times, slow the sample atoms down by lowering the temperature to, for example, four degrees Kelvin so that the STM can scan the atoms.31
The raster scan allows STM users to convert tip-surface interactions into a camera of sorts. In so doing, the STM does not beam electrons (like a television camera) to assemble an on-screen image; instead, the STM forms what could be called a haptic camera, a “camera haptica” (playing on “camera obscura”), as the STM converts a series of interactions between atoms in the vacuum into a series of spatially arranged data points. The use of interactions as measurements moves the visualizing process closer to the sense of touch than the sense of vision or, more accurately, in a merging of the two senses—to haptic vision—because the tip gathers data about local interactions over a series of contiguous spots, and then presents the interaction data in a two-dimensional matrix, an image form. (The concepts of “camera haptica” and haptic vision are discussed further in Chapter 2.)
The rastering movement plays a role in the design of experiments, as researchers engage in practices that rastering affords. For example, Kandel and Weiss, in the experiment mentioned above, record images “with the tip rastering quickly along different directions, and . . . see a correspondence between this ‘fast scan’ direction and the locations at which atoms in the cluster either attach or detach” (8103). Kandel and Weiss present four versions of the same atoms that have been scanned in different ways, and use these versions to explore atomic properties (8104). Kandel and Weiss’s experiment design is one example of how the tip’s movement forms an experimental tool, allowing researchers to interact with surface atoms through the STM.
The ability of the tip to move up and down in the z direction also affords manipulation, because the tip can measure the three-dimensional electronic or topographic qualities of the surface. The ability to move in the z direction has made the distance between the tip and the sample a critical component of the operation of the STM from the beginning: Soon after Binnig and Rohrer developed the STM, before they achieved atomic resolution, Binnig and Rohrer “had to struggle with resolution, because Au [gold, their sample] transferred from the surface even if [they] only touched it gently with [the] tip” (“From Birth” 398). In 1987, R. S. Becker, J. A. Golovchenko, and B. S. Swartzentruber repeated this “mistake” of touching the surface with the tip; as a consequence, the atoms moved. In a letter published in Nature, Becker, Golovchenko, and Swartzentruber reported “atomic-scale modification” of a sample surface after they applied voltage to the tip. They attributed the modification to the transfer of a tip atom to the sample surface (421). Others repeated the experiment of using voltage pulses to “pin” molecules and atoms to a surface (J. Foster 29–32). In another experiment with the STM, for example, R. C. Jaklevic used the STM to dent a piece of gold, and then re-scan the same sample to measure how quickly the dent filled itself in (six to nine atoms per minute), thus using the STM as a tool to make the event of atomic movement measurable (Jaklevic 659).
The x, y, and z directional dynamics structure what can be seen with the STM. The STM images events through measuring change and movement, much like other visualization technologies; PET scanning, for example, images dynamic processes. The combination of x, y, and z directions that intensifies multi-directional interaction dynamics also affords researchers opportunities to manipulate individual atoms. For example, in another experiment, researchers J. G. Kushmerick et al. moved a nickel atom in order to observe how far (and how) atoms hop from place to place (2983). However, until Eigler and Schweizer published the article in Nature with the “IBM” images, researchers had not announced the manipulation of atoms by actually picking them up, “dragging” them, and repositioning them—a practice that the ability to move in x, y and z directions made possible. Although large-scale or automated manipulation of individual atoms is not a common use of the STM, even as I write this, manipulation of the surface becomes almost encouraged due to the dynamics created by the microscope’s data-gathering process.
The coordinated x, y, and z directions also structure possible interactions between user and atoms through the arrangement of atomic interactions. The focus on measuring a series of interactions—and the ability to intervene, to change the interactions without destroying the sample—all encourage users to participate in the data-gathering process, as the measuring process becomes part of the experiment. The STM, then, is both a data-gathering probe and an experimental probe—and also an image processor, as explained in the next section.
GUI Image Processing Dynamics
The data produced through coordinated tunneling, rastering, and z-direction interactions is arranged in a matrix to create an image of atomic phenomena. Following the general process of scientific visualization, the raw data is arranged and thus transformed into images that viewers can interpret (see, for example, Brodbeck, Mazza, and Lalanne 30–31). The computer makes image creation relatively easy; however, despite the fact that Binnig and Rohrer worked for IBM when they developed the STM, they did not use computers to develop or operate the first STMs. Instead, Binnig and Rohrer created the first STM images by mentally building up atomic visions using an oscilloscope monitor’s two-dimensional trails of the tip’s rastering (Mody, “Intervening”). For a time, Binnig and Rohrer resisted the computer when they first began presenting their work. The first public STM image, outside of their own personal visions, was created from oscilloscope lines that Binnig and Rohrer traced onto cardboard, cut out, and glued together to form a three-dimensional model of the sampled surface (Binnig and Rohrer, “From Birth” 401).32 The history of STM image development suggests that Binnig and Rohrer participated in imaging processes that encompassed or paralleled computer imaging capabilities, but did not entirely stem from use of the computer. (Chapter 3 further discusses imaging practices beyond computer-aided visualization, in relation to pictorial conventions.) However, Binnig and Rohrer switched over to the computer to generate images for publication. Other STM users followed suit when arranging their data, as the GUI (also developed in the 1970s, when the wave of non-lens-based scientific visualization technologies mentioned above occurred) afforded extensions of the interactive dynamics initiated by the data-gathering operations of the STM.
Characteristics of GUI also structure interaction, affecting what is imaged and how the resulting image is communicated. The fact that GUI use is now ubiquitous in scientific and medical visualization technologies (as well as in other technologies) makes the interaction dynamics of the GUI seem almost invisible; however, the interaction dynamics the GUI encourages help trace the influence of the GUI in STM use and in the production of STM images. The GUI affords STM users interaction with on-screen visual objects to manipulate in order to explore, change, or image the data. Interaction can affect the visual objects, as well as the data, with which the user engages. As information visualization researchers Dominque Brodbeck, Riccarde Mazza, and Denis Lalanne explain, with the use of computers (and the GUI), “graphical objects are not static anymore but can be interactively manipulated and can change dynamically” (29). In GUI interactions, the user expects to respond to visual objects (such as elements of images, or icons) as behavioral cues for manipulation, not solely in order to understand their meaning, or signification (Drucker, “Reading Interface” 215).
GUI characteristics affect STM use as well as image processing, as the STM user interacts with the GUI screen image in multiple ways, including while conducting the experiment, interpreting data from the experiment, processing the image for publication, and processing the image further if the image is intended to function outside of scientific journal articles (such as in press releases or on research group web sites). The user’s interactions with other dynamics, such as x, y, or z direction, allow the on-screen image to operate as an interface, thus allowing the image to function as an experiment and to help marshal evidence. GUI interactions structure the space in which the user interacts with the image, but also create a space in which the user interacts with the data and through the image with the nanoscale. The range of possible interactions that GUI enables reinforces the multi-directional and manipulable affordances created by electron tunneling, raster scanning, and z-direction moves.
Images and Experiments
GUI interactions also extend the time
