Question is fairly self-explanatory, but I couldn't find an answer on Google. I would expect ion thrusters because of their efficiency.
According to the James Webb Space Telescope Initial Mid-Course Correction Monte Carlo Implementation using Task Parallelism, J. Petersen et al. (PDF):
3.1 Propulsion System Overview
Two sets of thrusters comprise the observatory’s propulsion system. The first is a set of Secondary Combustion Augmented Thrusters (SCATs) that are the main thrusters for the MCC maneuvers. The SCATs are bi-propellant thrusters and draw from two separate tanks for a hypergolic reaction. Two pairs of SCATs exist: one for MCC-1a and -1b, and one for MCC-2. The MCC-2 SCAT is also employed in station keeping throughout the life of the mission. Two pairs are required because the center of mass of the observatory changes between MCC-1b and MCC-2 because of the sun shield deployment; the thrust vector from a SCAT is directed through the center of mass at the time of that maneuver. Each pair is composed of a primary and a redundant thruster (the respective thrusters are coupled to the prime and redundant flight systems; only one system is on at a time). SCAT 1 and 2 are the primary and redundant pair for MCC-1a/b while SCAT 3 and 4 are the primary and redundant pair for MCC-2 and stationkeeping. For a given maneuver only one SCAT is on throughout that maneuver. The second set of thrusters is composed of eight Dual Thruster Modules (DTMs), each comprising of a primary and redundant Monopropellant Rocket Engine, 1 lbf, (MRE-1) thruster. A schematic of the propulsion system appears in Fig. 3. The MRE-1s consume hydrazine fuel only (and no oxidizer) from the same tank as the SCATs. The MRE-1s are subdivided into two sets: MRE-1s 1 through 4, which are directed generally along the observatory J3 axis for pitch and roll control, and MRE-1s 5 through 8, which are directed radially about the J3 axis for yaw control
. A picture of the spacecraft bus with the body J frame and orientation of the SCATs relative to the J frame appears in Fig. 4. Together, the MRE-1s on-pulse throughout a maneuver to control the attitude. Because of the location and alignment of MRE-1s 1 through 4, any firings to control pitch and roll during a maneuver may result in an additional ΔV to the MCC maneuver. Therefore, thrust contributions from the MRE-1s must be included in the design of the MCC maneuver.
Figure 3. Schematic of the JWST propulsion system
Figure 4. An independent view of the spacecraft bus showing the orientation of the body J frame along with the orientation of the SCATs relative to the spacecraft bus. The +J1 axis points in the general direction of the OTE boresight. The +J3 axis is roughly perpendicular to the multilayer sunshield and pointing parallel to the primary mirror (Image credit: jwst.nasa.gov)
Gidanian, D. 7.0. ACS design & analysis: Thruster Configuration and Modeling. In James Webb Space Telescope (JWST) Spacecraft Attitude Control Subsystem Hardware Critical Design Audit. Northrop Grumman, 28 August 2013.
Hammann, Jeff. JWST Propellant Budget Document, Northrup Grumman. 19 July 2013. Document Number D40258
I wanted to share this as it describes JWST stationkeeping thrusters in detail, and as you can see in text and the attached images:
Also see James Webb Space Telescope Orbit Determination Analysis, Yoon et al., NTRS 20140008865 (PDF) for explanation of JWST's three mid-course correction (MCC) maneuvers mentioned in the quote above.