Expansion during sintering are usually related to the Kirkendall effect, for example in Fe-Zn systems. I dont have information for your system Cu-C but it worth to take a look in binary systems and difusivity rate for Cu->C and C->Cu that could give you a hint about what's going on.

You did not indicate if you really get densification during the sintering.

What are the measured porosity values and what are the densities in your sample.

I am not sure for your system, but you could end up forming a phase with a lower density.

Copper acetylide? Be careful, it can explode then :)

Dear Azmi,

It is hard to answer to your question due to the lack of information that you have provided.

I would say that expansion of your composites can be due the formation of Cu2O since in the argon sintering presence of oxygen impurities can not be avoided. This can be checked by XRD analysis. In alternative you can either coat the graphite fibers with cooper by electroplating or sinter your composites in a vacuum furnace. Another aspect that you should bear in mind is that the thermal expansion coefficient (CTE) of the materials making up the composite is different. Depending on the composition of your composites, you may have an effect of CTE on the resulting material. As you might know, copper at has a CTE of 17x10-6 K-1 at room temperature, whereas graphite is known for its CTE anisotropy: typically for main crystal directions it is−1×10-6 K-1 for basal plane of its hexagonal structure,and 26×10-6 K-1 perpendicularly to basal plane. Hence, polycrystalline graphite average CTE can vary between 0–5×10-6K-1. Such difference in CTE can lead to thermal stress. Depending upon their magnitude, some degree of microcracking may occur thereby allowing oxygen to permeate through the composite.

Finally, you should look at polished cross-sections of your sintered composites and evaluate their porosity distribution. If it is mainly localised at the surface my guess is that expansion is due to oxidation of the graphite fibers. Indeed, CO/CO2 entrapped gas resulting from graphite oxidation might have been occurred thus resulting in overall expansion of the samples.

Kind regards,

Fernando Costa Oliveira

Lisboa, 2nd October 2015

Thank you Leonardo for the interesting highlight. 

Herbert Giesche, the sample cracked so I could not get a good measurements on the density and porosity. I will try xrd to investigate the formation of new phase. Thank you for your answer.

It it cracked and you could not measure the density, how do you know it expanded??

Roman Kubrin, thank you for your answer. You just make my day upside down. As crazy as it sounds, Copper acetylide is also a possibility because my sample has high sensitivity and reactivity towards water. 

Thank you Fernando Almeida Costa Oliveira for your responses.

1. On the oxidation of graphite issue, I already used TGA to confirm that the graphite oxidation did not occur. 

2. The graphite fiber was coated with copper by using electroless coating process. TGA result showed no change in weight when I heated the coated graphite up to 950 deg C under argon atmosphere.

3. I used dilatometer to find that the expansion of the green compact was detected at temperature as low as 150 deg C. It keeps on expanding as I heated the sample up to the sintering temperature. I used Cu-SiC sample as comparison and found that the Cu-SiC shrunk but the Cu-C expanded. 

4. I did try to look at the polished cross section of the sintered Cu-C but the sample has high reactivity toward water. Is that normal?

OK, some "expansion" is normal during heat-up. (if your sample has a higher thermal expansion as compared to your sample holder).

It will depend on how you calibrated (blank-run) your dilatometer.

I typically measure the heat-up and cool-down (without any form of correction).

Yes, there is sometimes a hysteresis (lag) due to the different temperatures in sample and sample holder, but you can see from the general slope of the dilatometer line during cool-down, what thermal expansion value you might use to correct your data during the heat-up cycle (obviously you assume that there are no changes in the crystal phases of your sample between heating and cooling). So, only take the "difference" between data at a specific temperature during heat-up versus the data at the same temperature during cool-down.

Roman Kubrin, the expansion was measured based on the sample thickness before and after the sintering process.

Herbert Giesche, I used std alumina for the dilatometer calibration. Thank you for the "thermal expansion short cut". Appreciate it a lot.