via del fosso del Cavaliere 100, Roma, Italy
b CEA/Liten/DTNM/LCEI, 17 rue des Martyrs, Grenoble 38054 Cedex 9, France
from "Organic Electronics" 13, 2012 2017-2027

[aengel]

Kudo etal. discuss the differences between inorganic and polymeric materials, and describes the advantages of the VFET versus the lateral design of the OFET :


"In fabricating inorganic SITs, however, perfect crystal
technology is necessary, because the imperfection of
the semiconductor layer on the gate region results in
the degradation of the device characteristics. However,
most organic materials form molecular crystals or
amorphous films and their physical properties are
sometimes insensitive in the crystal imperfection as
compared with inorganic semiconductors. For
example, organic films grown by Van der Waals epitaxy have small amounts of defects due to the nonexistence of dangling bonds. From this point of view, a
simple method such as a vacuum evaporation technique
can be applied in fabricating organic SITs. From
these advantages, the vertical type organic FETs are
suitable for a high-speed and high-current density driver
for display devices.



If the device structure, such as the device size,
strongly reflects on the device characteristics, an excellent
device performance is expected in the vertical type
FET. Fig. 2 shows the comparison of device parameters
of the lateral and vertical type FETs. As shown in Fig.
2, the vertical type FET has a short channel and large
effective area and should show a fast and large current
operation. Furthermore, the effect of surface roughness
in the lateral type FET might be strong, because
the channel is formed along the interface of the gateinsulator
and semiconductor film. The factors mentioned
above predict that the vertical type FETs should
have a excellent device performance.

Thin Solid Films 393 2001. 362-367
Device characteristics of lateral and vertical type organic field
effect transistors
Kazuhiro Kudo, Masaaki Iizuka, Shigekazu Kuniyoshi, Kuniaki Tanaka

[chris]

We did publish this story from Mark Peplow in 2014 which was based on a PNAS paper about printing RFID electronics into paper:

https://www.thenakedscientists.com/articles/interviews/electronic-labels-printed-inkjet

Chris

[yor_on]

You made me think of this SciS :)

http://spectrum.ieee.org/nanoclast/semiconductors/nanotechnology/atomically-thin-circuits-made-from-graphene-and-molybdenite
https://gizmodo.com/ibms-graphene-circuit-a-genius-reminder-of-how-far-gr-1512452994

https://en.wikipedia.org/wiki/Graphene