Latest revision as of 12:48, 8 June 2023

A microfluidic general purpose synthesizer (MGPS)?

Eventually an integrated general purpose system would be nice that
is both the size and the cost as early day consumer level computers.

Use cases in the context of atomically precise manufacturing targeting advanced productive nanosystems

• This could greatly boost foldamer R&D including e.g. structural DNA nanotechnology and de-novo protein design
• Eventually early chip like systems could be interated. See: Modular molecular composite nanosystems
  and these could eventually evolve into gem-gum factories.

Components on an MGPS

This is a huge topic ...
(wiki-TODO: add existing graphic, discuss it a bit, eventually factor out)
See: Automated research - transflooder

Hurdles for an MGPS to be developed and widely adopted

The incentivisation structure is completely different to computers though.
No games, or other enticing visual media there.

Availability of basic standard ingredients

The prospect of making medical drugs locally may be an incentivising factor.
But some raw ingredients still need to be brought in.
They are not as simple as electricity for electronic computers.

Miniaturizability of critical sub-technologies

Some parts of the chemical processing chain (like separation and purification of proteins) may be inherently hard to miniaturize.

Especially analytic technologies can be quite hard to miniaturize.
That is: Once something one has produced something one often needs to confirm
that the stuff produced is actually what was intended to be produced.
And even if miniaturizable things quickly can become very expensive.

One fundamental physical problem is smaller measurement systems give higher noise.

Examples (state 2021):

Electron microscopy is a particular case that (as it stands) is utterly unminiaturizable.

• Optical standing wave electron acceleration might allow a big miniaturization step - but this is very questionable (to investigate if even applicable)
• Neutral matter helium wave microscopy might be better miniaturizable (no need of high energy acceleration) – But such technology with atomic resolution (!) does not exist yet.

Big benefit of neutral matter microscopy is that it is completely nondestructive and is only imaging the surface.
The complete opposite of electron microscopy. Highly destructive and deeply penetrative. Big downside however is that even the low resolution actually existing prototypes are still highly experimental technology as of today.
See: Matter wave microscopy

Scanning probe microscopy (imaging with a needle in various ways) can be made quite small.
Achieving atomic resolution with a very small desktop sized machine has only shown for scanning tunneling microscopy on very inert surfaces (Gold, HOPG) in air. Atomic resolution atomic force microscopy is more challenging even

Proprietaryness (institutional – profit)

Then there is the aspect of biotechnology still being a stronghold of proprietaryness
This is very much contrary to software where technology already arrived at in (2021),
where a lot of critical core infrastructure is already fully open source.

Proprietaryness seems to be a direct result of the still tremendously high development cost of biotechnology.
Developing isntitutions want to and need to regain their development costs.

Why could this perhaps be an opposing factor for MGPS?

An MGPS could mean pirated recipes. We know the story from pirated software.
In the case of computers the interests of mainframe computer developers did not stop the rise of personal computers.
Comparability is questionable though. The use cases of PCs where drastically different to the use cases of mainframe computers.
We'll entually see how the use cases of biotech factories will differ from future MGPSs.
But the use cases seem to overlap more right from the get go, not boding well for MGPSs making them in a more direct competition.
(offtopic: let's see how this turns out with quantum computers)

Software piracy (and maybe future MGPS recipe piracy), while not to condone,
may have put and may will put pressure on "proprietary overstay". To elaborate:
A problem with any proprietary developed technology is that it seems hard to tell from the outside when the development costs have been amortized and some reasonable excess profit has been made. So some developers can (and will if they can) try to milk their cash cow ad infinitum. "Overstay their welcome" so to say.

The issue with MGPS recipe privacy is that consequences could be much more severe because chemical substances can
be much more dangerous than software in at least three ways: drugs, poisons, and explosives.
Only the former two seem relevant as an opposing force against MGPSs.
Explosives need high quantities of mundane stuff and can be made already today with scarily little effort.
Ah yes and potentially bio-hazardous stuff can also me handled with microfluidics (viruses, bacteria, fungi, ...).

Danger of recreational drugs and genetic manipulation (institutional regulatory)

People will (without fail in attempts) try to hack MGPS machines to make some
more or less nasty and more or less illegal recreational drugs.
There will be very big questions from a regulatory standpoint.

To note is maybe that preventing something to be hacked,
when there is full physical access is even more difficult than preventing remote hacking.

Genetic manipulation of life and especially of humans is obviously an extremely controversial topic.
And the existence of (hacked) MGPSs may make the execution of
potentially dangerous experiments of questionable ethics in secrecy much more likely.

Backlash of hype waves

See infamous story of "Theranos".
No further details here.

Related

• Chemical synthesis
• Foldamer R&D
• Microfluidics can make use of microelectromechanical systems (aka MEMS)
• Automated research -- See the Transflooder idea. (wiki-TODO: Add the concept graphic here to this page or a dedicated page eventually.)

External links

Manually reconfigurable modular microfluidic system

Papers on "microfluidic LEGO":

• 2014 -- Discrete elements for 3D microfluidics
• 2015 -- Predicting the behavior of microfluidic circuits made from discrete elements
• 2020 -- mdpi.com - 3D Printed Reconfigurable Modular Microfluidic System for Generating Gel Microspheres

Specialized 3D printer

Video: "New Tools, New Possibilities - 3D Printing for Lab-on-a-Chip | Greg Nordin | TEDxBYU" by Greg Nordin (uploaded 2018-05-01)
This is about a custom high resolution resin 3D printer especially targeted at making microfluidic 3D circuits.
Results where quite promising (integrated micropomps) As the slides give away, the modelling software used was the free and open source programmatic CAD software: OpenSCAD
Associated slides: https://www.slideshare.net/gregnordin/miniaturizing-3d-printed-microfluidics

It is by far not as high resolution as two photon lithography. But that would be too small anyway because:

• viscosity at that small scale would likely be way to big and
• product throughput quantity would likely be way too small even with massive parallelism

Misc

• Video: Elveflow Microfluidics - Microfluidics interviews: Greg Nordin, Brigham Young University
• Company: http://www.elveflow.com/

DNA oglionucleotide service: https://eu.idtdna.com/pages/products/custom-dna-rna/dna-oligos