diff --git a/src/eo_datascience/clean_nb.py b/src/eo_datascience/clean_nb.py
index 2c3a8b8..8e9fe03 100644
--- a/src/eo_datascience/clean_nb.py
+++ b/src/eo_datascience/clean_nb.py
@@ -1,6 +1,7 @@
 import os
 import nbformat
 from pathlib import Path
+import re
 
 def clean_up_frontmatter(dir = './notebooks', save=False):
     # Define the path to the notebooks
@@ -41,14 +42,27 @@ def convert_refs(dir="./notebooks", save=True):
         for i in range(len(nb.cells)):
             if i != 0:
                 if nb.cells[i]["cell_type"] == "markdown":
-                    nb.cells[i].source = nb.cells[i].source.replace(r"[@", r"{cite}`").replace(r"]", r"`")
-        
+                    nb.cells[i].source = quarto_ref_person_replace(nb.cells[i].source)
+                    nb.cells[i].source = quarto_ref_time_replace(nb.cells[i].source)
+
         # Save the notebook
         if save:
             nbformat.write(nb, nb_path)
         else:
             return nb
 
+def quarto_ref_person_replace(quarto):
+    bibs = re.findall(r"(?<=\[\@)[^\]]+", quarto)
+    for i in bibs:
+        quarto = re.sub(r"\[\@" + i + "\]", r"{cite:p}`" + i + "`", quarto)
+    return quarto
+
+def quarto_ref_time_replace(quarto):
+    bibs = re.findall(r"(?<=\@)[^\s]+", quarto)
+    for i in bibs:
+        quarto = re.sub(r"\@" + i, r"{cite:t}`" + i + "`", quarto)
+    return quarto
+
 def find_ipynb(dir):
     root = Path(dir).resolve()
     nb_paths = [root / file for file in os.listdir(root) if file.endswith('.ipynb')]
diff --git a/tests/mock.ipynb b/tests/mock.ipynb
index 34c2609..05e623f 100644
--- a/tests/mock.ipynb
+++ b/tests/mock.ipynb
@@ -15,7 +15,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "[@ref1] "
+    "lorem ipsum [@anon2024] and [@anon2025] and lorem ipsum @anon2024 and @anon2025"
    ]
   }
  ],
diff --git a/tests/test_quarto_nb_conversions.py b/tests/test_quarto_nb_conversions.py
index fcbeae9..87f405a 100644
--- a/tests/test_quarto_nb_conversions.py
+++ b/tests/test_quarto_nb_conversions.py
@@ -1,10 +1,16 @@
 import nbformat
 from pathlib import Path
 import pytest
-from eo_datascience.clean_nb import clean_up_frontmatter, convert_refs
+from eo_datascience.clean_nb import clean_up_frontmatter, convert_refs, quarto_ref_person_replace, quarto_ref_time_replace
 
 def test_remove_front_matter():
     assert clean_up_frontmatter("./tests", False)["cells"][0]["source"] == "# This a mock Jupyter file\nWe use it for testing\n"
 
+def test_ref_regex():
+    quarto = [r"lorem ipsum [@anon2024] and [@anon2025]", r"lorem ipsum @anon2024 and @anon2025"]
+    quarto[0] = quarto_ref_person_replace(quarto[0])
+    quarto[1] = quarto_ref_time_replace(quarto[1])
+    assert quarto == [r"lorem ipsum {cite:p}`anon2024` and {cite:p}`anon2025`", r"lorem ipsum {cite:t}`anon2024` and {cite:t}`anon2025`"]
+
 def test_conversion_of_refs():
-    assert convert_refs("./tests", False)["cells"][1]["source"] == r'{cite}`ref1` '
\ No newline at end of file
+    assert convert_refs("./tests", False)["cells"][1]["source"] == r"lorem ipsum {cite:p}`anon2024` and {cite:p}`anon2025` and lorem ipsum {cite:t}`anon2024` and {cite:t}`anon2025`"
\ No newline at end of file