A ConFlat iris diaphragm device for direct beam control and alignment inside a vacuum chamber

We describe an easy-to-assemble and robust design of a ConFlat (CF) iris diaphragm device installed in a vacuum environment with its aperture size directly adjustable by users outside the vacuum. This design involves commercially available vacuum equipment, 3D-printed but vacuum-compatible components and a minimal need of professional machining to be straightforwardly taken advantage by a wide range of research groups. The iris diaphragm is centered in a 4.5 in. o.d. double-sided CF flange with user-customizable mounting orientation to allow a maximum range of aperture opening from 0.5 mm to 15 mm in diameter. Installation of this device does not require an additional pump for differential pumping across the iris diaphragm. The functionality of this device is examined at a pressure of ∼7 ×10−9 Torr to provide continuous control on the cross section of a light beam passed through the aperture.We describe an easy-to-assemble and robust design of a ConFlat (CF) iris diaphragm device installed in a vacuum environment with its aperture size directly adjustable by users outside the vacuum. This design involves commercially available vacuum equipment, 3D-printed but vacuum-compatible components and a minimal need of professional machining to be straightforwardly taken advantage by a wide range of research groups. The iris diaphragm is centered in a 4.5 in. o.d. double-sided CF flange with user-customizable mounting orientation to allow a maximum range of aperture opening from 0.5 mm to 15 mm in diameter. Installation of this device does not require an additional pump for differential pumping across the iris diaphragm. The functionality of this device is examined at a pressure of ∼7 ×10−9 Torr to provide continuous control on the cross section of a light beam passed through the aperture.


I. INTRODUCTION
Iris diaphragms are commonly used on optical tables for precisely aligning laser beams in a user-defined path. Placing it directly in front of a sensor can limit the amount of detected photons in order to prevent damage by oversaturation. These properties of an iris diaphragm come from the smooth adjustment of a circular aperture via controlled movements of a set of well-positioned thin metal leaves. However, many research groups around the globe perform experiments using laser and/or molecular beams in vacuum (with examples of recent publications given here [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17], and the alignment of the laser beam path inside a vacuum space is usually achieved via placing two iris diaphragms outside the optical windows on both ends of the vacuum chamber. If one of the windows is replaced by a detector, precise re-alignment would inevitably involve vacuum breaking. In cases where the laser beam has a long traveling path inside the vacuum chamber, a small angle of deviation at the optical window could result in a completely off alignment at the target. It would thus be preferable if an iris diaphragm can be installed in vacuo along the beam traveling path or directly in front of a highly sensitive detector, and it is imperative that its aperture size (the movement of the iris leaves) can be finely controlled outside the vacuum chamber. Such device can also be used for atomic or molecular beam experiments in terms of controlling the amount of atoms/molecules entering the interaction region. Major features of this device would necessarily include 1 an easy-to-construct design; 2 vacuum compatibility for all the constituent materials; and 3 a completely vented architecture that precludes dead volume within the device and facilitates the original vacuum condition prior to installation.
No such device is currently available from major vacuum equipment manufacturers. As part of an effort to directly align a laser beam towards, as well as to control the amount of molecules arriving at, a UHV cryogenic scanning tunneling microscope, we have designed and built a CF iris diaphragm device with the abovementioned key properties. Design consideration for this device centers around maximum usage of commercially available parts and minimization of professional machine shop work so that fellow researchers can easily adopt this design into their own instrument. 3D-printed but vacuum-compatible materials are used as a straightforward method to make small but intricate components for this device, but these components can also be easily made by a professional machine shop. Diagrams of the key components are provided ARTICLE scitation.org/journal/adv in the supplementary material document. The functionality of this device has been tested below 10 −8 Torr.

II. INSTRUMENT DESIGN
The primary design philosophy for this device is the ease of integration into existing vacuum systems. We chose a double-sided 4.5 in. o.d. CF flange as the base of this device. As a result, it can be coupled with other CF flanges of different sizes via commonly available CF reducers. Part A of Figure 1 provides an overview of the entire device, while its key components are individually presented in parts B to H. Its foundation, a double-sided 4.5 in. o.d. CF flange welded with a 1.33 in. mini top port, is a part of a readily available 4.5 in. UHV viewport shutter from MDC Vacuum Products (part number 454001) and is illustrated in part B of Figure 1 with unused components drawn in blank. This flange can also be individually ordered from MDC (without the shutter blade and the manual rotator) with the part number of 945411. The advantages of using this specific flange include 1 it is commercially available and can be directly adopted into our design with minimal amount of modification; 2 it has a 1.33 in. port welded on the side to be connected to an operating mechanism for out-of-vacuum control; 3 it offers complete venting on both sides of the aperture while providing sufficient coverage of a stainless steel (SS) plate with 0.1 in. thickness inside the flange to securely install the iris diaphragm; and 4 since the inner SS plate is parallel to the cross-section of the 4.5 in. o.d. CF flange, it automatically secures the movements of the aperture leaves on the same plane when the iris diaphragm is mounted parallel to this SS plate.
Part C is a stand-alone, vacuum compatible iris diaphragm made of bare SS case and leafs without black anodization (OptoSigma Corporation part number VIH-15) and is used in our design without any modification. This iris diaphragm is mounted inside the 4.5 in. flange using a tightly-fit holder (part D) that is designed by us and 3D printed using sterling silver (93% silver, 3% tin and 4% copper from Shapeways, www.shapeways.com). This silver material has been previously demonstrated to be UHV compatible. 18,19 The iris diaphragm is secured on the holder using a SS NF 6 − 40 × 1/4 in. allen cap set screw at the bottom of its case (position (a) in Figure 1) with a vent hole that is already drilled by the manufacturer on the iris body (position (b)). With the assistance of the professional machine shop here at Washington State University, we tapped two threaded holes of size NF 1-72 in. on the inside plate of the 4.5 in. flange to securely mount the holder and the iris diaphragm (positions (d)). The positions of these two threaded holes can be user-modified based on the purpose of individual experiment: As shown in part A of Figure 1, the 3D-printed base is installed at an angle vs. the vertical neck (position (e)) connecting the 1.33 in. o.d. port. This particular orientation provides a maximum opening of the aperture of ∼9.5 mm in diameter, which is more than sufficient for our purpose of study. It is however straightforward to install this holding base parallel to the flange neck in order to take advantage of the entire opening range of the aperture from 0.5 mm to 15 mm in diameter. This feature has been embraced in the original design in a way that the length of the 3D-printed base is calculated so that once both of its corners (labeled in red in Figure 1) touch the inner wall of the flange, the center of the aperture is aligned with the center of the flange regardless of the installation angle between the base and the flange neck. Most importantly, the shape of this holder ensures that the motion of the iris leaves is perpendicular to the traveling path of the molecular or laser beam.
Parts E and F in Figure 1 illustrate two small 3D printed silver adapters that are designed to tightly hug the turning knob of the iris diaphragm using two pairs of SS NF 1-72 in. socket cap screws with hex nuts to clamp down at their "wings". The shapes of the inner cavity of these adapters are designed to exactly match the geometry of the iris diaphragm turning knob together for tight fitting. There is also a hole on each adapter (positions (f)) made for venting. Due to the friction of motion for the iris leaves (in opening and closing the aperture) is adjusted via rotating the turning knob (similar to ARTICLE scitation.org/journal/adv adjusting a set screw), a protruding kink (position (g)) is included in the design of the adapters to help glide along the iris diaphragm case without causing further rotation of the knob in operation. There is however a ∼0.5 mm separation between the kink and the iris case to avoid metal-metal grinding. The direct transfer of motion from the user fingers to the iris leaves is accomplished using the combination of a linear motion feedthrough with 2.0 in. stroke distance (part G in Figure 1, This rod is made by the machine shop instead of being 3D printed due to its length being longer than the allowed printing size by Shapeways. Such long length is necessary because the space required for the two set screws at positions (h) of Figure 1 is larger than the inner space of the neck (position (e)) on the 4.5 in. CF flange. In addition, the minimum length of the linear feedthrough measured from its 1.33 in. CF flange is 3.55 in. (same length is found for equal products from other manufacturers). Thus, it is crucial to have two standard 1.33 in. CF nipples (shown semi-transparent in part A) installed in between the linear feedthrough and the 1.33. in. port on the 4.5 in. base flange. Note that the entire device and parts illustrated in Figure 1 represents a plug-in-and-play design concept with the in vacuo functionality as the top priority and the space required outside the vacuum chamber as a minor concern. Following this concept, our design only requires machine shop services to drill two thread holes inside the 4.5 in. flange (positions (d)) and to fabricate a simple aluminum rod (part H). All other parts, including the 3D printed pieces of parts D, E and F, can be ordered, received in mail and used without further modification. It can however be anticipated that using a customized linear feedthrough of almost zero minimum length and a customized neck of bigger i.d. on the 4.5 in. flange will effectively eliminate the necessity of the two 1.33 in. nipples and reduce the out-of-vacuum space required for this device.

III. DEVICE OPERATION
Once assembled, operation of this iris diaphragm device is straightforward. The extension of the rod (part H in Figure 1) into vacuum is directly adjusted by the fine rotation of the linear feedthrough, and the adapters on the iris diaphragm (parts E and F) are correspondingly moved to open or close the aperture. In our particular case the range of diameter change for the aperture from ∼0.5 mm to ∼9.5 mm corresponds to the scale readings on the linear feedthrough of ∼16 mm to ∼50 mm (which is the maximum extension for the linear feedthrough). These starting and ending positions on the feedthrough scale can be customized, but in doing so the length of the rod also needs to be modified. Although being on the same plane (which is perpendicular to the traveling beam path), the motion of the rod is linear while the adapters are moving in a partially circular route. As a result the point of contact on the cylindrical protrusion of the adapters (position (j)) by the rod, or equivalently the angle between the adapter protrusion and the rod, is constantly changing while the device is in operation. We have purposely designed the length of the rectangular hole on the rod (0.200 in.; position (i)) larger than the size of the adapter protrusion (0.080 in. in diameter) in order to prevent clasping between the two parts. Since the region of contact between these parts is not visible once this device is installed in vacuum, forced operation of the linear feedthrough without knowing the parts were clasped could easily damage the iris leaves. Unnecessary friction and silver-aluminum grinding is also effectively avoided by this size difference.
This "loose" fitting at the region of contact nevertheless introduces inactive linear motion for the rod. This inactive motion, however, does not affect the functionality of this device, and the relationship between readings of the linear feedthrough vs. the actual aperture size in diameter is plotted in panel (A) of Figure 2. The physical contact of the rectangular hole and the adapter protrusion at the turn-around positions are shown in panels (B) to (E) as indicated by the red circles, and each panel contains images taken on the feedthrough scale as well as on the iris diaphragm at the same stage of operation. Panels (B) and (C) both correspond to the minimum opening state of the aperture with a diameter of ∼0.5 mm;  The aperture is gradually opened as the linear feedthrough extends from ∼19 mm to ∼50 mm on its scale. Similarly for the closing process, operating the linear feedthrough from its scale reading of ∼50 mm (panel D) to ∼47 mm (panel E) does not move the iris leaves. The aperture is gradually closed as the linear feedthrough retracts from ∼47 mm to ∼16 mm (back to panel B) on its scale. The overall relationship of aperture diameter vs. the feedthrough scale reading is illustrated in panel (A) for our particular set-up, in which (◯) represents the opening process while (∎) represents the closing process. This relationship will vary if one customizes the installing orientation of the base (part D in Figure 1) or the length of the aluminum rod. The two arrows in Figure 2(A) indicate the above-mentioned inactive linear motions at the fully closed and opened stages. The (◯) and (∎) trends can be used to determine the aperture diameter from the feedthrough scale readings when this device is in operation inside a vacuum chamber. Note that as shown in the plot, similar inactive motions are also expected when the operational direction is reversed at any other stages of aperture openings.

IV. FUNCTIONALITY
To examine the functionality of this iris diaphragm device in a vacuum environment, a compact system is assembled and shown in panel (A) of Figure 3, which consists of (a) the device of study in this work; (b) a turbomolecular pump (Pfeiffer HIPACE 80); (d) an UHV pressure gauge (Inficon BPG400); (e) a foreline pressure gauge (Inficon TV90) and a dry scroll pump (Edwards nDXS10i, not shown in the photo). This system is pumped down to ∼7 ×10 −9 Torr within 24 hrs using a turbomolecular pump (with a foreline pressure of ∼2 ×10 −3 Torr), which demonstrates effective pumping across this device without the need of differential pumping.
In order to demonstrate complete opening and closing of the aperture in operation, a simple LED flashlight (which has much larger beam cross-section than a laser) is used as the light source. The LED light source is place in front of the 2.75 in. CF window (part g) (with the end of the flashlight touches the window and at most 0.5 cm from the window to the led bulb), and this window is ∼34.5 cm away from the iris diaphragm device. A piece of paper containing a series of concentric circles with 0.05 in. difference in diameter is placed and centered in front of the exit window (part f) that is ∼4 cm away from the aperture. Four stages representing the passages of different amount of light through the aperture (and further through the piece of paper) is recorded using a camera aiming at the "bull's-eye" and shown in panels (B) → (E) in Figure 3. This closing process from panel (B) to panel (E) corresponds to the feedthrough readings of 47 cm → 37 cm → 27 cm → 17 cm, which is exactly the (∎) progress plotted in Figure 2(A). Figure 1.