Kinetic trap
Misassembly trap
In thermally driven self assembly if
- specificity is low but
- bonding strength is high
things may self assembly in a wrong way and stay there for good.
If the bonding strength is deliberately kept low then even after the fact of a faulty self-assembly-bonding-event
by thermal vibrations the formed bond can be jostled apart again
and the part gets a another chance for assembling into one of the intended binding spots.
In first approximation that typically works the better the slower the cooling profile during selfassembly (thermal annealing).
Assembly path cutoff trap
Even without any faulty self-assembly-bonding-events a kinetic trap can emerge.
If the order of assembly is by chance such that access to yet unfinished parts gets and irreversibly stays blocked, that part can never be finished.
Example:
DNA bricks in structural DNA nanotechnology may assemble such that some voxels that should be filled with an DNA oglionucleotide strands
just remain unfilled since the structure around was built up before they got a chance to be filled.
The degree to which this happens may be hard to tell since the imaging technology cryo-TEM-tomography
for the highest resolution images needs to use averages over a large number of images of assemblies to get the fimal pictures (lots of fourier math involved).
Actually maybe these shoulkd be calles "steric traps".
Related
- One pot binding site finding - this can sometimes allow for steric kinetic traps
- Iterative binding site finding
- this allows to avert steric kinetic traps by determining the assembly order
- this also vcan lead to an exponential drop in yield though. - Exponential drop in yield - in the synthesis of chain molecules via conventional thermodynamic means.
- Iterated one pot binding site finding - a combination of the former two - the best of both worlds?
External links
self-assembly group wiki
The technology presented there makes heavy use of intentionally weak bonds to avert kinetic traps.
It's a trade-off/balance where even the optimum may be insufficient for some too ambitious ideas.
(Was ist about multi-input comuting hitting the limits of this?? Needs a re-read.)