Since do not know population standard deviation, would use critical t-value (as opposed to critical z-value) for CI dtermination.
CI for population mean mean +/- critical t value * standard deviation/sqrt(sample size)
For 95% confidence, alpha value is 0.05. Since confidence interval two sided, look for critical t corresponding to alpha/2 = 0.025 and sample size - 1 = 63 degrees of freedom. If 63 DOF not available, look for closest DOF, below 63 (which is likely 60.