Gwendolyn Midlo Hall wrote a book about this called Slavery and African Ethnicities in the Americas: Restoring the Links.
By looking at the manifests of slave ships, she concludes that slavers had preferred "hunting grounds" where they established links with local chieftans and had a low risk of hostility from people on the African coast. At the other end, thorough documents of colonial governments or plantations, she notes a clustering effect.
For instance, she points to a large population of Akan and Coromanti slaves in Jamaica, predominately Kongo slaves in South Carolina, and large populations of slaves from Upper Guinea and Igbo slaves in the Chesapeake region. A proposed reason for this clustering effect is the perceived characteristics of slaves from different regions. For instance, slaves from Senegambia and the Gold Coast were brought to Brazil for their ability to mine for gold. Likewise, Spanish officials preferred Senegambian slaves for their skill at blacksmithing.
She extends this proposition to argue that such clusters of ethnicities would have a "founder effect" and that later populations would have to assimilate to some degree into the established slave culture that they encountered. That last point is not uncontroversial, and she directly contradicts older narratives such as Rhett Jones' 'Sub-Africanites in Africa and the Americas' where he contends that the extended process of plantation slavery in the americas forced slaves to shed distinctive elements of Igbo or Akan or Wolof etc customs and identities for synchronized 'african' customs and identities.
Leaving that debate aside for a moment, there are certainly moments that we can point to as evidence of African national identities manifesting themselves in the Americas.
One instance would be the Stono Rebellion of 1739. In African Dimensions of the Stono Rebellion John Thornton argues that colonial documents mention details such as slaves coming from Angola, their ability to somewhat understand spanish, familiarity with muskets and practice of marching with banners and drums point to a core of slaves who were former soldiers in the Kongo civil war. Their facility with spanish is explained by the Kongo kingdom's extensive trade and religious contacts with Portugal since the 1400s.
Another example would be the Bahia revolt in 1835. In the decade before the revolt, that area of Brazil saw a large influx of slaves, including a large number of Muslim Yoruba slaves. In court depositions after the revolt, participants told how they would meet together to eat, as part of a custom to eat meals prepared by Muslim hands. Additionally, behind closed doors participants would wear an Abada, which is a white shirt or frock. During the revolt, participants wore the Abada in order to identify their co-conspirators.
Certainly, there were participants in the Bahia revolt that were not ethnic Yorubas, but these depositions do demonstrate how communities could maintain their religious identity in Brazil in secret.
1 Slavery and African Ethnicities in the Americas by Gwendolyn Midlo Hall. 2005, University of North Carolina Press
2 "African Dimension of the Stono Rebellion" by John Thornton, The American Historical Review 96.4 (1991): 1101–1113.
3 Slave Rebellion in Brazil: the Muslim Uprising of 1835 in Bahia by Joao Jose Reis. 1993, Johns Hopkins University press.