The cytoplasm of a cell is incredibly complex, with several thousands of different types of molecules floating within it. On top of that, the distribution of such molecules can widely vary depending on the species the cell was drawn from. Most recreations of a cell cytoplasm are through computational methods in molecular dynamic simulations:

Article Diffusion, Crowding & Protein Stability in a Dynamic Molecul...

Article Molecular simulations of cellular processes

But you're talking about semi-in vivo, which I suppose means you want to make a physical synthetic cell whose cytoplasm has the molecular composition of some reference cell. That would be very labor-intensive, even if you only utilize the most common compounds in the cytoplasm.

As a tangent, there's an article about a so-called "synthetic cell," but that merely reused the cellular environment/cytoplasm of a real cell, and injected the genome of another species into it.

http://science.sciencemag.org/content/sci/328/5981/958.full.pdf

My guess is even they knew how hard it is to fully recreate cell cytoplasm.

So, to my knowledge, no one has done this before. But if you have an efficient system that can accomplish this, potentially at scale, you truly have something publication-worthy. :)

Hi Yueying, I've worked with synthetic cells for a couple of years, and they've enabled us to build algorithms for the real cells. here is the link to our paper. I also have the algorithms to solve the problems, happy to help. https://www.dropbox.com/s/hq4sgldzd0wxf3q/Ushizima_ISBI_firstPlace.pdf?dl=0

Hi Yueying, were you able to find the answer you were looking for?