Java xPath/DOM - When importing a node into a document, xPath query is unable to find new node (and sub node)

I am trying to import a node into a 'new' DOM Document and then execute xPath queries on the newly imported nodes. The XML that I'm working with has no namespaces. I have followed the standard procedure to create a document, as shown:

    Document newDoc = null;
    DocumentBuilderFactory factory = DocumentBuilderFactory.newInstance();
    try {
          DocumentBuilder builder = factory.newDocumentBuilder();
          newDoc = builder.newDocument();   
    } catch (ParserConfigurationException pce) {
      // Parser with specified options can't be built
      pce.printStackTrace();
    }

and to import the nodes...

    // nodeToImport is a node from a prior xPath search (part of a node set)
    Node docToRunXPathOn = newDoc.importNode(nodeToImport, true);

Later, when I try to run an xPath query such as

    XPathFactory xPathFactory = XPathFactory.newInstance();
    XPath xpath = xPathFactory.newXPath();

    if (docToRunXPathOn != null) {
        try {
            XPathExpression expr = xpath.compile("//version_id/text()");
            return (String)expr.evaluate(doc, XPathConstants.STRING);
        } catch (XPathExpressionException e) {
            logger.error(e);
        }
    }

The evaluation returns an empty string. A sample XML is below:

    <?xml version="1.0" encoding="UTF-8"?>
    <Version>
       <version_id>51312</version_id>
       <description>Some Description</description>
    </Version>

The solution that I've employed is to serialize the DOM into a string, and then immediately parse that string to re-build the structure as a new DOM. This seems incredibly inefficient and backwards - what am I missing with this import and xPath? I had thought it might be because the node's parent and such are not being set properly, but when I did the clone/adopt method (or adopting after importing) the same behavior is seen.

Any answers/explanations/suggestions/thoughts appreciated.