Robust standard errors for mixed effects models in R

Question

I understand that you cannot fit a glmer(via "lme4") then test using robust SE estimator.

Ie

model <- glmer(Sepal.Length ~ Species * Sepal.Width + Petal.Length + (1 | grp), data = restricted2, family = poisson(link = "log"))

then do

coeftest(model, vcov = sandwich)

Is there any other packages that allow you to do the same using lme4 packages?

I've had a look at merDeriv package but I'm still not to sure how to go about this.