Using innovative technology tools in organic chemistry education: bibliometric analysis

2.1 Database selection and search query

The database of the study was determined as Scopus because many researchers have used Scopus to trace the articles published in journals (i.e. Gao et al. 2022; García-Tudela & Marín-Marín, 2023) because this is an index and selective full text database of educational research. For instance, Hassan et al. (2022) used Scopus database to illustrate the organic chemistry education’s trends for their bibliometric analysis study. Indeed, Scopus provides at least two times more research metrics than other available abstract and citation databases (Scopus, 2024). Besides, it is one of the databases that was used by the primary bibliometric analysis tools, such as Vosviewer and Biblioshiny. In this study, the data were analyzed through PRISMA (2020) guidelines. There are three phases according to PRISMA (Moher et al., 2009); identification, screening, and inclusion. In the following, each phase is introduced.

Identification: This phase included a data mining process and the control of duplicate publications. No duplicate results were identified in this study. The data was obtained with the following retrieval strategy: (TITLE-ABS-KEY) (“organic chemistry” AND “education”) AND (TITLE-ABS-KEY) (“augmented reality” OR “artificial intelligence” OR “3D print*” OR “virtual reality”). The query yielded a total of 56 publications (see Figure 1).

Figure 1:

Prisma flow diagram for the data selection [adapted from PRISMA (2020)].

Screening: This phase consists of selection according to the inclusion/exclusion criteria and the eligibility process. The inclusion and exclusion criteria of the studies were as below:

1. The screening was limited to the Title, Abstract, and Keywords of the publications.

2. Languages other than English are excluded to be able to read all parts of the article by the authors for the eligibility process.

3. In document type, other than the journal articles are excluded (i.e., proceeding papers, book chapters, editorial materials, meeting abstracts).

4. The publication must focus on using innovative technology tools (i.e., augmented reality) in organic chemistry education.

Removing the languages other than English left 56 publications (see Figure 1). Moreover, having selected the article as the only type in terms of the document type, 38 publications appeared. The reason behind this selection is these ‘journal papers’ are deemed reliable because of peer review. Moreover, publications focusing on research trends in science education have tended to focus on particular journals as the content of their investigations (i.e., Lin et al., 2019). Then, these 38 articles were scrutinized regarding the abstract as well as the titles by two researchers to ensure that they fit the criteria for the study. At this stage, studies were dismissed if the articles were not focused on innovative technology tools in organic chemistry education. For instance, one article, titled “Information and Communication Technologies Combined with Mixed Reality as Supporting Tools in Medical Education” was deleted from the dataset because this article was about the implication of IT in medical education instead of organic chemistry. Moreover, another article named “A Case Study on Using Uncritical Inference Test to Promote Malaysian College Students’ Deeper Thinking in Organic Chemistry” was removed from the list because it does not focus on the implications of IT in organic chemistry learning.

Inclusion: After reading and screening 38 articles, 30 of them were deemed eligible in the final data analysis as the last phase of the PRISMA (2020) protocol, Inclusion. These 30 articles were listed in Appendix 1 in terms of the year of publication, total citations (TC) and type of innovative technology tools used in the organic chemistry education. Appendix 1 could also be useful to see which study is the most cited one in descending order for each innovative technology tool used in this study.

As seen in Figure 2, a considerable proportion of the articles in the dataset were related to using Virtual Reality (30 %), 3D printing (30 %) and Augmented Reality (20 %) in organic chemistry education. Therefore, it would be expected that the articles with the highest citation would follow the same order above, as also presented in Appendix 1.

Figure 2:

The percentages of the innovative technology tools used in 30 articles. Augmented Reality (AR) is represented by blue color, Virtual Reality (VR) by orange color, Artificial Intelligence (AI) by red color, 3D Printing by green color and the study including both VR&AR by purple.

2.2 Data analysis

Performance analysis and Science/Bibliometric mapping were used to discover the research trends of using innovative technology in organic chemistry education. Performance analysis is an established quantitative method for assessing academic output for productivity, quality, and scientific impact by detecting principal contributors. Science or bibliometric mapping analysis presents the structural and dynamic aspects of the data extracted from the research (Börner et al., 2003). In the current study, we used Vosviewer and Biblioshiny in data visualization and in discovering the relations in citations, co-authorship, and bibliographic coupling.

3.1 Overview of the analyzed data set

Table 1 shows the information on 30 articles published in the period between January 2014 and June 2024, which was extracted from the Scopus database. All articles related to innovative technology tools were published by 13 different journals and 123 authors, who used 115 different authors keywords. Scopus records include two types of keywords: Author Keywords, and Keywords Plus. Author Keywords comprise a list of terms that authors have chosen themselves to represent the content of their paper (Li et al., 2009). On the other hand, Keywords Plus are generated automatically by computer algorithm, which consist of words or phrases that can appear anywhere and are not limited to the author’s keywords or content, even including the titles of an article’s references. For this study, we used Authors Keywords because although Keywords Plus is effective for searching articles, it is less comprehensive in representing an article’s content. The average number of co-authors in each publication is 4.33. This value is compatible with the other disciplines (i.e., green marketing, social sciences research) where the most common authorship numbers consist of more than 3 authors (Saleem et al., 2021). Moreover, considering the results obtained from average citations per publication (21.97), we can say that “using innovative technology tools in organic chemistry education” is a topic that has not only gained a place but also solidified its position in the academic sphere. Besides, the percentage of international co-authorships (23.33) demonstrates that a significant portion of the studies conducted in this field have included international collaboration. Similarly, Prahani et al. (2024) also observed high collaboration percentage (21.82) in their bibliometric study, aimed to analyze the scientific creativity literature in the last 20 years. This means that the scientific creativity field is also likely has a high collaboration index. However, some other bibliometric studies in the literature did not find high percentage of international co-authorships. For instance, Mata et al. (2024) indicated moderate level of global collaboration (17.11) in their bibliometrics study focused on new software product development and the researchers stated the necessity of improvement of international cooperation in this field.

Table 1 also includes descriptive statistics for VR, 3D, AR, and AI individually in addition to innovative tool (IT) statistics. As presented in Table 1, 3D printing was the first innovative tool among the selected ones to be used in organic chemistry education while AI is the most recent one to start to be used in this context. VR and AR had the highest average citations per publication.

What is more, looking at the statistics for co-authors per publication and international authorship, the highest values belong to AI. In the other studies made, there have not been observed a relation between being the most recent or a higher value for co-authors per publication and international authorship (i.e., Authors 1, 2). In addition to this, to verify this relationship or lack of relationship, we have also compared the co-authors per publication and international authorship values for 3D printing (which has the oldest average age in this dataset) with AI (which is the most recent one) between the years 2022–2024 in Table 2. As can be seen, the studies with 3D still have lower co-authors per publication and international co-authorship values compared to those of AI. Indeed, the recent studies that are made on 3D do not represent any international collaborations.

In this dataset, we have found three studies about 3D printing for the period 2022 January to 2024 June, all of which belonged to the year 2023. The first of these is titled “Hydrophilic and Conductive 3D-Printed Electrocatalysts in Hydrogen Evolution Reaction for Undergraduate Experiments”. The current affiliation country of all 5 authors of this article is seen to be China. The second article is named “Do-It-Yourself 5-Color 3D Printing of Molecular Orbitals and Electron Density Surfaces”, and its authors are from the USA. The last one is titled “The Approach to Aromatherapy and Essential Oils Extraction as a Generating Theme for Teaching Organic Chemistry”, the three authors of which are from Brazil. This demonstrates the point that there has not been international cooperation for the period. However, analyzing the same situation for AI, international collaboration is more common. For example, the first article that was published in 2022, named “A CASE (Computer-Assisted Structure Elucidation) for Bench-Top NMR Systems in the Undergraduate Laboratory for de Novo Structure Determination: How Well Can We Do?” boasts the collaboration of 12 authors who are from three different countries (Canada, Spain and Germany). Similarly, another example belonging to AI is an article made by China, the USA, and Saudi Arabia, with the name “Data Integration Method of Multi-source Feedback Evaluation for Remote Teaching Quality”. In addition to these studies, it is observed that many other AI studies in the literature are conducted with significant global collaboration. For instance, Polat et al. (2024) conducted a bibliometric analysis to investigate the educational studies on ChatGPT and observed that each publication had approximately four co-authors (n = 3.96). Moreover, the researchers reported international co-authorship in more than one-fourth of the publications (26.42). Radu et al. (2024) also reported 23.54 % of the documents focused on AI and competency-based education have international co-authorship. Furthermore, Fidan and Kasimi (2023) also reported the high percentage of international co-authorships (37.29) in their bibliometric study aimed to examine AI articles in the subject of language education.

3.2 Research productivity in terms of publications and citations

Scopus was used to investigate the distribution of the number and citations of the articles per year. The annual research productivity is demonstrated in Figure 3. According to this graph, the first article appeared in 2014 and was related to using 3D printers in teaching of structure−energy relationships in molecular conformations and in chemical reactions. This article was cited 22 times. Despite the dearth of publications until 2019 as seen in Figure 3, there seems to be a considerable increment in the coming few years. The year 2020 and 2023 had the highest numbers of publications (n = 6), while the highest number of citations (n = 203) were received in the year 2019. As the year 2024 is not finished yet, the drop in this year can be accepted as normal. The number of articles produced after 2020 (during Covid) constituted 76.66 % of total publications. It seems that most of the articles have been published after the Covid started and later. This finding is also supported by the average age of the articles, which is found to be 3.3.

Figure 3:

Number of (a) publications and (b) citations per year from January 2014 to June 2024.

3.3 Leading countries

The total number of countries that have contributed to the studies conducted on the subject was twenty-two. Out of these countries, Figure 4 depicts the research productivity of the most productive countries in terms of publications and citations. Findings identify that the most productive countries were the USA (n = 16), Canada (n = 3) and China (n = 3). The rest of the countries had two or fewer publications. Figure 4 also shows that the USA (n = 374), Malaysia (85), and Canada (n = 73) have the highest citations in the related field. As seen in Figure 4, although China had 3 articles, the number of citations was very low. On the contrary, Malaysia had only one article, their number of citations was one of the highest. The unequal distribution of publications and citations among the countries has been seen in the literature very often (i.e., Authors 1, Kaya-Capocci, 2023). We investigate the descriptive statistics of these two countries as presented in Table 3 to better understand the unequal distribution among the countries (see Figure 4). However, because the number of the articles was very low, it is not easy to say what caused this unequal distribution in this situation. When we looked at the article belonging to Malaysia, we found that the article titled “Haptic Virtual Reality and Immersive Learning for Enhanced Organic Chemistry Instruction” is the most cited publication in this dataset.

Figure 4:

Trends of publications and citations of the countries from January 2014 to June 2024.

3.4 Most productive authors

Table 4 shows the most productive authors with their current institutional affiliation, the total number of publications by each author (TP) and the respective years these publications belong to, the total citations associated with these publications (TC), and the h-index (h). As is seen in Table 4, only seven authors out of 123 authors published more than one paper on the topic. Table 4 also presents the most productive and most cited authors in the dataset. In this table, A.D. Cheok has a very high impact with an h-index of 41, meaning this author has 41 papers with at least 41 citations each. Moreover, top authors’ production over time is also presented in Table 4. It is seen that M. T. Gallardo-Williams and C. L. Dunnagan, the most productive and most cited authors, made their two joint publications in the same year (2020). Moreover, A. J. Dood started publications in this field in 2023 and published in 2024.

3.5 Most influential journals

All articles in the database were published by 13 different journals (see Table 1). Journal of Chemical Education produced 60 % of the articles (n = 18) in the database and the other journals each published only one article on this topic. Moreover, Journal of Chemical Education had the first published article in the database, named “3D printers can provide an added dimension for teaching structure-energy relationships and STEM publications”.

3.6 Most frequently used words in the dataset

115 author’s keywords were detected in 30 articles in the dataset (see Table 1). Figure 5 presents a visualization of the top 15 words that appeared most frequently in the database. While the location of the words within the illustration does not have specific meaning, the prominent words with a larger font size are placed in the middle so that they are more visible. The most repeatedly used keywords were “Organic Chemistry” (n = 20), “second-year undergraduate” (n = 11), “hands-on learning” (n = 8), “computer-based learning” (n = 6), “molecular modeling” (n = 6), “multimedia-based learning” (n = 6), “upper-division undergraduate” (n = 5), and “virtual reality (n = 5)”.

Figure 5:

The word cloud of the author’s keywords illustrates words in various sizes depending on how frequently they appear.

To visualize the top 5 keywords’ frequency over time, Figure 6 was developed using the Biblioshiny with precise graphical parameters: field as “authors’ keyword”, occurrences “per year” with no confidence interval with the top 5 keywords considering their maximum frequencies. As seen in Figure 6, during the first five years of our dataset (2014–2018), “organic chemistry” does not show up as an author’s keyword in any article. However, during the next 5-year period 2019–2023, this number rose from zero to 14. This number is 4 for the first 6 months of 2024. This is a meaningful trend in the number of authors keywords, signifying the place and importance of “organic chemistry” in such studies.

Figure 6:

Mapping the word growth of the top 5 keywords based on the annual occurrence. Each keyword is denoted with an individual color to distinguish in the graph, which are respectively: Computer-based learning (Red), Hands-on learning (dark green), molecular modeling (light green), organic chemistry (blue), and second-year undergraduate (fuchsia).

3.7 Co-authorship network map

Biblioshiny software was employed to conduct a visual analysis of the authors’ collaboration network. For this analysis, Figure 7 illustrates authors who co-wrote at least one article. As seen from the thickest linkage in Figure 7, the strongest collaboration relationships belong to two groups of authors: the first pair includes C. L. Dunnaganand and M. T. Gallardo-Williams while the other group have three authors M. Abdinejad, H. S. Qorbani and S. Dalili. Each group of the authors collaborated on 2 papers together in the dataset. The other authors connected in Figure 7, have only 1 article together.

Figure 7:

Co-authorship network map for showing the authors who co-wrote at least one article together with the thickest linkage showing the strongest collaboration relationships belong to two groups of authors.

3.8 Collaborative networks between countries

Vosviewer software was utilized to visualize the analysis of cooperation relations between countries, and the results are presented in Figure 8. Each country is signified by a circle, the size of which depends on the number of connections produced in that country. As indicated by the color blue, the USA is the most collaborative country in the dataset of this study; hence, given the largest circle. The curve linking the two circles denotes the cooperation between the two connected countries. The thicker the curve is, the stronger the cooperation between the two countries. For this analysis, the minimum number of articles for a country was determined as 1. The 3 clusters developed from 22 countries. The first cluster, the most crowded one, contained three countries (Argentina-Cuba-Switzerland). Clusters 2 and 3 had two countries inside. These coupled countries in these clusters are as follows, respectively; China-Saudi Arabia (Cluster 2); Poland-USA (Cluster 3).

Figure 8:

The co-authorship across all countries denoting each country by a circle, the size of which depends on the number of connections produced, connected by a curve denoting the cooperation between the two connected countries.

3.9 Bibliographic coupling of countries

Figure 9 gives the bibliographic coupling of countries in the dataset. Bibliographic coupling of countries occurs when publications from two countries reference publications from a third country. It is a sign of the similarities between two countries working on a related subject matter. Bibliographic coupling is deemed to be strong or weak depending on the total number of references or citations of other third documents that they share. In Figure 9, the strength of the contributing countries is represented by the size of the blocks. The colors and proximity of the circles have been deliberately used to determine the clusters. Countries with a minimum of 1 document or more were included; 22 of 22 countries meet the criterion. Countries are clustered as one for their close relations in content; clusters’ connections are categorized using quantitative network indicators. Bibliographic couplings of counties in the dataset were classified into 6 clusters. Argentina is placed along with Cuba, and Switzerland in cluster 1 (red cluster). Canada is in cluster 2 with Germany and Spain (green one). The United Kingdom, Portugal and Malaysia are in Cluster 3 (blue one). The United States, Poland and Italy are placed in Cluster 4. Moreover, Cluster 5 contains China and Saudi Arabia. Lastly, France is with Singapore in Cluster 6. Countries with the top bibliographic coupling action included the United States (288 total link strength); Argentina (159 total link strength); Cuba (159 total link strength) and Switzerland (159 total link strength). As shown in Figure 9, the United States has the largest network on the map, and it has the largest block. This means that it is the most productive country for related literature. The mapping analysis of bibliographic coupling of countries suggests that the USA is an influential contributor to IT-Organic Chemistry and that other countries are coupled to the USA.

Figure 9:

Bibliographic coupling of countries divided into different cluster with regards to color and proximity in terms of references; their strength is shown by the size of the blocks.

3.10 Co-occurrence network mapping

With the Vosviewer program, “co-occurrence” and “author keywords” were chosen as the analysis unit. Then, when the minimum repetition count was selected as 2 for keywords, 29 keywords met this threshold. The network structure of the relationship between authors’ keywords is presented in Figure 10. Each circle refers to a keyword. The size of a circle indicates the frequency of the keyword. Clusters of keywords are represented with different colors. Lines refer to co-occurrence links between keywords and the thickness of the line refers to the strength of the relationship between them. Five clusters were observed after the analysis. These clusters were composed of 4–9 keywords. The largest circle of each cluster indicates the dominant keyword. “Organic chemistry” (72 total link strength) for the green cluster, “Second-year undergraduate” (56 total link strength) for the fuchsia cluster, “Hands-on learning” (47 total link strength) for the red cluster, “Molecular modeling” (26 total link strength) for the purple cluster, and “computer-based learning” (24 total link strength) for the blue cluster were the dominant keywords. The words “Organic chemistry”, “Second-year undergraduate”, “Hands-on learning”, “computer-based learning”, “upper-division undergraduate” and “molecular modeling” are situated at the center of Figure 10 because they have the highest connections with others.

Figure 10:

The co-occurrence of the authors’ keywords represents circles referring to a keyword; and the size of a circle depends on the frequency of the keyword while the clusters of keywords are denoted with different colors.