Diffusion time lag of polymer based thin film solar cells |

Hello Diffusion Platform:

I have a question on polymer based thin film solar cells, and specifically on an article "Protecting OLEDs and Solar Cells from Moisture Using Nanotechnology and a Nanoengineered Barrier" on www.azonano.com. I am not an expert in the field of Water Vapour Transmssion in Nano layers or Solar Cell / Photovoltaics, but several key statements in this article seem at least utterly nonsense. Among which is a time lag of 2300 hour for a nano layer material (which is physically impossible for any of the materials under consideration). Could you shine your light upon this issue of water permeation in flexible thin film equipment?

Philip Ball

I have a question on polymer based thin film solar cells, and specifically on an article "Protecting OLEDs and Solar Cells from Moisture Using Nanotechnology and a Nanoengineered Barrier" on www.azonano.com. I am not an expert in the field of Water Vapour Transmssion in Nano layers or Solar Cell / Photovoltaics, but several key statements in this article seem at least utterly nonsense. Among which is a time lag of 2300 hour for a nano layer material (which is physically impossible for any of the materials under consideration). Could you shine your light upon this issue of water permeation in flexible thin film equipment?

Philip Ball

Philip,

The time lag of nano barrier layers - no matter the materials used titanium, silicium oxide, carbides is usually in the order of seconds. This can be demonstrated with a simple penetration calculation for expected diffusivities in tight barrier films (note that this is just a bootstrap and penetration for multilple layers can not be solved analytically but only by means of finite element CheFEM routines ):

x= SQRT (pi x diffusion coefficient x time)

Hence:

time = (x^2) / ( pi x diffusion coefficient)

Take x=1 E-8 meter and the lowest expected diffusivity 1 E-17 m2/s then:

time = 3 seconds.

All nano barrier layers have time lags in this order of magnitude (usually lower).

In the article, they are probably pointing at a laminate of the nanolayer with another substrate material, expected PEN (Polyethylene Naphtalate). Then it depends - at least theoretically - on the thickness of this PEN layer.

However if this PEN layer is - which is expected - in the order of micrometers, then again with finite difference time lag calculations it can be shown that the the obtained time lag is impossible. Underlying - and probably missed principle is - that a laminate which has a very low Water Vapour Transmission, also contains little water (so the accumulation of water inside the barrier material is small), then lower barrier diffusivities will not decrease time lag to a significant extent after a certain point (on the other hand the decrease in WVTR continues with improvement of the barrier film, so here I will not argue with the results presented in the article).

The ones involved in the research should perhaps once look to polymers that swell under the load of a permeant (so not water in PEN, but e.g. water in Polyamide). If they do the correct time lag analysis they will notice that time lag is not too bad for polymers that swell: these polymers first accumulate a lot of permeant before transmitting it to the surroundings (and please do not use the standard time lag formula since this only applies for Fickian diffusion in one single polymer layer! - I am sorry but real ife is much more complex than that ;-).

Rodney

The time lag of nano barrier layers - no matter the materials used titanium, silicium oxide, carbides is usually in the order of seconds. This can be demonstrated with a simple penetration calculation for expected diffusivities in tight barrier films (note that this is just a bootstrap and penetration for multilple layers can not be solved analytically but only by means of finite element CheFEM routines ):

x= SQRT (pi x diffusion coefficient x time)

Hence:

time = (x^2) / ( pi x diffusion coefficient)

Take x=1 E-8 meter and the lowest expected diffusivity 1 E-17 m2/s then:

time = 3 seconds.

All nano barrier layers have time lags in this order of magnitude (usually lower).

In the article, they are probably pointing at a laminate of the nanolayer with another substrate material, expected PEN (Polyethylene Naphtalate). Then it depends - at least theoretically - on the thickness of this PEN layer.

However if this PEN layer is - which is expected - in the order of micrometers, then again with finite difference time lag calculations it can be shown that the the obtained time lag is impossible. Underlying - and probably missed principle is - that a laminate which has a very low Water Vapour Transmission, also contains little water (so the accumulation of water inside the barrier material is small), then lower barrier diffusivities will not decrease time lag to a significant extent after a certain point (on the other hand the decrease in WVTR continues with improvement of the barrier film, so here I will not argue with the results presented in the article).

The ones involved in the research should perhaps once look to polymers that swell under the load of a permeant (so not water in PEN, but e.g. water in Polyamide). If they do the correct time lag analysis they will notice that time lag is not too bad for polymers that swell: these polymers first accumulate a lot of permeant before transmitting it to the surroundings (and please do not use the standard time lag formula since this only applies for Fickian diffusion in one single polymer layer! - I am sorry but real ife is much more complex than that ;-).

Rodney