idref="DRAWINGS">FIG. 14 as 174) is a second segmented optical element that consists essentially of a first segment (174-1) of a first refraction index through which the pair of second center rays (including a third pair of marginal rays 68-2 and a fourth pair of marginal rays 69-2) propagate and a second segment (denoted in FIG. 14 as 174-2) of a second refraction index through which the two pairs of marginal rays propagate, wherein the first refraction index differs from the second refraction index.

According to an embodiment, the first spherical telescope is configured to provide a demagnified image (the demagnification factor may range between six to twelve or may equal ten or may be of any other value) of the entrance pupil at a focal plane of a first spherical telescope output lens.

According to an embodiment, the second spherical telescope is configured to provide a demagnified image of the entrance pupil at a focal plane of a second spherical telescope output lens.

According to an embodiment, the first collimated relay includes a first relay input spherical lens and a first relay output spherical lens.

According to an embodiment, the second collimated relay includes a second relay input spherical lens and a second relay output spherical lens.

According to an embodiment, the auto-focus system includes a pair of first prisms located at the focal plane of the first spherical telescope output lens.

According to an embodiment, the auto-focus system includes a pair of second prisms located at the focal plane of the second spherical telescope output lens. The pair of second prisms moves the rays of the first pair of rays away from each other.

According to an embodiment, the auto-focus system includes a first collimated relay movement mechanism that is configured to change a distance between the first relay input spherical lens and the first relay output spherical lens.

According to an embodiment, the first collimated relay movement mechanism moves the first relay output spherical lens and also moves a first sensor—to maintain a distance between the first relay output spherical lens and the first sensor. According to an embodiment, the first collimated relay movement mechanism includes a first motor.

According to an embodiment, the auto-focus system includes a second collimated relay movement mechanism that is configured to change a distance between the second relay input spherical lens and the second relay output spherical lens.

According to an embodiment, the second collimated relay movement mechanism moves the second relay output spherical lens and moves a second sensor—to maintain a distance between the second relay output spherical lens and the second sensor. According to an embodiment, the second collimated relay movement mechanism includes a second motor.

According to an embodiment, different distances between the relay input spherical lens and the relay output spherical lens are associated with different tradeoffs between a dynamic range of the auto-focus system and a sensitivity of the auto-focus system. Larger dynamic ranges are associated with lower sensitivity. Different distances are associated with different effective focal lengths of the collimated relay.

According to an embodiment, different distances between the first relay input spherical lens and the first relay output spherical lens are associated with different effective focal lengths of the first collimated relay.

According to an embodiment, the first collimated relay consists essentially a first relay input spherical lens and a first relay output spherical lens.

The relay includes a relay input spherical lens and a relay output spherical lens that have imperfections or aberrations so that the optical power at different points of any of the relay spherical lenses differs from each other. The different optical powers introduce a residual optical power even along a secondary axis of any of the relay spherical lenses. According to an embodiment, a value of an optical power of a point of the relay spherical lens is related to a spatial relationship between a center of the relay spherical lens and the point. The optical power is associated with a magnification.

According to an embodiment, the first collimated relay exhibits a main optical power along a primary axis, and when receiving a skew ray (an off-axis ray)—that is oriented to the primary axis of the collimated relay and to the collimated axis of the collimated relay—the skew ray impinges on a point that is located outside the center of the relay spherical lens (an off-center point) and exhibits an optical power along both axes.

The skey-ray is skewed in part due to the non-zero angle between an optical axis formed between the mask and the splitter and to an optical axis of each one of the first branch and the second branch.

According to an embodiment, the auto-focus system includes a selectable depth of field unit that is configured to select a reference focal plane of the auto-focus system.

According to an embodiment, the selectable depth of field unit includes a first selectable depth of field unit—that is selectively positioned within the first branch or outside the first branch. According to an embodiment, the first selectable depth of field unit is a first window selectively positionable within the first spherical telescope. When positioned within the first spherical telescope the depth of field of the first branch is located at a first position, when positioned outside the first spherical telescope the depth of field of the first branch is located at a second position which differs from the second position.

According to an embodiment, the selectable depth of field unit includes a second selectable depth of field unit—that is selectively positioned within the second branch or outside the second branch. According to an embodiment, the second selectable depth of field unit is a second window selectively positionable within the second spherical telescope. When positioned within the second spherical telescope the depth of field of the second branch is located at a third position, when positioned outside the second spherical telescope the depth of field of the second branch is located at a fourth position which differs from the first position.

FIG. 21 illustrates an example of method 501 for auto-focusing on an auto-focus system (10), the method includes:

- a. Illuminating (510) a sample with illumination beams (30) that form multiple spot arrays (40) on the sample, wherein the spot arrays include a first set of spot arrays (42) formed on a first side of an imaging area (39), and a second set of spot arrays (44) formed on another side of the imaging area (39).
- b. Collecting (520) collected beams emitted from the sample along a collection path that includes an entrance pupil.
- c. Focusing (530) the collected beams along a first axis (91) while imaging the entrance pupil along a second axis (92) to provide optically processed beams.
- d. Generating (540) detection signals (100) that represent the optically processed beams.
- e. Determining (550) a focus state of an evaluation beam (38) that impinges on the imaging area (39).
- f. Responding (590) to the focus state. The responding may include changing the focus state of the evaluation system.

According to an embodiment, steps 510-550 are repeated multiple times, at different points of time.

According to an embodiment, step 540 followed by step 560 of generating, an initial auto-focus estimate of the future focus state of the evaluation beam obtained when the evaluated beam reaches a defined position of the first set of spot arrays. Step 560 may be included in step 590.

For example—assuming that the sample is scanned so that the first set of spot arrays precedes the evaluation beam-then the initial auto-focus estimate may be based on the focus information embedded in the first set of spot arrays. The initial auto-focus estimate may equal the focus state as reflected by the first set of spot arrays—or may differ from the focus state as reflected by the first set of spot arrays.

For example—assuming that the sample is scanned so that the second set of spot arrays precedes the evaluation beam—then the initial auto-focus estimate may be based on the focus information embedded in the second set of spot arrays. The initial auto-focus estimate may equal the focus state as reflected by the second set of spot arrays—or may differ from the focus state as reflected by the second set of spot arrays.

According to an embodiment, method 501 includes step 570 of updating the initial auto-focus estimate in timing proximity to the reaching of the imaging area (39) to the defined position. Step 570 may be included in step 590.

The updating may be based on the focus state of the imaging area (or on the focus of any of the sets of sport arrays) obtained during one or more iterations of steps 510-550 that follow the iteration that provided the initial-auto-focus estimate.

See, for example, FIG. 2. The right part of FIG. 2 illustrates sample 99 and the first and second sets of spot arrays at a first point in time during which an initial auto-focus estimate of the focus state of imaging area 39 is made—based on the focus information focus information embedded in the first set of spot arrays. Before the imaging area 39 reaches the location (at the first point of time) of the first set of spot arrays—step 570 may update the initial auto-focus estimate. The left part of FIG. 2 illustrates the imaging area 39 as reaching the location (at the first point of time) of the first set of spot arrays.

According to an embodiment, each one of the multiple spot arrays (40) and the first set of spot arrays (42) is staggered along the first axis (91) and the second axis (92).

According to an embodiment, one, some or all spot arrays are linear spot arrays.

According to an embodiment, one, some or all spot arrays are nonlinear spot arrays.

According to an embodiment, the optically processed beams form pair of spots per each one of the multiple spot arrays, wherein a distance between spots of each pair (at a sensor plane) is indicative of a focus state associated with a corresponding spot array.

According to an embodiment, step 550 includes ignoring detection signals (100) based on sample elements illuminated by at least a part of the illumination beams (30). The ignoring may be based on design information or other type of information indicative of regions of the sample (for example highly dense logic element region) that once illuminated will provide a low signal to noise ration signals and/or diffracted signals, and the like.

According to an embodiment, step 550 includes determining at least one of a pitch angle and a roll angle of the illumination beams. Any of these angles is detectable by comparing focus information from different spot arrays.

According to an embodiment, at least one of method 500 and method 501 includes adjusting a distance between optical elements in a collection path of the evaluation system to optimize the focus state of the evaluation beam.

According to an embodiment, adjusting the distance between optical elements includes moving a lens or a mirror in the collection path of the evaluation system.

According to an embodiment, the method includes adjusting the intensity of the illumination beams used by the evaluation system to optimize the focus state of the evaluation beam.

According to an embodiment, adjusting the intensity of the illumination beams of the evaluation system includes controlling the power of a light source or adjusting the aperture of an illumination path of the evaluation system.

According to an embodiment, at least one of method 500 and method 501 includes determining a depth of field of the auto-focus system based on the focus state of the evaluation beam.

According to an embodiment, at least one of method 500 and method 501 includes adjusting the position of the imaging area based on the determined depth of field.

According to an embodiment, at least one of method 500 and method 501 includes compensating for aberrations in the collection path of to improve the focus state of the evaluation beam 38.

According to an embodiment, compensating for aberrations includes adjusting the position or shape of one or more optical elements in the collection path of the evaluation system.

According to an embodiment, at least one of method 500 and method 501 includes determining a focus metric based on the detection signals to quantify the focus state of the evaluation beam. The focus metric may be indicative of the focus along any axis, a relative focus error between beams, and the like.

According to an embodiment, step 590 includes adjusting the focus state of the evaluation beam based on the determined focus metric.

According to an embodiment, step 590 includes capturing, by an evaluation system associated with the auto-focus system, an image of the sample based on the focus state of the evaluation beam.

According to an embodiment, step 590 includes analyzing the captured image to extract information about the sample.

According to an embodiment, step 590 includes adjusting the position or orientation of the sample based on the focus state of the evaluation beam.

According to an embodiment, step 590 includes determining a focus error signal based on the detection signals (100) to provide feedback for adjusting the focus state of the evaluation beam.

According to an embodiment, step 590 includes using the focus error signal to control the position or movement of one or more optical elements of the evaluation system.

According to an embodiment, there is provided an auto-focus system (10), that includes:

- a. Illumination path that is configured to illuminate a sample with illumination beams that form multiple spot arrays on the sample that includes a first set of spot arrays formed on a first side of an imaging area, and a Second set of spot arrays formed on another side of the Imaging area.
- b. Controller.
- c. A collection path that includes:
  - i. A first branch that includes a first spherical telescope, a first field curvature compensator, a pair of first prisms, and a first collimated relay that consists essentially of a first relay input spherical lens and a first relay output spherical lens.
  - ii. A second branch that includes a second spherical telescope, a second field curvature compensator, a pair of second prisms, and a second collimated relay that consists essentially of a second relay input spherical lens and a second relay output spherical lens;
  - iii. A shared module that is configured to (i) receive first collected beams and second collected beams, (ii) optically process the First collected beams to provide to the first segment, a pair of first rays per first collected beam, and (iii) optically process the second collected beams to provide to the second segment, a pair of second rays per second collected beam;
- d. A sensor that follows the first branch and the second branch and is configured to generate detection signals indicative of light outputted by the first branch and of light outputted by the second branch.

According to an embodiment, the shared module includes a mask that is located at an entrance pupil, the mask includes a pair of off-axis slits for truncating each collected beam to provide a pair of rays per each collected beam.

According to an embodiment, the shared module further includes a splitter that is configured to (i) direct to the first branch, first rays associated with the first collected beams, and (ii) direct to the second branch, rays associated with the second collected beams.

According to an embodiment, the first collimated relay exhibits a main optical power along a primary axis and exhibits a residual optical power along a secondary axis.

According to an embodiment, the auto-focus system includes a selectable depth of field unit that is configured to select a reference focal plane of the auto-focus system.

According to an embodiment, the selectable depth of field unit includes a first window selectively positionable within the first spherical telescope and a second window selectively positionable within the second spherical telescope.

According to an embodiment, there is provided a system, that includes:

- a. An illumination path that is configured to illuminate a sample with illumination beams that form multiple spot arrays on the sample that includes a first set of spot arrays formed on a first side of an imaging area, and a second set of spot arrays formed on another side of the imaging area.
- b. A sensor.
- c. Controller.
- d. A collection path that is configured to receive first collected beams and second collected beams from the sample. The collection path includes:
  - i. A mask that is located at an entrance pupil, the mask includes a pair of off-axis slits for truncating each collected beam to provide a pair of rays per each collected beam;
  - ii. Splitter that includes an optical splitting element that includes (i) a first reflecting facet that is configured to direct to the first branch, first rays associated with the first collected beams, and (ii) a second reflecting facet that is oriented to the first reflecting facet and is configured to direct to the second branch, rays associated with the second collected beams. The first reflecting facet and the second reflecting facet are located at a reflecting facets location that (a) is outside an image plain, and (b) in which there is a separation between the first rays and the second rays.

According to an embodiment, the optical splitting element is a prism.

According to an embodiment, the prism is a knife-edge right angle prism.

According to an embodiment, the optical splitting element is shaped and

positioned to prevent vignetting of any of the first rays and the second rays.

According to an embodiment, the optical splitting element is shaped and positioned at a furthest location from the intermediate image plane in which any of the first rays and the second rays are not cut by the prism.

According to an embodiment, the splitter includes a housing and a mechanical interface that is connected to the housing and to the optical splitting element.

According to an embodiment, the housing includes an input opening, a first rays output opening and a second rays output opening.

FIG. 22 is an example of method (600) for field manipulation, the method includes:

- a. Illuminating (610) a sample with illumination beams that form multiple spot arrays on the sample, including a first set of spot arrays formed on a first side of an imaging area and a second set of spot arrays formed on another side of the imaging area.
- b. Collecting (620) first collected beams and second collected beams from the sample.
- c. Truncating (630) each collected beam using a mask (110) located at an entrance pupil, wherein the mask (110) includes a pair of slits (112) that are off-axis slits to provide a pair of rays per each collected beam.
- d. Directing (640) the first rays associated with the first collected beams to a first portion of a system, using a first reflecting facet of an optical splitting element located at a reflecting facets location outside an intermediate image plane. The first portion may be the first branch or any other portion that is not used for auto-focusing.
- e. Directing (650) the second rays associated with the second collected beams to a second portion of a system using a second reflecting facet of the optical splitting element, wherein the second reflecting facet is oriented to the first reflecting facet. The second portion may be the second branch or any other portion that is not used for auto-focusing.

According to an embodiment, steps 640 and 650 are based on a separation between the first rays and the second rays at the reflecting facets location.

According to an embodiment, steps 610-650 are used for auto-focusing a measurement system.

According to an embodiment, steps 610-650 are used for purposes other than auto-focusing.

According to an embodiment, the optical splitting element is a prism.

According to an embodiment, the prism is a knife-edge right angle prism.

According to an embodiment, the method includes shaping and

positioning the optical splitting element to prevent vignetting of any of the first rays and the second rays.

According to an embodiment, the method includes shaping and positioning the optical splitting element at a furthest location from the intermediate image plane in which any of the first rays and the second rays are not cut by the prism.

According to an embodiment, the method includes connecting a mechanical interface to a housing and the optical splitting element of the splitter (130).

FIG. 1 illustrates an example of auto-focus system 10 and evaluation system 11 that uses the auto-focus system 10 to maintain an evaluation beam at a desired focus position on sample 99. The evaluation system 11 includes valuation illumination 11, evaluation system beam splitter 14, and evaluation sensor 13. The evaluation system outputs an evaluation beam 38 that passes through first dichroic mirror 24, and is focused by objective lens 19 to impinge on the sample 99 to form a returned beam that is collected by objective lens 19, is directed by the first dichroic mirror 24 towards the evaluation system beam splitter 14 and then directed to the evaluation sensor 13.

The first dichroic mirror 24, and the objective lens 19 are also sued by the auto-focus system 10.

The auto-focus system 10 includes auto-focus (AF) illumination unit 19, initial mirror 21, illumination/collection beam splitter 22, third mirror 23, mask 110 (having slits 112), splitter 130, first mirror 125, first branch 140, first sensor 50-1, second mirror 126, second branch 170, and second sensor 50-2.

The first branch 140 includes a first spherical telescope 142, a first field curvature compensator 144, and a first collimated relay 146.

The second branch 170 includes second spherical telescope 172, a second field curvature compensator 174, and a second collimated relay 176.

FIG. 1 illustrates the mask pupil first conjugate plane 111 and the mask pupil first conjugate plane 113.

FIG. 2 illustrates sample 99 and multiple spot arrays at two different points in time.

The multiple spot arrays include a first set of spot arrays 42 and a second set of spot arrays 44 that are staggered along the first axis 91 and the second axis 92.

The first set of spot arrays 42 includes a first spot array 42-1, a second spot array 42-2 and a third spot array 42-3.

The second set of spot arrays 44 includes a fourth spot array 44-1, a fifth spot array 44-2 and a sixth spot array 44-3.

FIG. 3 illustrates the relationship between a focus state of a collected beam having wavefront 31 that reaches mask 110 and the rays outputted from the slits 112 that are off-axis slits. When focused, the wavefront 31 is parallel to the mask and the rays that pass through slits are parallel to each other and normal to the mask. When unfocused, the wavefront is either convex (wavefront 32) or concave (wavefront 33) and the rays that pass through slits are oriented to each other and not normal to the mask.

FIG. 4 illustrates that the slits 112 of mask 110 convert each spot array of the first set of spot arrays and of the second set of spot arrays to a pair of spaced apart rays. For values of index x between 1 and 3, spot array 42-x is converted to pair of rays 46-x and 47-x, and spot array 44-x is converted to pair of rays 48-x and 49-x.

The distance, following mask 110, between rays of a single pair of rays is indicative of the focus state related to the pair of rays. FIG. 4 illustrates first distance D1 45-1 between first rays 46-1 and 47-1, second distance D2 45-2 between second rays 46-2 and 47-2, third distance D3 45-3 between third rays 46-3 and 47-3, fourth distance D4 45-4 between fourth rays 48-1 and 49-1, fifth distance D5 45-5 between fifth rays 48-2 and 49-2, and sixth distance D6 45-6 between sixth rays 48-3 and 49-3.

FIG. 4 also illustrates a demagnified image 51 of the entrance pupil at a focal plane of a first spherical telescope output lens.

FIG. 5 illustrates the impact of the wavefront of a ray at the focal plane of a spherical telescope output lens—on how the rays are sensed by a sensor.

The focus stage should be determined based on the distance between the rays that impinge on the sensor. The distance can provide a clear indication of the focus state when the wavefront 1031 is parallel to the sensor. The distance can be ambiguous when the wavefront 1032 is concave or convex 1033—depending on the location of a focal point associated with the ray. In order to resolve the ambiguity—the rays of the pair of rays are spaced apart from each other—by using one prism or a pair of prisms.

FIGS. 6-8 and 9-18 illustrate the auto-focus system or components of the auto-focus system having fixed collimating relays. FIG. 9 illustrates the auto-focus system with collimated relay movement mechanisms that are configured to change the distance between relay input spherical lens and relay output spherical lens of the collimated relays. According to an embodiment, any auto-focus system of FIGS. 6-8 and 9-18 includes the collimated relay movement mechanisms.

FIGS. 6 and 7 illustrate the auto-focus system as including entrance pupil 80, mask 110, splitter 130, first mirror 125, second mirror 126, first spherical telescope input lens 142-1, first selectable depth of field unit 203-1, first field curvature compensator 144, first spherical telescope output lens 142-2, first prisms 145, first relay input spherical lens 146-1, first relay output spherical lens 146-1, first sensor 50-1, second spherical telescope input lens 172-1, second selectable depth of field unit 203-2, second field curvature compensator 174, second spherical telescope output lens 172-2, second prisms 175, second relay input spherical lens 176-1, second relay output spherical lens 176-1, second sensor 50-2.

The top part of FIG. 8 illustrates the auto-focus system having the first selectable depth of field unit 203-1 within the first branch 140 and having the second selectable depth of field unit 203-2 within the second branch 170.

The bottom part of FIG. 8 illustrates the auto-focus system having the first selectable depth of field unit 203-1 outside the first branch 140 and having the second selectable depth of field unit 203-2 outside the second branch 170.

The depth of focus is determined by moving the second selectable depth of field unit and the first selectable depth of field unit—while other optical components of the first branch and of the second branch maintain static-which increases the accuracy of the auto-focus system.

FIG. 9 illustrates the auto-focus system at two points in time—and at two different distances between the lenses of the collimated relay.

The auto-focus system includes:

- a. First lower folding mirror 222-5 and first upper folding mirror 222-4 that are located between the first relay input spherical lens 146-1 and the first relay output spherical lens 146-2.
- b. Second upper folding mirror 222-6 and second lower folding mirror 222-7 that are located between the second relay input spherical lens 176-1 and the second relay output spherical lens 176-2.
- c. First collimated relay movement mechanism 202-1 that is configured to change a distance between the first relay input spherical lens 146-1 and the first relay output spherical lens 146-2.
- d. Second collimated relay movement mechanism 202-2 that is configured to change a distance between the second relay input spherical lens 176-1 and the second relay output spherical lens 176-2.

The auto-focus system of FIG. 8 allows a continuous change of the effective focal length of the collimated relay.

FIGS. 10 and 11 illustrate an example of mask 110, first mirror 125, second mirror 126, first spherical telescope input lens 142-1, and second spherical telescope input lens 172-1, a separation of two pairs of rays by splitter 130, a deflection of pair of rays 66-2 and 67-2 by first mirror 125, and a deflection of pair of rays 68-2 and 69-2 by second mirror 126.

FIG. 12 illustrates that the slits 112 of mask 110 convert each spot array of the first set of spot arrays and of the second set of spot arrays to a pair of spaced apart rays and the splitter spots formed on splitter 130. For values of index x between 1 and 3, spot array 42-x is converted to pair of rays that form splitter spots 66-x and 67-x, and spot array 44-x is converted to pair of rays that form splitter spots 68-x and 69-x.

FIG. 13 illustrates splitter 130A that includes an optical splitting element (such as prism 133) that includes first reflecting facet 131 and a second reflecting facet 132. Splitter 130A also includes housing 134 and mechanical interface. The housing includes housing top 135, input opening 136, first rays output opening 137, and second rays output opening 138.

FIGS. 14 and 15 illustrate the optical processing of pairs of rays by various components of the first branch and of the second branch.

In FIG. 14 the various components include first spherical telescope input lens 142-1, first selectable depth of field unit 203-1, first field curvature compensator that include a first segment 144-1 and a second segment 144-2, first spherical telescope output lens 142-2, first prisms 145, and first relay input spherical lens 146-1, second spherical telescope input lens 172-1, second selectable depth of field unit 203-2, second field curvature compensator 174 that includes third segment 174-1 and fourth segment 174-2, second spherical telescope output lens 172-2, second prisms 175, second relay input spherical lens 176-1.

In FIG. 15 the various components include first field curvature compensator 144, first spherical telescope output lens 142-2, first prisms 145, first relay input spherical lens 146-1, first relay output spherical lens 146-1, first sensor 50-1, second field curvature compensator 174, second spherical telescope output lens 172-2, second prisms 175, second relay input spherical lens 176-1, second relay output spherical lens 176-1 and second sensor 50-2.

FIGS. 14 and 15 illustrates a first pair 401 of rays (that includes first rays 46-1 and 47-1 of FIG. 4), a second pair 402 of rays (that includes second rays 46-2 and 47-2 of FIG. 4), a third pair 403 of rays (that includes third rays 46-3 and 47-3 of FIG. 4), a fourth pair 411 of rays (that includes fourth rays 48-1 and 49-1 of FIG. 4), a fifth pair 412 of rays (that includes fifth rays 48-2 and 49-2 of FIG. 4), and a sixth pair 413 of rays (that includes sixth rays 48-3 and 49-3 of FIG. 4).

FIG. 16 illustrates the optical processing of pairs of rays by various components of the first branch and of the second branch.

The various components include first selectable depth of field unit 203-1, first field curvature compensator 144, first spherical telescope output lens 142-2, first prisms 145, and first relay input spherical lens 146-1.

FIG. 16 illustrates first rays 46-1 and 47-1, first pair