**2. Materials and Methods**

Web of Science (WoS) and Scopus are the most important databases of scientific literature. However, there are works comparing them and concluding that using Scopus, the largest database of peer-reviewed scientific articles in the world, is the best option [63]. Studies determine that, while 54% of Scopus titles are indexed in WoS [64], 84% of WoS titles are also indexed in Scopus. Thus, comparative studies conclude that it is much more effective to use Scopus than WoS in bibliometric analysis. Therefore, in bibliography, a large number of bibliometric works using this database can be found [64,65]. Scopus contains almost 40,000 titles belonging to more than 10,000 publishers, and the analysis of these allow the scientific community to identify where, who, when, and how research is taking place in a given scientific area. Therefore, Scopus was the database chosen by us to carry out this analysis.

In the present analysis, a full search of the Elsevier Scopus database was conducted using [TITLE-ABS-KEY (Wastewater and "advanced oxidation")] as the search query. This resulted in 3208 documents obtained between 1990, the year of first publication, and 2018, the last full year from Scopus database. If the search criteria are modified, the results obtained can be significantly different. Similarly, continuous updates or modifications of the database may also result in certain differences in the result obtained. On the other hand, it must also be taken into account that some of the elements analyzed are the keywords entered by the author or the editor, and these do not always fit the subject matter of the articles. These small anomalies do not invalidate the methodology, and Scopus is considered the best option in bibliometric analysis.

In analyzing the keywords, it was considered that there are terms mentioning the same concept. For example, ozonation and ozonization. Therefore, these keywords were considered as one. On the other hand, keywords that do not contribute to the analysis, such as "article", were discarded. The items

analyzed were: evolution of scientific output, publication distribution by countries and institutions, and keywords. The software tool VOSviewer (http://www.vosviewer.com/) was used in the detection of scientific communities. These scientific communities are represented in graphs, and they are studied as nodes that establish connections between them giving rise to complex networks that determine the relationships within the whole.
