In Part 1  of Going Global, I explained why you might want to collaborate and how to identify potential collaborators. Now imagine that you've decided to tell one or more potential collaborators that your piece of science might be of interest to their research. These people you've identified could well be working on the other side of the world. Unless you have already met them at a conference (in which case they may not remember you anyway), they have almost certainly never heard of you, or, quite possibly, your supervisor's group. So, whoever they are, you will need to start by introducing yourself.
You could try cold-calling on the telephone, but chances are the people you are trying to contact are either at a conference, are in meetings with their own people, or are just too busy to want to speak with you. The good thing about using e-mail for first contact is that your are not pressuring the other people by demanding their immediate attention. The bad thing is that the other people can easily choose just to ignore you. And that's the first hurdle in getting collaborations off the ground. Busy, successful scientists can receive upwards of 50 e-mails a day, so to ensure you message is even read, you need to get yourself noticed with an eye-catching subject line. Write it in fairly short prose, but make sure you include their keywords; make it sound like something they will want to read. They probably don't want to hear about your results--they want to know what's in it for them. So, for example, you could use 'Opportunity to collaborate with group interested in the function of XYZ'. Naturally, ensure that XYZ includes one or more of their keywords.
Once your message is opened, you face the next hurdle: convincing them it's worth their while bothering to write a reply. So state your name and whom you work for, then summarise the general area of your research in no more than two lines! If you talk about yourself too much they may well 'switch off' and start scanning down the page. Some more obvious issues: Get to the point; keep it brief and state exactly what you would like from them (i.e., say " please"), and not what you want. Give them numbered points to avoid misunderstandings.
But what happens if they don't reply after, say, a week? Well don't give up. I once sent three e-mails and made four phone calls trying to track down a possible collaborator. After almost giving up hope, I got a reply by e-mail that was full of apology for the delay and gave me exactly the news I had been hoping for. Some scientists really are unbelievably busy, even if they spend too much of their time sitting in airport lounges. The above example was an extreme case, but I always remind myself to follow up with at least a second--and if it's really important, a third--e-mail spread over the course of a month. This allows for absences due to conferences and other visits. The second time around, include your original message in case the first one got deleted. Be polite and simply restate your interest, rather than rudely asking if they have had the time to read your message. If all else fails, ring them or ring their lab and speak to someone who can help you get a message to them.
If you get lucky and receive a favourable reply (most people eventually respond well to a brief, intelligent request), talk to your boss before you respond. There are lots of people out there who'd like privileged access to your unpublished results, and your boss, not you, is the one person who should decide what detail, if any, you share with them. Thank goodness I have no personal experience of having my results 'scooped', but several naïve PhD students must have fallen foul of unscrupulous scientists saying 'thank you very much and goodbye' once they had got the information they wanted.
The politics of collaboration can be a daunting prospect for PhD students who haven't even met all the key players in their own immediate fields. Your advisor must play a key role if you are to negotiate (literally, negotiate) your way through the subtleties of scientific protocol, especially because this process will probably be carried out mostly via e-mail. Without access to the human voice and body language, this can all get a bit cagey in my experience, and it can certainly open you eyes to how much science--like any other profession--is all about personalities. If the other person is a bit of an unknown quantity, your boss may want to start by offering a small interesting titbit of information from your lab-book to see what is given in return. Or you could write a general statement about your work that catches the other party's interest but doesn't really tell any useful detail. The person will soon see where you are coming from, and so can begin a tennis match of e-mails that can rapidly become a daily source first of excitement, then furious discussion and another quickly but carefully drafted response. All this is very satisfying and, for me, what science is all about: people with nothing else in common getting very excited about their own little corners of the natural world.
Last comes the real business of any collaboration. This is when tangible things like scientific materials and PhD students start moving from one lab to another to open up experiments that would otherwise be impossible. There's little to match the excitement of receiving a parcel from Japan or the United States with potentially PhD-saving contents inside. And you never know, your name could end up on a joint paper!