1. Read words from words.txt file into @X.
2. Go through temp.txt and remove any lines that contain any of the words present in @X.
This is all you are doing, right? The following code is untested.

open (TEMP,"<temp.txt");
open (TEMP_MOD,">temp_mod.txt");
while($line=<TEMP>)
{
   $flag=0;
   foreach $word (@X)
   {
      if($line=~/$word/) #Does $line contain this $word?
      { 
          $flag++;
          last;
      } 
   }
   if($flag==0)#None of the words from @X is present in $line
   {
      print TEMP_MOD $line ;
   }


}
close TEMP;
close TEMP_MOD;
[download]

You can also split the line read from TEMP into an array (say @arr) and then apply the logic of intersection of array elements given in Programming Perl for arrays @X and @arr.
If you get any elements in the interesection set, don't write the line into TEMP_MOD.