50 kilometer Deimos tether - minimum length to remain aloft.
Even with a safety factor of three, needed Zylon mass is tiny. Less than a quarter kilogram of tether could handle a 10 tonne payload.
There is no net acceleration at L1 and L2, so docking at ports at these locations would be like docking with the I.S.S.
This first step could serve as a scaffolding additional tether infrastructure could be added onto.
2942 kilometer lower Deimos tether - ZRVTO to Phobos tether
Given an ~1000 upper Phobos tether and a ~3000 lower Deimos tether, it is possible to move payloads between the two moons with almost no reaction mass. The tether points connected by the ellipse match the transfer ellipse's velocities. See my Upper Phobos Tether post.
So even with a safety factor of 3, the elevator's Zylon mass is modest. 1 tonne of Zylon can handle 11 tonnes of payload.
The red transfer orbit pictured above is called a ZRVTO - Zero Relative Velocity Transfer Orbit. At either end of the transfer orbit, relative velocity with the tether at rendezvous point is zero. ZRVTO is a term coined by Marshall Eubanks.
The idea of ion driven interplanetary vehicles excites me. The Dawn probe has demonstrated ion rockets are long lived and amenable to re-use. An ion rocket's fantastic ISP means a lot more mass fraction can be devoted to the dry mass structure and payload.
However ion rockets have pathetic thrust. They suck at climbing in and out of planetary gravity wells.
Here Mark Adler talks about ion rocket trajectories:
alpha, But perhaps possible so I will go with that number.
At Deimos' distance from Mars, gravitational acceleration is about 80 mm/s^2. The Hermes' acceleration over Mars gravitational acceleration at that orbit is about 1/40. A small fraction but a lot larger than the 10^-3 fraction Adler mentions.
Deimos moves about 1.35 km/s about Mars. With an impulsive chemical burn, it would take about .56 km/s to achieve escape. But with a 2 mm/s^2 acceleration, it would take about 5 days and and .8 km/s to achieve escape.
To spiral down to low Mars Orbit, it'd take Hermes more than 17 days and 3 km/s. So the Deimos rendezvous saves about two weeks and more than 2 km/s delta V.
Once in heliocentric orbit, it is the sun's gravitational acceleration that we put in the denominator. Here is a chart of gravitational acceleration at various distances from the sun:
If the rocket's acceleration is a significant fraction of central body's acceleration, we can model burns as impulsive. The trajectory would be more like an ellipse than a spiral. At earth's distance from the sun., Hermes 2 mm/s^2 acceleration would be about a third the sun's gravity. At Mars, it's about four fifths. In the asteroid belt, Hermes acceleration exceeds acceleration from sun's gravity.
Ion rockets may not be great for climbing in and out of planetary gravity wells. But they're fine for changing heliocentric orbits, especially in the asteroid belt and beyond.