I have used both lipofectamine and PEI for my experiments. Lipofectamine is appropriate for siRNA transfection. I have reached up to 90% of knock down in gene expression.

I've tested a few options.

Viromer Green has the highest transfection and knockout efficiency (25uM - 70+% KO, 100uM - 80+%KO), the only problem is that it's very toxic to my cells. I worked off their protocol, and used the higher end of the recommended dose, so that might be part of the issue. Will be testing further to see if I can reduce some of the toxicity. Also, and this is more nit picky than anything else, their protocol isn't exactly the clearest thing I've come across. Personally I wish they would just write their protocols using pmol dosages. It makes more sense because different cells require different media volumes, so it doesn't make much sense to scale up the dosage simply because you use more media. pmol makes much more biochemical sense.

lipofectamine 3000 & messengerMAX - 50-60+% KO, higher cell vitality, overall not a bad option.

RNAiMAX - 20-40% KO, very toxic, transfected siRNA seem to localize in nucleus (not good)

electroporation of siRNA - doesn't work at all

electroporation of miR RNAi - 40-50% KO, also another good option if you plan to do combinatorial knockdowns as you can create constructs with cocistronic expression of multiple siRNA.

Lipofection-based approaches are good for adherent cell lines like SH-SY5Y, however not all reagents are applicable to all cell lines; many are picky, and because SH-SY5Y is a neuroblast cell line, you'll need to have cell line-specific reagents (https://altogen.com/product/sk-n-sh-transfection-reagent-neuroblastoma-cells-htb-11/) to go along with it. Generics can work, but they won't give the most efficient results.