Scopus API in Python
By Nick Peinitz, Vincent F. Scalfani, and Avery Fernandez
The Scopus API, provided by Elsevier, offers programmatic access to a comprehensive database of abstracts and citations from peer-reviewed literature. It supports advanced search capabilities, author and affiliation retrieval, and citation analysis, facilitating a wide range of academic and research applications.
This tutorial content is intended to help facilitate academic research.
Please see the following resources for more information on API usage: - Documentation - Scopus - Scopus API - Elsevier API Documentation - Elsevier API Use Cases - Terms - Elsevier API Service Agreement - Elsevier API Policy - Data Reuse - Elsevier Research Data Policy - Scopus Platform
NOTE: Please see access details and rate limit requests for this API in the official documentation.
These recipe examples were tested on August 4, 2026.
Setup
Import Libraries
The following external libraries need to be installed into your environment to run the code examples in this tutorial:
We import the libraries used in this tutorial below:
library(httr2)
library(dotenv)
library(xml2)
library(dplyr)
library(purrr)
library(tibble)
Import API Key
An API key is required to access the Scopus API. You can sign up for one at the Scopus Developer Portal.
We keep our API key in a separate file, a .env file, and use the dotenv library to access it. If you use this method, create a file named .env in the same directory as this notebook and add the following line to it:
SCOPUS_API_KEY=PUT_YOUR_API_KEY_HERE
load_dot_env()
API_KEY <- Sys.getenv("SCOPUS_API_KEY")
if (API_KEY == "") {
message("API key not found. Please set 'SCOPUS_API_KEY' in your .env file.")
} else {
print("Environment and API key successfully loaded.")
}
[1] "Environment and API key successfully loaded."
1. Get Author Data
Number of Records for Author
BASE_URL <- "https://api.elsevier.com/content/search/scopus"
query <- "AU-ID(55764087400)"
httpAccept <- "application/json"
data <- tryCatch({
response <- request(BASE_URL) |>
req_headers(
"X-ELS-APIKey" = API_KEY,
"Accept" = httpAccept
) |>
req_url_query(
query = query
) |>
req_perform()
resp_body_json(response)
}, error = function(e) {
message("An error ocurred: ", e$message)
NULL
})
We can take a closer look at the data given back:
List of 1
$ search-results:List of 6
The JSON response is parsed into an R list. The top-level list contains an element named search-results.
str(data[["search-results"]], max.level = 1)
List of 6
$ opensearch:totalResults: chr "31"
$ opensearch:startIndex : chr "0"
$ opensearch:itemsPerPage: chr "25"
$ opensearch:Query :List of 3
$ link :List of 4
$ entry :List of 25
Inside the search-results list, there are six keys: * entry - the actual data we want * link - a link to the API endpoint * opensearch:Query - the query we used to get the data * opensearch:itemsPerPage - the number of items per page * opensearch:startIndex - the starting index of the items * opensearch:totalResults - the total number of results
str(data[["search-results"]]$entry, max.level = 1)
List of 25
$ :List of 24
$ :List of 22
$ :List of 25
$ :List of 24
$ :List of 25
$ :List of 26
$ :List of 26
$ :List of 24
$ :List of 26
$ :List of 24
$ :List of 25
$ :List of 26
$ :List of 26
$ :List of 24
$ :List of 23
$ :List of 24
$ :List of 25
$ :List of 24
$ :List of 24
$ :List of 26
$ :List of 26
$ :List of 26
$ :List of 25
$ :List of 26
$ :List of 24
The entry key contains a list of publication records, we can see that the first publication record in the list like so:
str(data[["search-results"]]$entry[[1]], max.level = 1)
List of 24
$ @_fa : chr "true"
$ link :List of 4
$ prism:url : chr "https://api.elsevier.com/content/abstract/scopus_id/105027280060"
$ dc:identifier : chr "SCOPUS_ID:105027280060"
$ eid : chr "2-s2.0-105027280060"
$ dc:title : chr "Teaching Computer-Assisted Retrosynthesis Reaction Prediction with Open-Source Software"
$ dc:creator : chr "Scalfani V.F."
$ prism:publicationName : chr "Journal of Chemical Education"
$ prism:issn : chr "00219584"
$ prism:eIssn : chr "19381328"
$ prism:volume : chr "103"
$ prism:issueIdentifier : chr "1"
$ prism:pageRange : chr "358-369"
$ prism:coverDate : chr "2026-01-13"
$ prism:coverDisplayDate: chr "13 January 2026"
$ prism:doi : chr "10.1021/acs.jchemed.5c00959"
$ citedby-count : chr "0"
$ affiliation :List of 1
$ prism:aggregationType : chr "Journal"
$ subtype : chr "ar"
$ subtypeDescription : chr "Article"
$ source-id : chr "24169"
$ openaccess : chr "0"
$ openaccessFlag : logi FALSE
Each element of the entry list is itself a named list that contains metadata for a single Scopus record.
entries <- data[["search-results"]][["entry"]]
df <- purrr::map_dfr(entries, function(x) {
x[sapply(x, is.null)] <- NA
x[sapply(x, is.list)] <- lapply(
x[sapply(x, is.list)],
function(z) as.character(jsonlite::toJSON(z, auto_unbox = TRUE))
)
as_tibble_row(x)
})
head(df)
# See the columns of the data frame
colnames(df)
[1] "@_fa" "link"
[3] "prism:url" "dc:identifier"
[5] "eid" "dc:title"
[7] "dc:creator" "prism:publicationName"
[9] "prism:issn" "prism:eIssn"
[11] "prism:volume" "prism:issueIdentifier"
[13] "prism:pageRange" "prism:coverDate"
[15] "prism:coverDisplayDate" "prism:doi"
[17] "citedby-count" "affiliation"
[19] "prism:aggregationType" "subtype"
[21] "subtypeDescription" "source-id"
[23] "openaccess" "openaccessFlag"
[25] "pii" "article-number"
[27] "freetoread" "freetoreadLabel"
[29] "pubmed-id"
# Number of rows
nrow(df)
# We can index data from our new data frame, df.
# For example, create a vector of just the DOIs
dois <- df[["prism:doi"]]
dois
[1] "10.1021/acs.jchemed.5c00959" "10.1080/0194262X.2026.2655257"
[3] "10.1016/j.acalib.2024.102984" "10.1007/s10755-023-09648-7"
[5] "10.29173/istl2766" "10.1515/pac-2022-1001"
[7] "10.1186/s13321-022-00664-x" "10.1007/s10755-022-09636-3"
[9] "10.1515/pac-2022-2019" "10.1021/acs.jchemed.1c00904"
[11] "10.29173/istl2566" "10.5860/crln.82.9.428"
[13] "10.1021/acs.iecr.8b02573" "10.1021/acs.jchemed.6b00602"
[15] "10.5062/F4TD9VBX" "10.1021/acs.macromol.6b02005"
[17] "10.1186/s13321-016-0181-z" "10.1021/acs.chemmater.5b04431"
[19] "10.1021/acs.jchemed.5b00512" "10.1021/acs.jchemed.5b00375"
[21] "10.5860/crln.76.9.9384" "10.5860/crln.76.2.9259"
[23] "10.1126/science.346.6214.1258" "10.1021/ed400887t"
[25] "10.1016/j.acalib.2014.03.015"
# Get a vector of article titles
titles = df[["dc:title"]]
titles
[1] "Teaching Computer-Assisted Retrosynthesis Reaction Prediction with Open-Source Software"
[2] "Engineering Outreach Through Open Data: A Library-Led Model for Research Visibility and Engagement"
[3] "Comparing impact of green open access and toll-access publication in the chemical sciences"
[4] "Citation Metrics and Boyer’s Model of Scholarship: How Do Bibliometrics and Altmetrics Respond to Research Impact?"
[5] "Creating a Scholarly API Cookbook: Supporting Library Users with Programmatic Access to Information"
[6] "The current landscape of author guidelines in chemistry through the lens of research data sharing"
[7] "Visualizing chemical space networks with RDKit and NetworkX"
[8] "The Power Law and Emerging and Senior Scholar Publication Patterns"
[9] "Cheminformatics: data and standards a Pure and Applied Chemistry special issue"
[10] "Using NCBI Entrez Direct (EDirect) for Small Molecule Chemical Information Searching in a Unix Terminal"
[11] "Enhancing the Discovery of Chemistry Theses by Registering Substances and Depositing in PubChem"
[12] "Using the linux operating system full-time tips and experiences from a subject liaison librarian"
[13] "Analysis of the Frequency and Diversity of 1,3-Dialkylimidazolium Ionic Liquids Appearing in the Literature"
[14] "Rapid Access to Multicolor Three-Dimensional Printed Chemistry and Biochemistry Models Using Visualization and Three-Dimensional Printing Software Programs"
[15] "Text analysis of chemistry thesis and dissertation titles"
[16] "Phototunable Thermoplastic Elastomer Hydrogel Networks"
[17] "Programmatic conversion of crystal structures into 3D printable files using Jmol"
[18] "Dangling-End Double Networks: Tapping Hidden Toughness in Highly Swollen Thermoplastic Elastomer Hydrogels"
[19] "Replacing the Traditional Graduate Chemistry Literature Seminar with a Chemical Research Literacy Course"
[20] "3D Printed Block Copolymer Nanostructures"
[21] "Hypotheses in librarianship: Applying the scientific method"
[22] "Recruiting students to campus: Creating tangible and digital products in the academic library"
[23] "Finally free"
[24] "3D printed molecules and extended solid models for teaching symmetry and point groups"
[25] "Repurposing Space in a Science and Engineering Library: Considerations for a Successful Outcome"
# Now a vector of the cited by count
cited_by = df[["citedby-count"]]
cited_by
[1] "0" "0" "5" "8" "3" "4" "89" "6" "0" "3" "0"
[12] "0" "26" "30" "9" "14" "29" "8" "15" "28" "1" "1"
[23] "0" "123" "7"
# Get sum of cited_by
sum(as.numeric(cited_by))
2. Get Author Data in a Loop
Number of Records for Author
# Create text file to store author names and their Scopus AUIDs
writeLines(
c(
"Emy Decker\t36660678600",
"Lindsey Lowry\t57210944451",
"Karen Chapman\t35783926100",
"Kevin Walker\t56133961300",
"Sara Whitver\t57194760730"
),
"author.txt"
)
# Load a list of author names and Scopus AUIDs
author_df <- read.delim(
"author.txt",
header = FALSE,
sep = "\t"
)
author_list <- split(author_df, seq_len(nrow(author_df)))
str(author_list)
List of 5
$ 1:'data.frame': 1 obs. of 2 variables:
..$ V1: chr "Emy Decker"
..$ V2: num 3.67e+10
$ 2:'data.frame': 1 obs. of 2 variables:
..$ V1: chr "Lindsey Lowry"
..$ V2: num 5.72e+10
$ 3:'data.frame': 1 obs. of 2 variables:
..$ V1: chr "Karen Chapman"
..$ V2: num 3.58e+10
$ 4:'data.frame': 1 obs. of 2 variables:
..$ V1: chr "Kevin Walker"
..$ V2: num 5.61e+10
$ 5:'data.frame': 1 obs. of 2 variables:
..$ V1: chr "Sara Whitver"
..$ V2: num 5.72e+10
# Get number of Scopus records for each author
num_records <- list()
for (i in seq_along(author_list))
{
author <- author_list[[i]][1, 1]
authorID <- author_list[[i]][1, 2]
query <- paste0("AU-ID(", authorID, ")")
httpAccept <- "application/json"
tryCatch({
response <- request(BASE_URL) |>
req_headers(
"X-ELS-APIKey" = API_KEY,
"Accept" = httpAccept
) |>
req_url_query(
query = query
) |>
req_perform()
Sys.sleep(1)
data <- resp_body_json(response)
number_of_records <- data[["search-results"]][["opensearch:totalResults"]]
num_records[[length(num_records) + 1]] <- list(
author = author,
authorID = authorID,
number_of_records = number_of_records)
}, error = function(e) {
message("An error ocurred: ", e$message)
num_records[[length(num_records) + 1]] <<- list(
author = author,
authorID = authorID,
number_of_records = NA)
})
}
[[1]]
[[1]]$author
[1] "Emy Decker"
[[1]]$authorID
[1] 36660678600
[[1]]$number_of_records
[1] "24"
[[2]]
[[2]]$author
[1] "Lindsey Lowry"
[[2]]$authorID
[1] 57210944451
[[2]]$number_of_records
[1] "10"
[[3]]
[[3]]$author
[1] "Karen Chapman"
[[3]]$authorID
[1] 35783926100
[[3]]$number_of_records
[1] "23"
[[4]]
[[4]]$author
[1] "Kevin Walker"
[[4]]$authorID
[1] 56133961300
[[4]]$number_of_records
[1] "13"
[[5]]
[[5]]$author
[1] "Sara Whitver"
[[5]]$authorID
[1] 57194760730
[[5]]$number_of_records
[1] "8"
Download Record Data
# Let's say we want the DOIs and cited by counts in a list
cites <- list()
for (i in seq_along(author_list))
{
author <- author_list[[i]][1, 1]
authorID <- author_list[[i]][1, 2]
query <- paste0("AU-ID(", authorID, ")")
httpAccept <- "application/json"
tryCatch({
response <- request(BASE_URL) |>
req_headers(
"X-ELS-APIKey" = API_KEY,
"Accept" = httpAccept
) |>
req_url_query(
query = query
) |>
req_perform()
Sys.sleep(1)
data <- resp_body_json(response)
entries <- data[["search-results"]][["entry"]]
author_df <- purrr::map_dfr(entries, function(x) {
x[sapply(x, is.null)] <- NA
x[sapply(x, is.list)] <- lapply(
x[sapply(x, is.list)],
function(z) as.character(jsonlite::toJSON(z, auto_unbox = TRUE))
)
as_tibble_row(x)
})
# Get the DOIs and cited by counts
dois <- author_df[["prism:doi"]]
cited_by <- author_df[["citedby-count"]]
# Create a list of lists with author, authorID, DOI, and cited by count
for (i in seq_along(dois))
{
doi <- dois[i]
cited <- cited_by[i]
cites[[length(cites) + 1]] <- list(
author = author,
authorID = authorID,
doi = doi,
cited = cited
)
}
}, error = function(e) {
message("An error ocurred: ", e$message)
num_records[[length(num_records) + 1]] <<- list(
author = author,
authorID = authorID,
number_of_records = NA)
})
}
# The cites variable is a list of list with the data
# View data for first 5 records
cites[1:5]
[[1]]
[[1]]$author
[1] "Emy Decker"
[[1]]$authorID
[1] 36660678600
[[1]]$doi
[1] "10.1007/s10755-024-09698-5"
[[1]]$cited
[1] "1"
[[2]]
[[2]]$author
[1] "Emy Decker"
[[2]]$authorID
[1] 36660678600
[[2]]$doi
[1] "10.1016/j.acalib.2024.102858"
[[2]]$cited
[1] "3"
[[3]]
[[3]]$author
[1] "Emy Decker"
[[3]]$authorID
[1] 36660678600
[[3]]$doi
[1] NA
[[3]]$cited
[1] "0"
[[4]]
[[4]]$author
[1] "Emy Decker"
[[4]]$authorID
[1] 36660678600
[[4]]$doi
[1] "10.1016/j.acalib.2022.102648"
[[4]]$cited
[1] "0"
[[5]]
[[5]]$author
[1] "Emy Decker"
[[5]]$authorID
[1] 36660678600
[[5]]$doi
[1] "10.1016/j.acalib.2022.102634"
[[5]]$cited
[1] "4"
# Add to data frame
cites_df <- bind_rows(cites)
cites_df
Save Record Data to a File
Here is one method if you want to loop over author queries and save all Scopus document data to a file:
# Load a list of author names and Scopus AUIDs
author_df <- read.delim(
"author.txt",
colClasses = "character",
header = FALSE,
sep = "\t"
)
author_list <- split(author_df, seq_len(nrow(author_df)))
str(author_list)
List of 5
$ 1:'data.frame': 1 obs. of 2 variables:
..$ V1: chr "Emy Decker"
..$ V2: num 3.67e+10
$ 2:'data.frame': 1 obs. of 2 variables:
..$ V1: chr "Lindsey Lowry"
..$ V2: num 5.72e+10
$ 3:'data.frame': 1 obs. of 2 variables:
..$ V1: chr "Karen Chapman"
..$ V2: num 3.58e+10
$ 4:'data.frame': 1 obs. of 2 variables:
..$ V1: chr "Kevin Walker"
..$ V2: num 5.61e+10
$ 5:'data.frame': 1 obs. of 2 variables:
..$ V1: chr "Sara Whitver"
..$ V2: num 5.72e+10
for (i in seq_along(author_list))
{
authorName <- author_list[[i]][1, 1]
authorID <- author_list[[i]][1, 2]
query <- paste0("AU-ID(", authorID, ")")
httpAccept <- "application/json"
tryCatch({
response <- request(BASE_URL) |>
req_headers(
"X-ELS-APIKey" = API_KEY,
"Accept" = httpAccept
) |>
req_url_query(
query = query
) |>
req_perform()
Sys.sleep(2)
data <- resp_body_json(response)
# Extract the 'entry' data and convert it to a data frame
if ("entry" %in% names(data[["search-results"]]))
df <- bind_rows(data[["search-results"]][["entry"]])
df[] <- lapply(df, function(col) {
if (is.list(col)) {
sapply(col, jsonlite::toJSON, auto_unbox = TRUE)
} else {
col
}
})
# Save to file
filename <- paste0(
gsub(" ", "_", authorName),
"_",
authorID,
"_ScopusData.tsv"
)
write.table(
df,
file = filename,
sep = "\t",
row.names = FALSE,
quote = FALSE
)
print(paste0("Data for ", authorName, " saved to ", filename))
}, error = function(e) {
message("An error ocurred for ", authorName, ": ", e$message)
})
}
[1] "Data for Emy Decker saved to Emy_Decker_36660678600_ScopusData.tsv"
[1] "Data for Lindsey Lowry saved to Lindsey_Lowry_57210944451_ScopusData.tsv"
[1] "Data for Karen Chapman saved to Karen_Chapman_35783926100_ScopusData.tsv"
[1] "Data for Kevin Walker saved to Kevin_Walker_56133961300_ScopusData.tsv"
[1] "Data for Sara Whitver saved to Sara_Whitver_57194760730_ScopusData.tsv"
df_author <- read.delim(
"Karen_Chapman_35783926100_ScopusData.tsv",
sep = "\t"
)
head(df_author)
# Get info about citedby_count
summary(as.numeric(df_author[["citedby.count"]]))
Min. 1st Qu. Median Mean 3rd Qu. Max.
0.00 4.00 6.00 13.88 12.00 71.00
# Get info about publication titles
x <- df_author[["prism.publicationName"]]
c(
count <- sum(!is.na(x)),
unique_vals <- length(unique(x[!is.na(x)])),
top = names(sort(table(x), decreasing = TRUE))[1],
freq = max(table(x))
)
"97"
"11"
top
"Behavioral and Social Sciences Librarian"
freq
"20"
3. Get References via a Title Search
Number of Title Match Records
query <- "TITLE(ChemSpider)"
httpAccept <- "application/json"
tryCatch({
response <- request(BASE_URL) |>
req_headers(
"X-ELS-APIKey" = API_KEY,
"Accept" = httpAccept
) |>
req_url_query(
query = query
) |>
req_perform()
data <- resp_body_json(response)
print(data[["search-results"]][["opensearch:totalResults"]])
}, error = function(e) {
message("An error ocurred: ", e$message)
data <- NULL
})
titleWord_list <- list(
"ChemSpider", "PubChem", "ChEMBL", "Reaxys", "SciFinder"
)
# Get number of Scopus records for each title search
num_records_title <- list()
for (titleWord in titleWord_list)
{
query <- paste0("TITLE(", titleWord, ")")
httpAccept <- "application/json"
tryCatch({
response <- request(BASE_URL) |>
req_headers(
"X-ELS-APIKey" = API_KEY,
"Accept" = httpAccept
) |>
req_url_query(
query = query
) |>
req_perform()
data <- resp_body_json(response)
# Extract the total number of results
numt = data[["search-results"]][["opensearch:totalResults"]]
# Compile saved Scopus data into a list of lists
num_records_title[[length(num_records_title) + 1]] <- list(
titleWord,
numt
)
# Delay 1 second between API calls to be nice to Elsevier servers
Sys.sleep(1)
}, error = function(e) {
message("An error ocurred: ", e$message)
num_records_title[[length(num_records_title) + 1]] <- list(
titleWord,
numt = NA
)
})
}
[[1]]
[[1]][[1]]
[1] "ChemSpider"
[[1]][[2]]
[1] "8"
[[2]]
[[2]][[1]]
[1] "PubChem"
[[2]][[2]]
[1] "115"
[[3]]
[[3]][[1]]
[1] "ChEMBL"
[[3]][[2]]
[1] "72"
[[4]]
[[4]][[1]]
[1] "Reaxys"
[[4]][[2]]
[1] "9"
[[5]]
[[5]][[1]]
[1] "SciFinder"
[[5]][[2]]
[1] "35"
Download Title Match Record Data
titleWord_list <- list(
"ChemSpider", "PubChem", "ChEMBL", "Reaxys", "SciFinder"
)
# Get number of Scopus records for each title search
scopus_title_data <- list()
for (titleWord in titleWord_list)
{
query <- paste0("TITLE(", titleWord, ")")
httpAccept <- "application/json"
tryCatch({
response <- request(BASE_URL) |>
req_headers(
"X-ELS-APIKey" = API_KEY,
"Accept" = httpAccept
) |>
req_url_query(
query = query
) |>
req_perform()
# Delay 1 second between API calls to be nice to Elsevier servers
Sys.sleep(1)
data <- resp_body_json(response)
# Extract the 'entry' data and convert it to a DataFrame
entries <- data[["search-results"]][["entry"]]
if (!is.null(entries))
{
for (entry in entries)
{
doi <- if ("prism:doi" %in% names(entry)) entry[["prism:doi"]] else NA
title <- if ("dc:title" %in% names(entry)) entry[["dc:title"]] else NA
coverDate <- if ("prism:coverDate" %in% names(entry)) entry[["prism:coverDate"]] else NA
# Append to the list
scopus_title_data[[length(scopus_title_data) + 1]] <- list(
titleWord,
doi,
title,
coverDate
)
}
}
}, error = function(e) {
message("An error ocurred for ", titleWord, ": ", e$message)
num_records_title[[length(num_records_title) + 1]] <- list(
titleWord,
doi = NA,
title = NA,
coverDate = NA
)
})
}
scopus_title_data_df <- bind_rows(
lapply(scopus_title_data, function(x) {
tibble(
titleWord = x[[1]],
doi = x[[2]],
title = x[[3]],
coverDate = x[[4]],
)
})
)
scopus_title_data_df
