There are indeed many examples of new species created by humans, and observed to arise by humans. Many via deliberate experimentation. You can find examples googling "examples of speciation", looking at the Wikipedia page for speciation or the Talk Origins one:
http://www.talkorigins.org/faqs/faq-speciation.html
http://www.talkorigins.org/faqs/speciation.html
https://en.wikipedia.org/wiki/Speciation
The issue is that there are two events of interest in "the origin of species" that are related but not the same thing: there is speciation, and then there is overall change. And as far as change itself goes you can have genetic change and phenotypic change which themselves aren't perfectly correlated.
Speciation is a very important evolutionary event, as it's the point at which genetic material is no longer exchanged between two groups and so they start evolving independently, meaning differences will accumulate over time (steadily and inevitably for genetic differences, more variably for phenotypic ones as that depends on the environment). But you can have differences accumulating without speciation (especially not with your "in vitro fertilization" criterion, which allows gene flow to be stopped for a long time without "speciation" by that definition happening), and you can have speciation happen without much genetic or phenotypic change.
And so on our human timescale we'll tend to see three categories of events:
1) large phenotypic change without speciation (especially by the in-vitro definition), because natural selection can cause large phenotypic change but random mutation cannot be rushed and that genetic drift is usually what will make groups no longer inter-fertile,
2) Speciation without large phenotypic change, again because speciation is just the beginning and all you need for "speciation" is just enough change to make interbreeding impossible or difficult, and so many examples of observed speciation involve just that, and
3) Speciation with large phenotypic change, which will usually be single-generation events with hybridization and polyploidy. Those are the most likely to fit your criteria that in-vitro fertilization shouldn't work. And they are legitimate and important evolutionary processes; polyploidy has happened many times in our lineage, it's an important source of genetic diversity, and hybridization and polyploidy are a major mechanism in the evolution of plants in particular. But they're not the major mechanism for the evolution of large vertebrates, so it's easy for Creationists to dismiss them.
Still, looking at the links I gave you I see a few examples that involve genetic change sufficient to block even in-vitro fertilization. The second Talk Origins link in particular refers to examples where speciation was determined by the offspring being sterile. There is also an article in the notes of the Wikipedia page about a recent discovery about the kind of genetic changes that can lead to speciation:
https://www.scientificamerican.com/article/wandering-fly-gene-suppor/
The general problem with giving Creationists examples though, is that the examples they want don't exist (and would usually contradict the theory of evolution anyway) and the ones that do exist they're not satisfied by (because the change involved isn't dramatic enough, or IS dramatic but as a result makes the dramatic look trivial because it happened via a small gradual underlying change - which, you know, is how evolution is thought to work).
I notice in fact that you are the one asking about the speciation with in-vitro fertilization, not your friend; your friend apparently just said that huskies and chihuahuas are not comparable to fish vs tetrapods. If this is personal curiosity from you then that's fine; if you are looking for arguments for your friend specifically then it looks like you're on the wrong track, since no amount of lack-of-interbreeding will make flies any less flies, which is usually the issue for people who focus on "fish to tetrapod" as a scale of change.
You might have better luck figuring out exactly what claims your friend makes, where they set the boundaries of what they think is possible and what they think it isn't, and then find an example that breaks that specific boundary. So they say "speciation can't happen !", you reply "speciation happened in fruit flies", and they answer "so what, fish haven't become tetrapods" - but, what is it they meant by "speciation can't happen" ? What mechanism do they think makes it impossible, what do they think this "speciation" is that's impossible, where do they set the boundary for themselves between "ordinary change that can happen" and "stuff that can't happen so evolution is false" ? They probably don't know clearly themselves, but figuring it out and addressing it is a better way of moving forward, and you both can learn things. And it makes it a bit easier for you too - you don't have to worry about finding proof that fish can become tetrapods, you can be like "OK fish can't become tetrapods whatever, but you said that new species cannot arise through genetic change, but here is a paper saying that a gene transposition makes it impossible for those two flies to interbreed, wouldn't that be an example of speciation via genetic change ? What do you think ?"
