Donating money to scientific research after my death

Funding people instead of projects is in fact a good idea if you are interested in a supporting fundamental research problems to be addressed. There are multiple ways to go about this, and it's a reasonably well established principle, so you might want to search around a bit how other funding organizations and (non-profit) foundations approach this.

1. Lump sum vs. funding from interest is really a choice between aiming at one or two breakthroughs within a period of 5 to 10 years vs. continuously supporting research in a specific field, on a smaller scale, over 50 to 100 years. This relates to the answer of the next questions...
2. Amount to give to a single professor: A strategy for short term funding (using up the capital) would be to fund the establishment of a new research group, i.e., pay the salary of someone early in their career (assistant or associate professor level) plus some funds for research staff (PhD students and post-docs). I would recommend a funding level of $2M (that is what the Consolidator Grant of the European Research Council provides for a period of 5 years) to $5M per professor, depending on how much expensive equipment and materials are needed to do this research. That means you could fund 2 to 5 such groups. After the period of funding at that level, these persons should have achieved sufficient research results to be self-sustainable, though you cannot be sure that they will continue working on the kind of questions you aimed to fund. For long term funding, I would recommend donating the whole sum as a Financial endowment (see Endowed professorships) to one university. If you assume an interest rate of 1 %, it will pay $100k per year, which might be just enough to pay a professor's salary, or, if the university agrees to pay the professor, could be used to fund 1 - 2 PhD students or post-docs working with the professor.
3. How to spend the money: Don't transfer that money to a professor's private account. Professors get research budgets from their universities, and if the funder wants to, the spending restrictions can be extremely flexible while still making sure that the money is only used on research related costs. So you either need to make an agreement with a university for them to administer the money according to the stipulations you agree on, or set up a foundation that administers the money and pays it to universities, who then put it in the research budget of the professors that it's meant to support. For the first option, you may want to contact a university's alumni or financial office, or maybe the dean of a department in the area you'd like to support.
4. Don't plan to select the persons you're funding yourself. First, it's better to fund young researchers, and if you talk to potential candidates while you still can do it, I'd hope that these will be old, or at least senior researchers, by the time the money becomes available. Instead, choose the persons that will do the selection. In the endowment case, that would be the university department that gets the endowment. Make sure that the general research direction and overall functioning of the department fits to your aims, and only select them if you're reasonably confident that they will use the money according to your aims. In the non-profit foundation case, I'd suggest setting up an advisory board with reputable scientists in the field you'd like to fund, and let them run the selection. Keep in mind that the composition of this group (institutions they come from, character, etc.) will be an important factor in who they select for funding. If this is going to operate over a longer time, make sure that there are regulations in place how new persons can enter the advisory board when you won't be there anymore.

You are in a fortunate position, and so are the ones who may benefit from your generosity.

I imagine that you will have no problem getting people to talk to you. As far as modalities are concerned, the easiest way to find out how to do this best is probably to talk to someone in the "development office" of university of your choice. Just transferring the money into a professor's bank account is not a winning strategy, because that person might have to pay taxes and you'd have no guarantee that the money is actually spent on research (as opposed to a nice vacation in the Caribbean). On the other hand, development offices are well equipped to deal with donors. Options include making your money a "no strings attached" grant to an individual professor, endowing a "chair", or endowing fellowships for graduate students or postdocs in a particular area of research -- in essence, universities will be very happy to take your money, but you can set the conditions on how it will be used. All of this includes helping you draft details of these arrangements that will then become part of your will and that will be executed when the time comes.

If you talk to someone in a research university's development office, you will find out that your situation is more common than you may think. Your contacts there will have seen every variation of arrangements and should be able to help you find ways to achieve what you want!

One suggestion to consider is endowing a Professorial chair. The cost of this I believe is something like $5 million.

The university would take care of the administration and it would be an ongoing contribution to science. It would also be clear where the money is going -- the Professor's salary. This is different to making a more nebulous donation.

You can choose the area the chair should be in and its name. Potentially you could have some limited input into selecting the inaugural holder of the chair.

You would need to approach the department/university to explore this option. But bear in mind that you need to convince them that you are serious about donating.