R5 throws a tilt into your R3 and R4 network since it is in parallel with R4.
Your sensor signal may not be getting amplified properly. Try connecting it to the inverting input rather than the non-inverting. The signal should go through R2.
By including a capacitor inline it may or may not adversely affect performance.
Does the sensor change in resistance? Does it produce a DC voltage? These factors will determine what bias level needs to be applied to the non-inverting input.
Your gain appears to be set up as 83 (and some hi-freq reduction). If the output stay at 0, or 5V, then try reducing gain until the trace moves into a middle region. Then you can shine light on the detector and see how it responds.
You'll have to experiment with the capacitor values. It's hard to predict which values will give you the desired performance. Impedance in the components makes a difference what time constant is at work.