Tagged: reproducibility

Another TEM “problem”?

Do you remember the 2013 TEM scandal ? Well, today we may have a new one.

Everything started with this post on reddit….

and in all honesty some of the pictures look little bit strange…. People are still debating if the copy-paste was done with MS-Paint or scissor and glue….

pZzuDuf-1

 

From: “L-Cysteine-capped ZnS quantum dots based fluorescence sensor for Cu2+ ion”, Masilamany Koneswaran and Ramaier Narayanaswamy, Sensors and Actuators B: Chemical, 2009, 139, 104–109, doi:10.1016/j.snb.2008.09.028″ and “Mercaptoacetic acid capped CdS quantum dots as fluorescence single shot probe for mercury(II)”, Masilamany Koneswaran and Ramaier Narayanaswamy, Sensors and Actuators B: Chemical, 2009, 139, 91–96, doi:10.1016/j.snb.2008.09.011″

B-N8HrLCUAA-d0_.png-large

From: “CdS/ZnS core-shell quantum dots capped with mercaptoacetic acid as fluorescent probes for Hg(II) ions”, Masilamany Koneswaran and Ramaier Narayanaswamy, Microchimica Acta, 2012, 178, 171–178, doi:10.1007/s00604-012-0819-0″

B-N_-PiCEAABvN1.jpg-large

 

And a beautiful 2015 paper from: “Ultrasensitive detection of vitamin B6 using functionalised CdS/ZnS core–shell quantum dots”, Masilamani Koneswarana and Ramaier Narayanaswamy, Sensors and Actuators B: Chemical, 2015, in press, doi:10.1016/j.snb.2015.01.017″

What if I tell you that the main author of the papers is also the editor of the journal that published 3 out of the 4 papers above-mentioned? Are you screaming blasphemies thinking of your rejected papers?

B-OBwQPCYAAIQMY.jpg-large

One of the (other) editors will be soon informed about the story. Hoping that somethings will happen I’ll keep you updated….

 

UPDATE (20-02-2015): in two of the four papers the FTIR from two different samples are practically overlappable….

“Ultrasensitive detection of vitamin B6 using functionalised CdS/ZnS core–shell quantum dots”, Masilamani Koneswarana and Ramaier Narayanaswamy, Sensors and Actuators B: Chemical, 2015, in press, doi:10.1016/j.snb.2015.01.017 (figure 3)

fig3and “L-Cysteine-capped ZnS quantum dots based fluorescence sensor for Cu2+ ion”, Masilamany Koneswaran and Ramaier Narayanaswamy, Sensors and Actuators B: Chemical, 2009, 139, 104–109, doi:10.1016/j.snb.2008.09.028 (Figure 3)

1-s2.0-S0925400508006606-gr3

 

But doctor, I AM PAGLIACCI…..

watchmen_minimalist_poster_by_itomibhaa-d4khrle.png

 

Blog-Syn and the reproducibility problem

Did you make a soufflé following a recipe? Did it looks like the picture on the recipe at the end? Not at all
Do you know why with only three ingredients (water, flour, yeast) the pizza is so delicious in Naples and so crappy in the rest of the world?
Have you ever tried to repeat nanoparticle synthesis described in literature? Did you get the same results? No f*king way.

Welcome in the non-reproducible world!

Few days ago a new blog was born: Blog-Syn. The main idea is to pick up a reaction described in literature and try to reproduce it in different labs around the globe. It started, more than a week ago with a call to arms on a recent publication on Fe-S catalysis. After that, three different chemists ( Organometallica, B.R.S.M. and Matthew Katcher) tried to reproduce the yield reported (80% – 90%) without much success (5% – 24% – 37%). The idea is good, but it’s not going to tackle or solve the reproducibility problem.

Why a reaction is not reproducible? 

For two main reasons:

a) Variables: A reaction is not a simple system, deal with it. There are hundreds of variables that we don’t take in consideration, or we are not writing them in the Supporting Information. Starting from the classical Room Temperature and ending with the shear stress of the magnetic stirrer. The humidity of the air, the quality of the solvent (what was the water concentration? Was it stabilized with something?). What was the stirring rate, was the flask in the centre of the magnetic stirring plate or somewhere else? How big and which shape was the magnetic stirring bar? What was the pressure of the systems? Was the flask under sunlight during the reaction? Which batch of reagents did you use and from which company?
And the purification is not easy to reproduce as well. How humid the silica? What the size of the column? How did you pack it?… and so on.
These are only few variables that can play a role in a single chemical reaction. Do you think I’m crazy? Maybe you are partially right but check some true stories:

  • Some time ago the group in which I’m working now discover that the outcome of a reaction was different if you shake or stir the flask (ref Science). Few years ago an amazing chemist noted that the amount of one single species was different depending by the position of the vial on the stirring plate
  • Same HPLC column, same company, same model, same filling, same sample: two different chromatograms. We bang our head for a while thinking that it was our system and not the column until we contacted the producer. They were using two different source material before processing it.
  • During my master I was working with phosphoramidite. I did a reaction a couple of times without any problem. In July everything was the same, but the reaction didn’t work at all. And then it was working again in September. We discover the the insane south Italian humidity was killing the reagent no matter which precaution I was using for dry reaction.

b) Observables / “feeling” for the reaction: We are not anymore describing all the things we see during the reaction or the purification. Was it changing color suddenly during the addition and then changing back to colorless? What was the smell? Fluffy stuff during the workup? And so on.
The “feeling” for the reaction is also an important point. When I’m dealing with a new reaction, even if it’s described in literature, I’m always going to speak with people that have more experience than me in that specific reaction. There are some tricks that are not described in the supplementary information. If we are doing the same reaction probably I’ll get 20% yield while a well reaction-specific trained chemist can get 80%.

What about the Fe-S catalyst? Is that reaction working? Can you get the same yield?
Probably yes, a well described experimental section will help. Just for fun, take the Blog-Syn entries and plot them against the yield obtained. The first try gave only 5% of yield while the third one got the maximum so far. This can be easily explained by the knowledge that the entry 1 and 2 gave to 3. If I’m going to try this reaction, I will have the knowledge shared by three chemists that worked on that reaction before me, and probably get higher yield.
Is the Fe-S catalyst as easy as pointed out in the paper? No, probably not. But that is the selling point of the paper and it’s mainly marketing. But this is another story.

YieldI agree, three points are not enough for a linear correlation, but still, my point is the more experience you have the better the yield.

How can we help the reproducibility? 

We chemists are innovator on certain aspects, but really conservative on others. Are you still using the log chart for calculating logs? I bet not. Are you still using a paper lab journal for your experiments? Probably yes. Are you writing the experimental section like 100 years ago? Yes.

We should move to share and publish raw data. From our lab journal to raw chromatograms. This can solve most of reproducibility problems.
Since my postdoc I switched to using Evernote (not the best program for that) for writing my lab journal. In one click I can share or publish pages of my lab journal. We do have the possibility of publishing online all the data that we collect in the lab. So, why we are not doing it?

 

RawI’m still on my same idea of raw data. Tweet of 16 Jul 2012.