It seems to be a so interesting problem. I think it could happen because of degradation of your framework. Of course, HKUST-1 shouldn't decompose in ethanol in "normal conditions". But if your added organic molecule forms a more stable complex with Cu, than trimesic acid, it will result the decomposition of your framework and release of Cu.

I think if you did the washing (solvent exchange) 10 times, there mustn't left any CuSO4. How long did you soak your sample in the solvent exchange process?

What was your added organic molecule?

Best wishes

Thanks for your answer. 

It took me about a day for the exchange process of 10 times. So the sample would have been soaked for about a day. so altogether it will be 2 days of HKUST-1 soaked in ethanol. 

Is there a possibility for Cu atoms to be entrapped, which are hard to be removed by solvent exchange ? 

Few days soaking have to be enough.

About entrapped Cu atoms... It would be surprising to me, but wonders can happen, if you do some experiments with MOFs :). For example, if you prepared your sample by solvothermal way, metal-oxides can form and occupy the voids. And metal oxides are insoluble, so removing them can be difficult/impossible.

If you would like to sure about entrapped Cu (for eg. oxide, hydroxide), you should do TGA measurments in oxidative atmosphere (synthetic air). You will get a mass los in two steps: the first step is the release of solvent from the voids (below 300 °C, activation step) , and the secoond step is the decomposition of your activated HKUST-1. Your result will be CuO from the decomposition but only if you do your TGA in oxidative atmosphere. If you do TGA in inert atmosphere, you will get CuO plus amorphous carbon (from linkers), so you won't get the stochiometryc composition of your sample.  According to the stochiometry of HKUST-1, you can calculate an "estimated mass loss" of the last step (Cu3(BTC)2 > 3CuO ). If the measured mass loss of last step is less, than estimated mass loss, entrapped copper is present.

Of course this method is only good for detect a significant quantity of entrapped coppers in the voids, not for traces.

Best wishes!

This was synthesized using a solvothermal method. Thank you for giving me the idea of CuO which is insoluble. Leaching is in very small amounts. But your method through TGA is interesting to try. But wont it give 3 steps in the TGA, first for solvent evaporation, second for entrapped CuO and third for Cu3(BTC)2 ?

Thanks alot for the answer.   

No, entrapped CuO won't give any mass loss in oxidative atmosphere. If you heat CuO in oxidative (!) atmosphere, it remains CuO. But if you have copper-hydroxide, it can decompose: Cu(OH)2 > CuO + H2O , so it can give a little third step.

So this is one reason, why you should use oxidative TGA. And, as I noted before, the other reason is to avoid carbon formation from organic linkers, by "burning them out".

An other idea for determining entrapped CuO... Did you do PXRD measurments on your HKUST-1 sample? It can help to determine phase purity, and it can be good for detecting traces of any other crystalline phases (for eg. CuO).

Yes i did a PXRD and it matched with the reported data. I herewith have attached the PXRD I got from my sample. But can you explain abit more on how to determine the purity through it ? 

Unfortunately, it is not so easy, especially for a chemist (I am a synthetic chemist, and as I can see you are a chemist student too)..

Believe me, your better choice is to consult a crystallographer about it.

As I can see on your PXRD, you calculated d plane distance values for your 2 theta values. According to d values and the crystal structure of HKUST-1, you should assign your 2 theta peaks to h,k,l reflections (for example, your first intense peak around 2 theta=6.7 is the reflection of h,k,l=2,00). If there are some redisual peaks, you could'nt assignated to any h,k,l reflection of HKUST-1, you should compare this residual peaks with PXRD of copper oxide (or cupric oxide).

I show you an example in this supporting information article:

https://www.researchgate.net/publication/271207950_Solid-type_supercapacitor_of_reduced_graphene_oxide-metal_organic_framework_composite_coated_on_carbon_fiber_paper/figures?lo=1

As you can see, in this case, there are 2 small peaks assigned to Cu2O in this case (around 2 theta= 36 and 41).

By this method, you can say, it is a pure phase, or not. So it is not so easy... To determine the quantity of an impurity from a PXRD is much more complicated...

Article Solid-type supercapacitor of reduced graphene oxide-metal or...

Crystallographic softwares can help for assignation of peaks. For example, with mercury you can predict assignated PXRD patterns for known crystal structures.

Thanks a lot sir. I learnt a lot from your answer. Will check on it. Thanks again for all the help  

Welcome. For me your PXRD seems to be a pure sample's PXRD so don't worry. Nice work.  I think it mathes well to lyrical one. But I am not a crstallographer

If I correctly understood the theme, Cu was not leached from solution but mainly from HKUST-1 net. This happens even without organic molecules interacting with Cu2+ ion, but merely with the presence of water media. If there are unstructured Cu2+ in the pores of the net, hence they should be displaced to the solvent phase through water/ethanol washings. But if you use adduct or chelating agents the scavenging of the metallic center from the net is driven through adduct, chelate or macrocyclic effect depending on the employed substance that interacts, with Cu in this case. Its own dissolution or coordinating assisted dissolution happens due to the fact that the HKUST-1 crystallites are surface open/labile materials containing -COOH, -COO- and -COOCu sides or functionalities directed to the interaction with the solvent. Therefore having these terminals, these components should displace to the media since Cu is efficiently solvated with water molecules. In this way the displacement of both Cu ions and trimesate or trimesic acid towards water is observed by UV-vis, therefore the loss of material is attained. Having less and larger crystals should diminish the dissolution rate, but if you have more and smaller crystals, the dissolution rate should be higher. I hope to be sufficiently clear in this explanation. If you need further details in order to understand the issue please contact me and looking forward to help in the analysis of your experiments. Best regards. Hiram