The OLS estimator with heteroscedasticity

Problem

In the linear regression model \(y_i=β_1+β_2x_i+ε_i\) , the error terms are heteroscedastic ( \(x\) -variable is exogenous). What is true about the OLS estimator?

1. Unbiased, efficient and consistent
2. Unbiased, efficient but not consistent
3. Unbiased but not efficient
4. Efficient but not unbiased
5. Neither unbiased nor efficient