I've had best results using the following Xpath to get Google US's 1st Organic listing (ignoring Answer Boxes & other SERP features):
(//div[@id='rso']//div[@class='g' and descendant::div[@class='rc']][not(ancestor::div[@class='xpdopen'])])//cite
I just tried it on google.com.br and it worked fine there as well. Note that this will give you the display address (usually in green) as in your example. If you want the actual URL from the href, replace …
//cite with …
//h3/a/@href. I usually collect both elements so I can identify whenever Google chooses to override such things (especially with mobile search).
When scraping Google SERPs, make sure you set the following in your HTTP settings (at least a 1sec delay):
<IntervalBetweenRequests RandomFrom="1000" RandomTo="2500" IfSame="Host" />
If you're getting banned, it might be time to consider using proxies. Hope this helps!