Visualizing molecular structures and shapes: a comparison of virtual reality, computer simulation, and traditional modeling

Background of the study

The fact that molecular geometry of molecules is difficult to visualize makes chemistry a subject challenging to study. Additionally, people see different things when viewing complex structures: what one sees depends on one’s background in chemistry and preparation. “An understanding of the concepts of atom and molecule is fundamental to the learning of chemistry. Any misconceptions and alternative conceptions that students harbor about these concepts will impede further learning.” (Griffiths & Preston, 1992)

Students learn science through a process of constructing, interpreting, and modifying their own representations of reality based on their experiences. According to the constructivist theory of learning, students will try to incorporate the mental representations of objects and concepts, along with the details provided by their instructor, thus adapting their prior knowledge to ultimately build their own understanding (Piaget, 2001). Students generate their own mental images when they hear the material presented in class or reading a text. Development of molecular visualization techniques can be seen as scaffolding methods and are based on constructivism and educational technology design (Clark & Mayer, 2008). In chemistry education, instructors use models to describe submicroscopic elements of matter and interactions that learners cannot see with their naked eye (Taber, 2002), and the use of visualization and development of visualization techniques is “the key to understanding chemistry concepts” (Jones & Kelly, 2015).

A multitude of studies have emphasized the benefits of including physical models, computer models, and animations into the chemistry curriculum (Abraham, Varghese, & Tang, 2010; Claus & Nelsen, 2009; Dori & Barak, 2001; Jones, 2001; Pfennig & Frock, 1999; Ruddick, Parrill, & Petersen, 2012; Sanger, Phelps, & Fienhold, 2000; Wu, Krajcik, & Soloway, 2001). It was concluded that students who were provided with 3-D models as physical models, computer models or even photographs of physical models performed better on the ‘Chemical Visualization Test’ than students who were provided only with 2-D images (Ferc, Vrtacnik, Blejec, & Gril, 2003).

The topic of molecular geometry involves spatial representation and is important in the understanding of the chemical and physical properties of molecules. Molecular structure affects molecular function. Recent studies by Cooper, Underwood, Hilley, & Klymkowsky (2010, 2012 suggest that many students are unable to construct representations of simple molecular structures. Despite different teaching interventions (Gillespie, 2004), it continues to be one of the most challenging topics for the chemistry student. In this exploratory study, we decided to use the topic of the molecular structure and the comparison of three visualization techniques: physical models, PhET simulations, and Virtual Reality.

Physical molecular models allow students to create three-dimensional models in which atoms are represented by spheres of different colors, and bonds are represented by sticks between the spheres. Physical models are low tech and relatively inexpensive. The use of this tool is well documented in the chemistry education literature (Erlina, Cane, & Williams, 2018; Jones, 2001; Pfennig & Frock, 1999). They are useful for representing simple molecules; however, their utility for the representation of molecules is limited. It would be cumbersome and time consuming to use models to represent larger, more complex molecules. In addition, the models of larger molecules could be confusing due to the convolutedness of the models as well as being difficult and somewhat fragile to manipulate.

Computer simulations are tools that may help students in relating, interpreting, and making connections across multiple representations for the same object (Steiff & Wilensky, 2003; van der Meij & de Jong, 2006). PhET interactive simulations are research-based tools for teaching chemistry. The PhET interactive simulations project, hosted by the University of Colorado-Boulder, has developed numerous simulations for teaching chemistry, which are accessible for free online. These simulations are designed to focus on specific ideas and to allow students to explore those ideas through the chosen interactive components (Moore & Perkins, 2014). Due to the fact that the physical molecular models (model kits) and simulations have been traditionally used as visualization tools, they could provide a baseline (“control”) for the study.

While VR technology has been used commonly for entertainment purposes, such as gaming and 3-D theaters, it is gaining attention as an educational and training tool to provide learners with a safe environment, where they can learn and develop skills using virtual elements. Educational applications of VR can be found in medicine, nursing, biology, and chemistry education (Hansen, 2008).

In chemistry education, VR technology could facilitate the learning process by overcoming the major restrictions of traditional educational methods (Georgiou et al., 2007). The use of VR in education in general and also in chemistry is increasing in popularity. Thus far, the VR simulations have included experiences in 3-D molecular structure, chemical bonding, and intermolecular interactions in molecules (Jiménez, 2019). In their research, Dunnagan et al. (2020) evaluated the use of instrumentation in a VR setting and compared it with the use of the real instrument (infrared spectrophotometer) in an organic chemistry lab. The investigation found no significant difference in the learning outcomes between the two groups of students- those that were exposed to the physical instrument and the other group that worked in VR. Other research explored the use of VR technology in the understanding of the enzyme chorismate mutase mechanism and dynamics. A student survey showed that the students found the VR technology more engaging than the traditional approach (Bennie et al., 2019).

Companies such as Hyden–McNeil and Labster have developed VR labs in different areas of chemistry. Even though these labs need to be evaluated and their effect assessed, they could be useful when the attendance of students is challenging due to a pandemic that we are currently experiencing or other reasons that may prevent the students from attending the in-person lab.

Virtual reality technology provides a learner-centered approach, which supports constructivist learning. The technical capabilities of VR are aligned with constructive learning in terms of representation, active learning, and cognition (Chen, 2010). Virtual reality is capable of providing 3-D representations of concepts and processes, which allows students to be motivated and engaged in the learning process. Also, such representations provide individual experiences where learners are allowed to formulate their own perspectives and interpretations. In their research, Georgiou et al. (2007), compared of three chemical experiments: (i) volumetric analysis of acid from base and vice versa, (ii) complexometric neutralization and hardness estimation, and (iii) volumetric reduction–oxidation performed in a physical and VR lab. According to the authors the virtual lab contained a combination of hypermedia and virtual reality features. The authors are emphasizing that in terms of active learning, VR provides learners with a safe environment to freely explore and manipulate a problem, where making a mistake becomes a learning experience with the ability to adjust the situation and move forward with no risk. In such an environment, students learn by doing and constructing their own knowledge. Virtual Reality technology provides interactive, multidimensional visual forms of microscopic elements and allows the user to manipulate these elements in tactile ways, creating an experience that is not possible with other forms of representations (Georgiou et al., 2007).

The use of VR technology in education suggests that it is an effective tool in teaching topics that are proven difficult to carry out in traditional instructional settings and serves as a means of motivating and stimulating students’ understanding of certain concepts (Shim et al., 2003; Limniou, Roberts, & Papadopoulos, 2008). A quasi-experimental study that investigated the use of VR technology in enhancing students’ learning of certain chemistry concepts such as molecular geometry concluded that students had an improved understanding of the 3-D nature of molecules if they did relevant activities in a VR environment compared to those students who used only 2-D images (Merchant et al., 2013).

In distance learning, VR technology is capable of providing students with interactive experiences regardless of their location at any time. This is a significant enhancement over common experiences limited to reading and watching videos or simulations. Virtual chemistry labs have been developed using VR technology, where students can access the lab and perform experiments multiple times, in a safe environment; VR labs could be a cost-effective alternative to traditional labs that are constrained by the cost, substance availability, number of trials and work hours (De Jong, Linn, & Zacharia, 2013)

Research questions

Participants

The focus of this research study was an undergraduate general chemistry laboratory course at a mid-size, western U.S. public institution. The general chemistry course at this institution is taught as a two-semester, four-credit hour lecture, and one-credit hour laboratory each semester. The student population of this course consists of primarily chemistry, sports and exercise science, biology, pre-health, and dietetics majors.

The participants in this study were from three laboratory sections of the first semester general chemistry course. Each group was composed of 15 students. In order to minimize the instructor effect, all three sections were taught by the same instructor. The meeting time of each section was on different days. A traditional approach to teaching and learning was employed in the general chemistry laboratory at this institution. Students were not required to have prerequisite chemistry knowledge in order to enroll in this general chemistry lecture and laboratory course.

Data collection

Data collection took place in the spring semester, 2018, toward the end of the semester with the experiment: Molecular Modeling. The students in each group were exposed to the concept of molecular modeling, though not in great detail, during lectures. Each of the three sections of lab explored the concept by using the same exercise but a different form of instruction. Group 1: composed of eight males and seven females that met at 11:00 am Tuesday morning performed the lab using the regular molecular model kits. Group 2: composed of nine males and six females that met on Wednesday at 2:00 pm was exposed to a PhET interactive simulation: “Molecular Modeling” (https://phet.colorado.edu/en/simulation/molecule-shapes). Group 3: composed of seven males and eight females that met on Thursday at 2:00 pm was exposed to molecules designed in VR. The sample of students is a convenient sample but also random since the students registered in one of the three sections at the beginning of the semester. The age of the students in these three groups varied between 18 and 25 years old.

Prior to beginning the experiment, the students in all three groups were presented with a demographic form (see Appendix 1) and the Purdue Visualization of Rotations Test (Bodner & Guay, 1997). The Purdue Visualization of Rotation Test (PVRT) is a 20-question test. Students were presented with an example of rotation of a 3-D object and then asked to perform the same rotation on a different object and choose the results from five options. Students are allowed 10 min to complete the test. Furthermore, the students that participated in the VR exercise were asked additional yes/no questions in the form of a short survey regarding their familiarity with the VR technology (see Appendix 2). The students in each of the three groups were also asked to take a pre- and post- instruction test. The pre- and post-test was administered to the students with the purpose of measuring the difference of the participants’ ability to solve the exercises correctly before and after using the assigned method and ultimately to assess the learning gains. The pre- and post-test was developed by the researchers and had identical content, asking questions regarding Lewis structure, electron domain, lone pair electrons, molecular geometry, dipole direction, and polarity of the following molecules: H₂S, PH₃, CH₃OH, CS₂, and CH₂O. The chemistry instructor who taught the lab graded both tests at the end of the lab. The pre- and post-test are presented in Appendix 3.

During the Molecular Modeling Lab, students in all three groups were instructed to build several molecules, for example: H₂, BeCl₂, CO₂, HF, H₂CO, SO₂, CH₃Br, NH₃, H₂O, PCl₅, SF₄, ClF₃, SF₆, and XeF₄. For each of these molecules, the students completed a table similar to the one presented in the test with information regarding the Lewis structure, electron domain, lone pair electrons, molecular geometry, dipole direction and polarity of the molecules. The learning activities were comparable across the three platforms: model kits, simulations (PhET), and VR.

In the designing of the experiment, the researchers consulted the PhET simulation and the same molecules were designed in VR. For the instruction using the VR technology, one of the researchers programmed the molecules. In this experiment, the VIVE™|VIVE Virtual Reality System was used. The molecules were built by using open development engines and programs such as: Unity 3-D engine, Godot Engine, and Unreal Engine. Also, a unity software subscription to operate VIVE, that includes custom animations, editing tools, and tools that help track performance reports, was used in the designing of the molecular modeling lab in VR. Students who participated in the VR exercise received instruction and a short time for practicing with the controllers. At the end of the lab, the students in Group 3 were also asked to complete a short survey regarding their experience with VR (Appendix 4).

In all three lab sections, students worked in groups of two, and there was one group of three students. Regardless of the tool used, each group spent on average a total of 1.5 h working the practice problems. In VR, students switched roles while one of the students was in VR and the other or the others (the group of three) were following the work on a screen. Each group spent 15–20 min in VR.

For some students the VR exercise may look like a video game with molecules. Even though the exercise presented in different visualization techniques featured the same molecules, they are different in the visualization abilities. For example, the PhET simulations provides a 2-D display in comparison with the physical and VR models that expose the students to a 3-D representation. In comparison with the PhET simulations, the VR models provide students with a sense of immersion and realism.

Data analysis

To answer the first research question, the scores from the pre- and post-test were analyzed using Statistical Package for the Social Sciences (SPSS), and the mean and standard deviation for each group were reported. An analysis of covariance (ANCOVA) test was performed in order to determine if there is a significant difference among the pretest scores in three groups are correlated to the Purdue Visualization Test (PRVT) performance.

To answer the second research question, the students’ comments in response to each open-ended question (see Appendix 4) in the VR survey were categorized. Another aspect of this research explored the participants’ attitude towards the use of VR for this molecular modeling exercise. Since this is a qualitative aspect, we analyzed the data manually by grouping the responses into several themes.