Building a Molecular Workbench: Underpinning Core Science Concepts
Dr. Boris Berenfeld, Dan Damelin and Dr. Qian Xie
Concord Consortium
http://www.concord.org
To develop a scientific way of reasoning and to understand macroscopic phenomena discovered by modern science and their applications to practical fields such as nanotechnology, molecular genomics or proteomics, students need robust mental models of atoms, molecules, and their interactions. Such understanding allows students to grasp what lies at the heart of modern biology, chemistry, and physics. There is considerable evidence that students have major misconceptions about atoms and molecules, and few ideas about how the forces and motions at this scale relate to macroscopic properties.
The Molecular Workbench project (http://www.concord.org/workbench/index.html), which is funded by the US National Science Foundation, is studying the effectiveness of using highly manipulable molecular dynamics models to enhance student learning. It facilitates learning through the use of guided inquiry about the connections between microscopic and macroscopic properties of materials. To this purpose we are creating atomic-scale models, several macro-scale simulations with interactive Flash applications, and are using an integrating language called Pedagogica. In collaboration with teachers, instructional materials using these tools are now being tested in a variety of middle and high school classrooms.
Using our Molecular Workbench model we can already demonstrate states of matter, gas laws, phase change, latent heat, diffusion, gas absorption, osmosis, thermal diffusion, conformational changes, and some of the properties of liquid crystals. Using these models and supporting curricula, students are able to predict new micro-macro connections.
The Molecular Workbench2D and Molecular Workbench3D (formerly called Oslet) are molecular simulation engines developed for the project. MW2D is written in pure Java. MW3D is written in Java and Java3D. The models are based on molecular mechanics, an important part of contemporary computational chemistry and biology. MW2D develops an object-oriented framework (OOF) for interactive molecular simulations. The core of OOF provides software abstractions of basic objects for atoms, molecules, fields, boundaries, data sets, and the interactions among them. MW2D consists of four major functional modules: a builder, a viewer, a simulator, and a data analysis environment. The software abstraction model in the OOF provides the interfaces for these modules to interact and integrate.
Pedagogica, which was developed by Paul Horwitz originally to control a genetics simulation called BioLogica, has now been interfaced with MW2D. Pedagogica simplifies the conversion of open-ended models and tools into learning activities that support guided inquiry and generate specific, useful assessment data. We call the combination of a model with Pedagogica a "hypermodel".
Pedagogica is a client application that controls models and their appearance, displays what options are available, receives input concerning the state of the model, controls the interactions with the user, and coordinates other resources that might be used with the model. Pedagogica not only facilitates student choices, but also queries the users and monitors their progress.
Curriculum activities encourage students to uncover key attributes of molecular systems, develop their own models, and test these against a professional model. Proprioceptive activities with visual feedback, a “virtual construction kit” to arrange molecules in polymer chains, explorations with interactive models followed by discussions revealing the explanatory and predictive power of the models -- all work to build and enrich the student's mental models of the connection between macroscopic phenomena and molecular worlds.
Eight curriculum activities on Atoms in Motion, and a week-long “mini-unit” on States of Matter, both based on atomic-scale models have undergone formative testing and are now undergoing summative testing in public middle school 8th grade classes and in 9th grade classrooms . The activities include integrated hands-on labs, computer modeling, and kinesthetic modeling activities. New research modules on Solutions, and Monomers to Polymers will be tested this Spring and Fall.
Our primary research questions are:
- Does experience with such a tool help students in 8-9th grades internalize accurate mental models of atomic-scale phenomena, and
- Does this in turn help them become more expert in reasoning about phenomena at different levels, from micro to macro.
We are conducting our research now. Students to date have been highly engaged, responsive in discussions, and have made connections between the various types of activities conducted. For example, they performed a computer simulation and a kinesthetic simulation of kinetic energy exchange between gaseous atoms during collisions—in other words, heat transfer through direct mixing of atoms. Their written responses regarding these activities indicated that most found both the computer and kinesthetic experiences helpful in understanding the concept and that doing both of them was better than doing one or the other.
For more details about the project, including the software, images and descriptions of activities, scripted hypermodels, and curricula, see: http://www.concord.org/workbench/download.html
Boris Berenfeld, co-PI: Dr. Berenfeld oversees the curriculum design and development. With a Ph.D. in radiation biophysics, he combines his interest in frontier science, curriculum design and networking technologies to advance the quality and accessibility of science education. His current research involves innovative approaches to teaching biochemistry through visualization. He has led innovative educational initiatives in Russia, Europe, Latin America, United States and in the Middle East, organized several East-West Conferences on New Technologies in Education, and served on the faculty of the Salzburg Seminar.
Qian Xie at the Concord Consortium is the primary scientist on the project, responsible for developing the algorithms and designing the Molecular Workbench code. Dr. Xie is a computational scientist with a PhD in Materials Physics from University of Science and Technology, Beijing, and post-doctoral experience at the Max Plank Institute in Dresden and at the University of Cyprus. He has developed the Molecular Workbench core software.
Dan Damelin teaches Chemistry and coordinates the computer lab at Lincoln-Sudbury High School. With a MAT from Tufts in Chemistry, Biology and General Science, and a triple major BA from Tufts that included computer science, he has been able both to develop and to test Molecular Workbench activities.