I feel you should try distillation using kugelrohr distillation apparatus. This should help you in easy removal of cyclohexanone and water. Or with present distillation assembly, you either increase vacuum (reduce pressure with vaccum pump) or increase temperature to get cyclohexanone distilled out.

Thank you for your answer.

I don`t know what is different between Kugelrohr and Rotary evaporator. Is that different mechanism or distillation process? I will find the distillation system and use. 

One more curious is that a cyclohexanone boiling point is 155C under 1 atm. Even though our pump can make over 60 mmHg in the rotary evaporator, the cyclohexanone did not evaporate. According to the theory (PV=nRT), a cyclohexanone should be evaporate at 51C under 18 mmHg. What was wrong? 

How long does it take for completely react with TDI and 1,4-butanediol? In many reference, they gave 24 h for reaction time. But, some of researcher gave just 3 h for reaction time. I am so confuse...   

Dear Kim,

Therotically you are correct but when you try the same practically many factors add and hence the results deviate from actual desired outcome. Like here if you are using rotary evaporator there are chances that the vapors formed are condensing back in to your container instead of getting distilled out. This is because lower ambient tempt. which causes condensation. Also, the tempt in heating bath and inside container are not always same. There is some difference, it is generally lower than the heating bath/mantle.  In such case, one need to raise tempt. Regarding your second point, about variation in reaction time, it depends on scale of reaction, reaction conditions and most important reaction temperature. If reaction tempt. is high reaction will proceed with faster rate and will require less time and vice-versa.

Thank you so much, Dr. Pawar. 

Your answer is so clear and make sense. I will optimize the reaction temperature and reaction time in my case. And, I just found the problem which was condensing-vapors re came back to the container.  

The general kinetic model for the step-growth polyaddition process involving TDI and BDO is known in literature and detailed analysis is covered in "Linear Polyurethanes" ed. P. Król (2008). The reaction follows the second-order kinetics and, even uncatalyzed, it is rather fast. The calculated reaction time for the initially equimolar concentrations of monomers is 60 min (at 86'C), and 95% conversion is achieved within 30 min. If TDI is used in excess, the prepolymer is formed even faster.

To minimize side reactions and thus improve quality of the prepolymer, I would suggest that you markedly reduce the reaction time and/or temperature followed by fast distillation under vacuum. The time of 2 h at 80'C should be sufficient for the reaction to be completed. I would also give a try to MEK to replace cyclohexanone as the solvent because of its lower boiling point (far easier to be distilled off) and good solubility of the monomers and prepolymer in MEK. In this case, 4 h at 70'C will also work.

Thank you so much, Dr. Sobecki.

I set up my experimental in TDI excess condition because I want to make TDI-prepolymer. According to your opinion, I will change solvent to MEK.  Fortunately, I have MEK now.

Many previous studies in self-healing microcapsule, they gave 24 h as a reaction time except just one paper. I don`t know why they use 24 h for the TDI-prepolymer formation. In the polymer science, the scientists used 2 to 6 h for TDI-prepolymer formation. I am so confused who is right. :(

It would be helpful if you could provide some more information as follows:

1) What molecular weight for the prepolymer you are targeting?

2) Should the polymer chains be linear or some branching/crosslinking is acceptable?

3) What is the origin of water you mentioned in the synthetic protocol that needs distilling off? It is unlikely for water to be present in the system after the reaction is finished. Water molecules would react with isocyanate groups yielding urea linkages in the product or TDI-dimer.

4) Do you use standard TDI mixture (80/20, 65/35) or any of the isomers?

5) I assume the reaction is not catalyzed. Is it acceptable for your application to use a catalyst?

In my hands-on experience with polyurethanes, the reaction between BDO and excess TDI80/20 proceeded in a polar solvent at 80'C to give a linear prepolymer is completed within 1 hour. You can easily monitor the progress of the reaction by FTIR. There are also advanced FTIR probes that allow for real-time in situ reaction monitoring.

You are so nice! Thanks for your help.

1) molecular weight: According to the reference, the molecular weight of the synthesized prepolymer will be 1270 for number-average (Mn) and 1690 for weight-average (Mw) by gel permeation chromatography.

2)  I don`t have information of this. But, I think that some branch type will be better for my application.

3) I mentioned that residual water should be removed in distillation step. This explanation was from the reference paper. As I told here, I just use 2,4-toluene diisocyanate (as a monomer, from Sigma-Aldrich), cyclohexanone (as a solvent, from Sigma-Aldrich) and 1,4-butanediol (as a chain extender, from Sigma-Aldrich). The monomer and chain extender were mixed as 3:1 molar ratio (18 mL: 4.57 mL) in my study.  Therefore, I did not input the water during the experimental.  I think that the butanediol may contain a little water or the air in the reactor can make moisture during the reaction. 

4)  I use tolylene-2,4-diisocyanate (Sigma-Aldrich, T39853). 

5) Yes, I can use a catalyst in my reaction. 

Thanks again.

I think I found a paper you refer to (link below).

Indeed, the molecular weight of the studied prepolymer was Mn = 1270 (Mw - 1690, PDI = 1.33). The average NCO content of the prepolymer is 20.1 wt % as determined by titration. This is a clear evidence of a branched polymer chain since the calculated functionality of the prepolymer is 6 (f=NCO%*Mn/4202). In contrast, a linear prepolymer chain with functionality of 2 would give NCO% of 6.6 wt%. The formation of the branched structure via allophanate linkages is a good explanation of the reasons for the prolonged reaction time and use of cyclohexanone as a high-boiling-point and polar solvent. Moreover, it is quite obvious that higher functionality (NCO content) of the prepolymer should be essential for capsule wall formation.

I also noticed a mistake in your calculations. The volume ratio V1/V2 = 18/4.57 in feed is equivalent to the molar ratio of n1/n2 = (V1/V2)*(d1*M2/d2*M1) = 2.4. Needless to say, the shortfall in TDI will result in the prepolymer having less functionality.

As I mentioned earlier, you can replace cyclohexanone with MEK and carry out the synthesis at 80'C under gentle reflux. After a few hours, just replace the condenser with a distillation setup and first distill most of the solvent under normal pressure and then under vacuum keeping the bath temperature at 80'C.

Another option is to modify the polymer structure by forming a linear oligomer chain (ca. 6-8 units) and subsequent branching thereof. The synthesis can be performed in a one-pot procedure using two steps. In the first stage, BDO is reacted with 20-30 mol% excess TDI. The excess TDI should not be lower than 20 wt% to prevent the product from precipitation. Once the oligomer is formed (after 0.5-1 h), the rest of TDI is added to the mixture to promote branching. By doing so, one can adjust both functionality and molecular weight of the prepolymer, and consequently tailor various mechanical properties of the capsules such as stiffness of the shell wall and compression strength. With this technique, one can also prepare solid prepolymers in powder form so they can be recovered from the reaction mixture by precipitation-filtration instead of vacuum distillation of the solvent.

Feel free to contact me directly in case you need further assistance.

http://pubs.acs.org/doi/abs/10.1021/ma801718v

Thank you so much. I will try according to your comments.

Could I ask more after the experiment, sobecki?