You need to re-evaluate your threat model and change your approach. As others have said here, a TOTP that doesn't work would attract
more attention that one that does or one that outright doesn't exist, all the way up to escalating the encounter from casual privacy-conscious user to alleged spy.
The best way is to legitimately not have anything on the phone or your online presence that would cause problems, and then just be transparent (honestly, they're not after your nudes or embarrassing texts). A lot of border checks are based on feelings and if you look the part they'll quickly flick through the phone for obvious stuff they're after and will let you go once they don't find it.
If you are actually doing something that would cause issues, then you keep this off the local device and onto a remote one. Use a YubiKey or other dual-use authenticator (that gives you plausible deniability for having it - you can use the same key on benign social media accounts, etc) to access it from a secure device once you're through.
