NIRSpec Micro-Shutter Assembly

The JWST NIRSpec MOS observing mode uses a micro-shutter assembly (MSA; fabricated at NASA Goddard Space Flight Center) comprising nearly 250,000 configurable micro-shutters that can be individually opened or closed for a custom pattern used for each exposure or set of exposures. The MSA configuration is the MOS slit mask, so adjacent shutters are often opened in the cross-dispersion direction (MSA column) to form a "slitlet," and each configuration may contain dozens to hundreds of slitlets aligned with specific science targets on the sky.

On this page

Layout of the MSA

The NIRSpec MSA has 4 quadrants of configurable shutters. These are labeled Q1 to Q4, as seen in Figure 1. Each quadrant comprises an array of 365 shutters in the direction of dispersion by 171 shutters in the cross-dispersion direction. The total on-sky extent of all 4 MSA quadrants together is approximately 3.6' × 3.4'. There are gaps of approximately 23" (in the dispersion direction) and 37" (in the cross-dispersion direction) between the 4 quadrants. Each quadrant has an active region of ~95" × 87" where shutters are available for source placement.  All dimensions provided above (and in Figure 1) are averages, based on component-level measurements on the ground, and confirmed to hold true for in-flight operations.

Figure 1. Layout of the MSA

An illustration of the MSA layout in the NIRSpec aperture plane, including the 4 MSA quadrants (frame in blue, shutters as a grey checkerboard), the fixed slits and IFU aperture (all in black). Each quadrant, labelled Q1 through Q4, contains 365 × 171 shutters. The projected locations of the detector arrays are shown in green, with green dotted lines indicating their extent and hence where the gap between detectors is hidden in the figure by the MSA. The red dashed rectangle marks the ~3.6' × 3.4' on-sky field of view, constrained by a field stop at the entrance of the instrument. (© Ferruit et al., 2022)

MSA shutters

Each shutter has an open area of approximately 0.20" × 0.46", with an ~0.07" separating wall on all sides. The total shutter pitch (center-to-center distance) is ~0.27" in the dispersion direction and ~0.53" in the spatial (cross-dispersion) direction. The shutter walls extend away from the shutter door, in the direction from which light enters, creating a “shield” to minimize shutter-to-shutter contamination and a “crate” for the shutter door to open into (see Figures 2 and 3). The magnet arm used in the reconfiguration of shutter patterns is on the wall side of the shutters.

Figure 2. Detailed diagram of two MSA shutters

A cut-away view of 2 micro-shutters of the MSA. Details of their design are shown by the labels. In this figure, light enters from the bottom having already passed the filter wheel, goes through the open shutters and moves onto the selected disperser in the direction of the detectors.
Figure 3. Close-up photo of MSA shutters

A microphotograph showing the MSA shutters from the light input side. In this image, the shutter doors are on the far side of the protective "crates" so light enters from the top of the page (opposite orientation compared to Figure 2.)

MSA shutter configuration occurs in a 2-step process: (1) All shutters are opened by sweeping the magnet arm across the MSA from primary park to secondary park (i.e., in the direction of dispersion) as indicated in Figure 4. (2) The magnet arm returns from secondary to primary park, and as it passes each column, individual shutters are either kept open or are allowed to be closed. During that return sweep, shutters commanded open are latched in place by an electrostatic charge on the sidewalls of the shutter crate, while those commanded to close have the voltage removed, so that the magnet arm gently swings them back to their resting closed state. Each shutter has a row electrode and a column electrode so that it may be addressed individually, allowing a tailored slit mask. The planning of MSA observations should try to minimize slitlet reconfigurations in order to maximize the operational lifetime of the MSA hardware.

Figure 4. Layout of the magnet arm and the shutter arrays

Layout and numbering of the shutter rows and columns in each quadrant (as in Figure 1). The direction of motion of the magnet arm is indicated in green—each shutter reconfiguration comprises opening all shutters when moving the arm from primary to secondary park, and then the pattern addressing as commanded by the user as the arm moves back from secondary to primary park. The inset figure (upper-right) highlights the shutter orientation: all the shutter hinges are on the secondary park side of the shutter doors. Light travels through this aperture plane into the page, so the shutter crates would protrude towards the reader (see Figure 3).

MSA shutter operability

See also: NIRSpec MOS Recommended Strategies, NIRSpec MSA Shutter Operability

The MSA contains 365 × 171 × 4 = 249,660 shutters.

A number of shutters are rendered unusable for science via several distinct failure modes, as described in this section. The location of these inoperable shutters will be updated in the observation planning software and the pipeline as the population changes over time. This section (including the shutter operability map in Figure 5) has been updated to include shutter operability knowledge as of March 2024. For the current operability map, please refer to the APT's  MSA Planning Tool (APT/MPT), which uses the latest operability reference file available.


The configurable area of the MSA is deliberately oversized compared to the instrument field stop in the NIRSpec fore-optics, with 24,024 shutters (9.6% of the total shutter population) fully or partially vignetted. Therefore, for on-sky science observations, these shutters are not usable. (Note that the vignetted shutters are illuminated during internal calibration lamp exposures and can be commanded in the usual manner.).

Masking for electrical shorts

The complex control circuitry required for the densely-packed, individually-addressable apertures sometimes exhibits electrical shorts. Several dozen electrical shorts have appeared on the MSA over the years during ground testing and flight, caused by either particulate contamination or warped shutters. The shorts could cause unsafe currents on the MSA hardware, often manifesting themselves as a bright glow which is detected as contaminating thermal emission in science data, and also risks erratic behavior during pattern reconfiguration. To prevent these unwanted effects, identified shorts are removed by masking rows or columns in the affected quadrant using a special short mask applied directly at the hardware level. These masked shutters are unavailable during configuration planning for science or TA when using APT/MPT. (Even if the planned configuration commands them open, the hardware will override the command and the shutter remains closed). While masked columns (cross-dispersion direction) only pose a minor inconvenience to planning and remove up to 171 shutters each, every masked row (dispersion axis) not only removes up to 365 shutters, but has a direct, impact on MOS multiplexing.

Since launch, a small number of new shorts were located and masked, bringing the total number of masked rows or columns to ~110 throughout the 4 quadrants. This mode of inoperabiliy now contributes the largest population of inoperable shutters, with 24,668 (10.9 % of unvignetted shutters) shutters affected by the short mask. The masked rows and columns are marked in brown in Figure 5.

Mechanical inoperability

The small physical size of so many moving parts inevitably leads to some shutters failing to act as commanded, as the doors warp and/or stick. This inoperability primarily manifests as 2 distinct populations:

  • Failed open shutters, which either do not close when commanded (stuck open) or have no shutter door from the time of manufacture. Failed open shutters are always open to the sky and always acquire spectra. They pose a risk of contamination for NIRSpec MOS and IFU mode science because spectra from a source or background that falls in that aperture will always be collected on the detector (see Figure 6). Failed open shutters are also detrimental to overall MOS multiplexing as the risk of contamination generally precludes the use of any other shutters in the full row of the MSA. When planning MOS slitlet configurations, the locations of science source slits can be limited to those that have no spectral overlap with failed open shutters. In the IFU observing mode, the detrimental effect of the failed open MSA shutters can be mitigated by dithering, by moving targets to new locations in the IFU field (and on the detector), or by acquiring IFU leakage correction exposures.
  • Failed closed shutters do not open during the initial sweep, or fail to latch open when the magnet arm performs the return sweep to configure the MSA. Failed closed shutters do not degrade the obtained science spectra and only affect MOS planning (no impact on IFU mode observations). They can cause an inconvenience during configuration planning as some areas of the field of view become unavailable, which may be most problematic when searching for available reference stars for target acquisition (TA). MSA regions heavily impacted by multiple failed closed shutters may decrease overall MOS multiplexing, especially for sparse input catalogs. However, as failed closed shutters generally have a lower impact than those which fail open, some of the latter type were "plugged" during ground testing, converting them from the more harmful failure mode into permanent failed closed shutters.

The population of failed open shutters has remained very low and stable since launch, with only 20 shutters identified in the unvignetted regions (see Figure 6). Eleven of 20 failed open shutters are located on Q3 (see Figure 5). Instances of temporary failed open behavior from other shutters remains very rare.

There is a larger population of failed closed shutters, with 17144 (7.6 % of unvignetted shutters) currently identified in diagnostic imaging as highly likely to remain closed when commanded to open. More than 70% of the failed closed shutters are found in 2 regions: as a dense group on the inside corner of Q4 and a loose collection on the outside corner of Q3 (see Figure 5). While it is known that not all of these shutters would stay closed every time they are commanded to open, the probability is deemed high enough for the planning software to remove them from optimization, especially as this does not heavily impact multiplexing.

Finally, there is also a known random failure mode, in which up to 4% of operable shutters (excluding those flagged as failed closed or masked for shorts) remain closed when commanded to open.

Further discussion of the inoperable shutter populations can be found in Rawle et al. (2022).

Figure 5. Shutter operability in the MSA

Click on the figure for a larger view.

Overall operability of the 4 MSA quadrants, as imaged onto the NIRSpec detectors (Q3 and Q4 on the left, Q1 and Q2 on the right, as indicated by faint grey markings). Light green shutters are vignetted, dark green are failed closed, and brown shutters are those rendered unusable by the short mask. Red shutters (additionally highlighted by red circles) are failed open. Vignetted shutters are plotted with transparency so the underlying failed and short-masked shutters are visible (illuminated by internal calibration lamp observations). Note that 2 of the failed open shutters in Q4 are vignetted and therefore will not cause contamination in on-sky exposures. The uppermost and lowermost 3 rows of the MSA are marked here as failed closed, but in actuality they illuminate the focal plane beyond the edge of the detectors, and therefore their state is unknown (they are included in the total number of vignetted shutters, for completeness). This diagram follows the layout from Rawle et al. (2022) and has been updated to include shutter operability knowledge as of March 2024. For the current operability map, please refer to APT/MPT, which uses the latest operability reference file available.

Figure 6. Failed open shutters in the MSA

Failed open shutters in the MSA

Exposures to test the contrast of the MSA were taken in imaging mode with the shutters configured to "all closed." These long exposure images with a flat field lamp reveal prominent bright spots, which are the shutters that are permanently failed open. Extremely low level ghosting (<0.01%) is seen to the left of the four quadrant. This is not expected to pose a contamination issue for science exposures.

MSA flux leakage

See also: NIRSpec MSA Leakage Subtraction Recommended Strategies

Ground testing of the NIRSpec MSA revealed that the shutters have a flux leakage problem (Figure 7) where light from bright field illumination can pass through closed MSA shutters at specific locations. This is the same issue that is discussed in NIRSpec MSA Leakage Correction for IFU Observations. Figure 7 shows the flat field images acquired with all MSA shutters configured closed to measure contrast (also shown in Figure 6), and overplotted is a zoomed region in MSA quadrant 4 that shows a very regular grid-like structure to the contrast illumination pattern. The precise instrument model knowledge of where the MSA shutters were located was compared with the image of illumination structure shown in Figure 6. As a result, it was determined exactly where the excess leakage flux comes from in the shutters.

Figure 8 shows a zoomed view of an MSA shutter, centered at (0, 0) in x and y coordinates. This is a super-sampled map of the leakage flux from the MSA; the color scale shows where the shutters are extremely opaque (blue) and where a moderate amount of flux leaks through to the spectrograph (red). The colors presented are relative flux levels, and the red color corresponds to a leakage level (measured closed flux ⁄ measured open flux) of ~0.004%. The average leakage structure is very regular and repeatable across most shutters in the MSA. The map in Figure 8 reveals that the majority of the MSA flux leakage arises from the location of the the bar separating shutters in the y direction (cross-dispersion direction). Additional investigation is being carried out to understand the origin of this MSA flux leakage problem.

Although the flux leaking through an individual MSA shutter is less than one part in 10,000, the leakage is a cumulative effect from the over 700 MSA shutters in the dispersion direction. As a result, the MSA flux leakage can accumulate to a level that is closer to one part in ~50 (2%), on average. Figure 9 presents a dispersed NIRSpec prism flat field image that shows science-like slitlet configuration, plus the elevated flux level from MSA flux leakage. The excess flux from the MSA leakage can affect background flux levels MOS mode observations, but the available NIRSpec MOS nod offset options can help to subtract excess flux from science observations. Additionally, the MSA flux leakage can affect the sensitivity of IFU observations, so the concept of IFU leakage correction exposures was adopted to mitigate the issues. The MSA Leakage Subtraction Recommended Strategies article discusses different strategies for different science cases.

Figure 7. MSA flux leakage structure

MSA flux leakage structure

A zoomed view of the data acquired to determine the opaqueness of the MSA reveals very regular structure within the contrast map.
Figure 8. Map of the NIRSpec shutter leakage

Map of the NIRSpec shutter leakage

This figure highlights how excess light passes through the MSA at particular locations. The black rectangles represent MSA shutters, and the color scale presents the opaque regions of a shutter (blue) to the regions that are less opaque (red). This structure is very regular and repeatable across most shutters in the MSA.
 Figure 9. Dispersed NIRSpec MSA flux leakage

Dispersed NIRSpec MSA flux leakage

This flat field calibration ground test exposure shows NIRSpec prism-mode spectra in science-like "slitlet" configurations. Also highlighted is the MSA flux leakage which can cause an accumulation of low level background flux. Note that the background from the MSA quadrants is elevated compared to the area where the FS are. The image stretch applied to the different detector images in this figure is different, accounting for the different contrasts seen between the left and right images.


NIRSpec Multi-Object Spectroscopy 
Presentation by T. Boeker at ESAC JWST "On Your Mark" Workshop, 26-28 Sep. 2016.

Kutyrev, A.S., Collins, N., Chambers, J. et al. 2008 SPIE, 7010, 70103d
Microshutter arrays: High contrast programmable field masks for JWST NIRSpec

Ferruit, P., Jakobsen, P., Giardino, G. et al. 2022 A&A, 661, A81
The Near-Infrared Spectrograph (NIRSpec) on the James Webb Space Telescope  II. Multi-object spectroscopy (MOS)

Rawle, T., Giardino, G., Franz, D.E. et al. 2022 SPIE 12180, 121803R
In-flight performance of the NIRSpec micro shutter array

Notable updates
    Updated numbers for shutter operability and replaced Fig 5, based on March 2024 results of Peter Zeidler

    • Improved MSA layout diagrams (Figure 1 and Figure 4)
    • Expanded the MSA shutter operability section and updated the operability statistics
    • Updated the operability map (Figure 5)

  • Changed wording to reflect that MOS planning around failed open shutters is optional.

  • Updated gap sizes based on current best information.
Originally published