Hallo Joseph,

you must distinguish between current and voltage! Only current has a rise time of about 0.7...1nsec, NOT voltage! This I wanted to focuse in my last reply, when I told you to think about time constant of ESD event.
By the way, normed ESD Generator current waveform is defined with shortet output. It's NOT the current to be observed when having connected a real load. It must depend, of course, on actual load, right?

Have a look at the following simplified equivalent circuit:



C2 is NOT an added capacitance, it's just the parasitic junction capacitance of here shown transzorb. As I told many times formely, this parasitic junction capacitance is one major benefit of a transzorb.
By the way, bidirectional transzorb is not well suited, when wanting to protect a port of microcontroller. So, 'slower' turn-on time of this transzorb need not to be taken into account for your purpose here.

In the following I solve differential equation for shown equilvalent circuit:



As you can see, time constant of ESD event is R times series capacitance of C1 and C2, and is about 47.7nsec, well above the psec range, isn't it?

To get Uc2(t), we only have to integrate over IR(t) and divide by C2 and we result in:

Uc2(t) = tau x IR(0) / C2 x (1- exp(-t/tau))

With tau = 47.7nsec and IR(0) = 8kV / 330R = 24.2A it follows, that after 1nsec voltage drop across transzorb is about 5.99V. This is even below threshold voltage! So, transzorb has really enough time to absorb ESD energy without permitting to pass any dangerous high-voltage.
Keep in mind, that above formulas only represent behaviour of parasitic junction capacitance. So, when voltage drop across transzorb rises to more than threshold voltage formulas are no longer valid. But this is not relevant, because we only wanted to know, how fast voltage across transzorb is rising before transzrob is turning-on.

I want to focuse, that a lonly transzorb is not enough to protect a port pin of microcontroller. Transzorb is only for rough limiting of overvoltages. Some fine limiting must be added, in most cases. For this, connect two Schottky diodes (e.g. 1N6263 or SMD-equivalent LL101A) from port pin to Vcc and 0V. This measure shows only wished benefit, when power supply at microcontroller is adequately decoupled!!!
Insert between transzorb and port pin a resistor of about 100R...1k. This resistor limits current through fine limiting Schottky diodes. An example: Due to ringing caused by unavoidable inductivity voltage at transzorb shall rise up to 10V, which represents a real worst case, never observed with proper design. Then, across current limiting resistor about 5V will drop. When supply voltage is switched-off, even 10V will drop. Current limiting resistor should limit steady-state current to about 10mA.

This protection scheme is what I would call the absolute minimum. But flavors for other developers are different. Others will suggest something different. Perhaps additional ferrite beads or capacitors, varistors, etc.

Kai