brPEI -(-NH-CH2-CH2-)n-(-CH2-CH(NH2-)m- is suitable for CO2 - it is commercial multitonne product. Linear PEI is available in catalogs in gram quantities and synthetically available via acidic deacylation of poly(2-ethyloxazoline).

Thank you for your answer. In my case, CO2 must react with polymer and then polymer chain may need to break. However, PEI capture CO2, but it not break. Is there any other polymer materials?

Is it possible to consider water vapour as atmospheric gas in your problem? Then you may use water sensitive polymers, such as polyesters (PLA, PGA, copolymers thereof, etc).

Polyaldehydes are oxygen sensitive - in this case polymer chain does not break, but changes properties and thus works as matrix for release capsules.

Thanks. That's interesting. I will try.

Solution for O2 is simple. Polymer (hydrocarbon based) should contain Fenton reaction catalyst (Fe2+, Cu+ , etc). Atmospheric oxygen is effective as H2O2 - so polymers producers trying to remove traces of such metals, to avoid autooxidation.

There is a number of commercial additives (d2w, etc) for hydrocarbon polymers (HDPE, etc), that make hydrocarbon polymers oxo- and UV degradable.

Thanks Vadim for example, d2W prodegradant is exactly Fenton reaction catalyst:

CAS No.: 3353-05-7

Other Names: manganese distearate

MF: (C17H35COO) 2Mn

EINECS No.: 222-119-9

Purity: 99.8%

We have recently published a paper in Advances in Polymer Science ; Gas-Polymer Interactions: Key Thermodynamic Data and Thermophysical Properties (2011, 238: 137-177) which might be of interest. Since it is some kind of a review you will find in it several reference on our work in the field

Polymers containing amidine functional groups are well known to reversibly bind CO2 in the presence of water/moisture. There are examples of micelles and vesicles that 'breathe' using this reversible chemistry. The same effect can also be accomplished with amino-containing (Co)polymers. There are two recent reviews on the topic, one by myself published in 2013 in Chem Soc Rev and one also by Prof Patrick Theato although i do not recall the citation information off the top of my head