> I have invoices that are linked to email addresses.
OK, that was a design mistake, because entities that can acquire invoices often have dupe and/or changeable email addrs.
> My goal is to get rid of email duplicates on my table
Logically impossible as above.
> before dropping those duplicates, to copy the second invoice
> for one same email to a new column on my table.
Continuing, indeed compounding the design problem. Email address is not a suitable key. For one small example: entity XYZ today places an order from email address
XYZ@gmail.com. By the time they make their next order, their address is
orders@xyz.com. Now it's impossible to track that entity's orders.
So let's see the result of Show Create Table for the existing table, so we can figure out a sound way to solve your problem.
Meanwhile, please read a primer on relational database design; for reasons that'll immediately be obvious to you, I suggest
https://www.artfulsoftware.com/mysqlbook/sampler/mysqled1ch01.pdf