I'm not a DMol3 expert, but I can try to help. What units are your smearing widths in? Have you tried increasing the k-point sampling?

Thank you for your response! Philip James Hasnip

smearing width unit is Hatree and actually i was trying to replicate a literature:@

Ren, B., Dong, X., Yu, Y., Wen, G., & Zhang, M. (2017). A density functional theory study on the carbon chain growth of ethanol formation on Cu-Co (111) and (211) surfaces. Applied Surface Science, 412, 374-384.

I see that Ren et al. also used DMol3, so presumably the k-point sampling and smearing width are not a problem. Have you double-checked your initial geometry, for example to ensure none of the bonds are too short? You also write "Cu+2" metal slab; what does the "+2" refer to?

Use DIIS Size upto 10 pulay and also increase the the k-points