Dear Immacolata Tufano, yes it is possible and it depends on the preparation technique. The lowest size GO are more applied in filtration membranes. Please check the following documents. My Regards

https://doi.org/10.1016/j.carbon.2020.01.022

https://doi.org/10.1039/C9TC03251G

Following

Are you sure you are performing DLS measurements correctly? 7000 nm is too big a size. What are the count rate and PDI of your sample? If you are confident of the accuracy of DLS measurements, I would suggest r filtration through 0.2-micron filter.

Hi;

Prolonged or repeated oxidation of graphite oxide can induce perforations and structural defects and the material loses many of its properties. It is possible to control the size of the GO sheets by controlling the kinetic parameters of graphite oxidation such as the duration and intensity of the agitation and sonication periods in the exfoliation stages. To obtain a GO with a large surface area, graphite with particles larger than 100 microns should be used.

Therefore, in order to control the size of the GO, the exfoliation process is a key factor. A practical strategy is to remove visible heavier particles in suspension using low-speed centrifugation. The GO contained in the suspension can then be separated by size using the differential centrifugation technique. Because the size of a sheet varies in weight, it is in fact possible to fractionate sheets of different sizes from the same suspension by increasing the speed at each stage.

The GO sheets collected at each fraction must be resuspended separately in water using mechanical agitation. These suspensions should then be filtered under vacuum using nylon membranes of different porosity to further homogenize the collected fractions and to determine the particle size in each fraction. To obtain GO powder, the freeze-drying technique should be used to completely remove the water and residual filtrate containing the smallest particles.

NB: To obtain a GO with a large surface area, graphite with particles larger than 100 microns should be used.

With my best regards

Hi Immacolata,

I agree with Nabodita. 7 microns particle size is pretty big to be present in supernatant after spinning at 8000 rpm (would prefer rcf instead of rpm) for 10 minutes!

Maybe you need to recheck your DLS results or sample preparation/testing conditions.

If possible, check with any other characterization tool, maybe AFM/SEM?

Immacolata Tufano

Carry out a Stokes' Law determination. With the low density of graphene (oxide) and the fact that it's sheet-like, combined with the settling, tells me that your material is in the 10's, if not 100's, of microns in size. And there's a factor of 109 in the mass/volume difference between nano and micro... the difference between an elephant and ladybug (Drexler).

The method you apply is oxidative chopping. Use Hummer's GO , add concentrated nitric acid to cover this , Use water bath and evaporate to dryness. Add water and evaporate to dryness till nitric acid free. Check with DLS , now repeat the process using fresh nitric acid , evaporation, washing and DLS . In each step you will find the size will reduce. So you can stop when you get your desired size range . Caution: Do not use hot plate to dry the mixture, it should be slow and under 100 degree C. GO finally gets oxidized to carbon dioxide on continuous such processes.

Immacolata Tufano

BTW, if your particulate system was < 50 nm, then you'd expect a transparent 'solution' (often colored e.g. a gold sol).

Thank you all for your help.

Doufnoune RACHIDA Rachida You suggest to use the differential centrifugation technique and then filter the collected fractions with nylon membranes of different porosity, but I know only membranes with pore size of 0.45 micrometers and 0.2 micrometers, are there membranes suitable for nanometric particles? Could this differential centrifugation protocol be appropriate for my purposes?

Phys.Chem.Chem.Phys., 2016, 18, 12312

I didn't find anything in literature.

Sabyasachi Sarkar Dear Prof, I'm sorry, but I didn't quite understand your protocol. I have to add concentrated nitric acid to the GO precipitate obtained after purification (washing with HCl, H2O and centrifuge) and I have to let it evaporate in a water bath at about 60°C and then wash with water and let it evaporate again. Right? Can you please give me some reference for this procedure?

Alan F Rawle Dear Prof, I hypothesized that the change of color from the first to the third oxidation (see photo) is an indicator of size reduction, but I am not able to isolate and wash this precipitate probably because I use inappropriate centrifuge conditions and I mistake the DLS characterization.

Immacolata Tufano

At these sizes laser diffraction is the appropriate technique. What does DLS show on your clear supernatant?

Alan F Rawle

Dear Prof, DLS on the clear supernatant shows a Size of 616.8 nm, PDI 0.744 with peak 1 207.7 nm, intensity 83.4%, st.dev. 22.07 and peak 2 2874 nm, intensity 16.6%, st. dev. 388.7.

Immacolata Tufano

Your results on the supernatant do not make sense. Did you have adequate count rate? Did you carry out 5 - 10 consecutive measurements? Can you post your report?

@Alan F Rawle This is my report. I carry out 3 consecutive measurement.

Immacolata Tufano

I see no attachment. 3 isn't statistically significant, unfortunately. Throw away the 2 outliers and you have a perfect result...

Sorry

Immacolata Tufano

what are the chemical ratio you are taking during synthesis??

Immacolata Tufano

The attenuator (11) in your report indicates that you have virtually nothing there. what does the quality report say?

dear Prof. I repeated the DLS after centrifuge at 14000 rpm. The obtained suspension is quite clear without precipitated residue. However the measurements of intensity and number of particles are contrasting. Is it possible that I am wrong in the measurement? Are there particular parameters to set for the measurement of non-spherical particle size?

Immacolata Tufano

I can’t read a .rar file. Please post the .dts file. Number distributions are ‘deprecated’ in DLS. See ISO22412:2020 & ASTM 2490-09(15). If you do a limited search on RG you’ll find out why.

Please post your file imma031120. Thank you.

Alan F Rawle thank you for your answer. I am looking for the ISO standards you have attached. I have posted the pictures of my DLS results. So there is no sense in considering the particle number measurement?

Hi Immacolata Tufano

You persist in attaching screen dumps in contrast to the smaller .dls files that tell more. Your attenuator is on 11 which shows virtually no scattering. Reasons for this include:

• No particles in the system (from your centrifuging this may be the case) - effectively 0 concentration
• Very small particles in the system (in this case you'll need to increase the [C] 100 - 1000 fold
• Virtually indexed matched particles in the system (we can rule this out)
• Poor experimental procedure including not having the cell fully depressed in the system, laser or detector failure etc

You have to get some results with the attenuator at 6, 7, 8 - or at least well away from 11. Forget your graphene for awhile. Run a standard latex (~ 100 nm), or if you can't do that or won't pay for a service visit then highly dilute some (pasteurized) milk and run this. In the 90 degree cell then a faint milkiness is probably ideal. Then post those results - ideally the .dls file if you can find it.

Alan F Rawle Is this the format you are referring to?

As a control I measured the size of an Oil/Water nanoemulsion.

Thanks

Immacolata Tufano

OK, finally there... I'll revert later when I have a little more time. One immediate point though - your emulsion results are fine, but you need a minimum of 120 second equilibration time and not 0 seconds as you've selected.

Alan F Rawle thanks

Immacolata Tufano

OK, maybe things are not as bad as first thought. Let's look at the GO-SMALL file first.

The attenuator is on 9 so that's pretty dilute. Reflects what little material you have in the supernatant. However, the distribution is broad and the far-field in the correlogram indicates large and settling material. At these sizes plus the broad range then DLS isn't ideal. The (converted to) volume plot does show material ~ 10 nm so you've got a low concentration of small material from your centrifugation stage contaminated by a tiny amount of larger. So, your preparation isn't ideal and I don't think you're achieving what you want to do. You could try filtering the supernatant through a 0.02 micron filter and seeing what you get. Use 120 seconds equilibration time (not 0 seconds) as default. That should help with the consecutive measurements. If you have the quartz cells use these for small material - the optical quality is much better than the disposable PS ones. Use 5 measurements in a row not 3.

OK, nanoemulsion file. Here you have many results on the excellent attenuator setting of 6 indicating that there's plenty of scattering. You may want to see how I've set up my screen to give me (some of) the diagnostic information I need. Attached. It appears (Records 43 - 45) that you're attempting some sort of 'blank' measurement. DLS does not work like that - you've just measured noise and the computer tries to deal with this. Your GO data are at a similar noise level (records 7 - 9) in the file. Your records 1 - 3 appear exemplary and are a perfect use of DLS, so the instrument is working well. Compare the correlogram in record 1 to that of your GO one.

So, my overall conclusion is that all is well with the instrument and technique. The issues revolve around your GO sample preparation and that's something you're in control of (or not in control of), not me...

Alan F Rawle Many thanks for your time. I am trying to improve my method, I hope that with your guidance I will be able to measure the dimensions correctly.

Best regards

Immacolata Tufano - thanks for posting the dts data files! In addition to Alan Rawle's insights above, your data are a little unusual in that the intercept is relatively low for a good overall scattering intensity. This may be due to the presence of larger aggregates/sheets - or it could be an effect of light absorption (but you indicate the sample is clear and there is no dark tint to it).

Concerning distributions you may find this post of interest https://www.materials-talks.com/blog/2017/01/23/intensity-volume-number-which-size-is-correct/

You may be able to get slightly better correlation data by measuring in manual mode, to average more shorter runs than the automatic mode, for example 20x5sec subruns. However the best way forward is to try to optimize the sample preparation.