Silicon shouldn't be dissociating from quartz walls and contaminating your substrates unless the quartz tube is very old, heavily used for silicon processing in the past, or has severe structural degradation. Either ways, better to change out the tube. Make sure all other quartzware such as sample and source boats, push rods etc are also new or clean. All the best. Contamination in the growth chamber is a difficult problem to track down, and often the best solution is to simply replace most parts with new ones. 

If this CVD was previously used for Silicon deposition like a-Si:H then it should be thoroughly cleaned with CF4 gas. Quartz is hight stable SiO2 and i don't believe you will obtain silicon due to dissociation from quartz (unless it is severely damaged - unlikely!). Carefully look at the seals and clean and replace parts that are either damaged or contaminated. Also, EDX peaks are sometimes similar or have the same position for different materials, please check properly. If you have not optimized dead time and other parameters; you may get a peak for Si from EDX detector which is again comprised of Si detector. Generally, research labs isolate CVD systems for carbon based and silicon based materials. these are just a few suggestions. All the best with troubleshooting you problem! 

Thank you for answering.

I also don't believe silicon was dissociated. But it may happpen that dissociating quartz itself. Right?

If it is quartz (crystalline SiO2) - it's melting point is >1500 C. Also fused silica sometimes people refer it is as quartz due to the similar composition. However, fused silica is amorphous and might become brittle after several times of use. It is most unlikely that it can dissociate due to a very low thermal expansion of both quartz and fused silica tubes that are used for thermal applications including CVD. Is this CVD used for deposition of any other materials? 

Krishna // No, it is being only used for making graphene. During graphene cvd quartz tube is heated to 1000 celsius by furnace. So I have no idea what it comes from...

Okay. 

1) You can consider replacing quartz tube if you believe quartz could be the problem and check if the problem is resolved. This is most unlikely - if you are using it at 1000C temp. 

2) in CVD systems appropriate O-rings, Viton seals and other sealants are used. However, it needs to be selected properly in regard to operating temperatures and gases being used. e.g. few O-ring such as silicones are primarily based on a strong sequence of silicon and oxygen atoms - if they are severely damaged then probably silicon source of contamination is possible. But I do not what if you CVD system has got any as such. However, they cannot withstand high temperatures.

Choosing appropriate sealants and fittings for CVD:- this link will help you - http://www.rlhudson.com/media/RLHudson-o-ring-guide.pdf

http://www.valqua.com/product_classification/packing_oring/

All these assumptions are based on assuming that you are sure that contamination is silicon. I.e. you are 100% sure about your EDX data. 

Maybe reading the following paper can also be helpful.

Article Contamination-free graphene by chemical vapor deposition in ...