VAR Model with Newey West Standard Errors

I found heteroskedasticity and serial correlation among the residuals of my vector autoregression model. I cannot increase the lag size as the serial correlation persists up to a high number of lags and the model would not be interpretable anymore with such a high number of lags. Is there any method to get Newey West standard errors in a VAR object? I already created all the regressions separately but then I cannot use the impulse response and granger causality functionalities anymore. Any help is deeply appreciated.