John is targeting large droplets for industrial use. Smaller droplets means increased control.

It is possible to control the droplet volume, but it means altering the flow rate at specific points in the flow tubes.

Question from Thorstein. (Sorry didn't catch it)

The carrier fluid sinks to the bottom and product sticks to the top. That is fine at the end, but often you have the need to extract the carrier fluid in an intermediate step. John's team passed the fluid through a grating of teflon, which the carrier sticks to, resulting in a continuous stream of solute.

How are the droplets extracted from the oil in the channels?

Not appreciably, the freedom in a flow reactor is the length of the tubing. Wasted volume can be compensated by flowing faster in a longer tubing.

To which extent does the introduction of the gas phase reduce throughput?

For clarification: The reactor is stable for more than 20 hours. The stability of the product is a secondary condition.

**Are the quantum dots stable for longer times than 20 hours which you showed? Is it necessary to stabilize them for longer times? **

Question time!

And that's where John concluded his talk

It is possible to automate the optimization process by iterating(automatically) through reaction parameters.

The results of the research is stated as "easily reproducible"

Flow reactors allow changing more parameters at once and thus optimize further.

He's showing empirical evidence that functionalizing the C-60 is dependent on temperature. And by varying that parameter the reaction can be optimized. It is also possible to optimize by reaction time.

He's attempting to make functionalized C-60 fullerene. By controlling reaction time you can get either single, double or tripple-functionalized C-60.

We shift our focus to fullerene acceptors.

This process has been repeated four times in a reaction. (but there seems to be no limit on how many times it is possible to do)

The solution to this is to add a gaseous phase. The gas is distorted around the droplet and the droplet is sweeped away by the incoming droplet.

It is hard to add more injectors and simply attempting to hit the droplets with more reactants. The likelihood to miss is simply too high.

We add to the complexity by now talking about multistep chemistry in a flow reactor.

Adding the channels in a spiral might average out the temperature difference, removing the problem.

Important that all channels provide an identical environment. There may for instance be a temperature gradient laterally across the channels; with the end result that you get different products at different channels.

Throughput increased by adding more channels, which is a viable solution because the reactions are very controlled.

They want the reactors to make up to a kilogram a day! And he says that he'll be surprised if global demand exceeded 100kg.

He emphasizes the importance of the carrier fluid by showing that precipitates can be sustained through the reaction, without affecting the reactor badly.

He also goes through how it's done for CdO quantum dots. And that by altering the composition you can easily determine the optimum

it is similar, but the reagents are mixed first, den put in a droplet former where catalyst is added and the mixture will be transported to an oil bath by the carrier oil to an oil bath from there.

(P3HT-co-P3HS)

The focus is now on copolymer synthesis.

It is possible to get a controlled environment in a flask as well, but it is a lot easier in a flow reactor.

The polymers made with flow reactors have similar performance in solar cells as commercial material of the same type.

The length of the polymers can easily be controlled by letting the reaction run for a longer, or shorter, period of time

By altering the reaction parameters during the synthesis you can alter the molecular weight distribution and thus get a wide variety of samples for further investigation.

Done by adding a nickel catalyst to the carrier fluid and making the droplets of the monomer precursor.

We shift our focus to semiconducting polymers. We want to make poly(3-hexylthiophene) in a droplet reactor.

Droplet reactors: Two tubes with reactants converge and create a droplet. When the droplet is sufficiently big the droplet will start flowing with a carrier oil. And thus you have the same benefits as in traditional microreactors, but no problem with coating.

He's interested in making macromolecules with microfluids, but there are of course problems with clogging and reactants getting stuck in the reaction tube. This may foul the reaction.

With microfluidics you can for instance control the flow rate of one reagent and thus accurately control for instance temperature in a reaction.

By controlling the parameters at the bottom layers you can get custom, tightly controlled, properties at higher levels.

Control means for instance accurate morphology and weight distribution

John will talk about microfluidic routes to the controlled synthesis of nanomaterials.