EDIT (9/29/2016): Videos of the concert.
“Statistics? Statistics? Statistics mean nothing, newbie. As doctors, we know that people diagnosed with pancreatic cancer have an 85% death rate within five years, whereas people having an appendectomy have a 95% survival rating, but we both know pancreatic cancer sufferers who are still alive and appendicitis patients who didn’t make it. Statistics mean nothing to the individual. Not a damn thing.” -Dr. Cox, Scrubs
I’ve neglected this for a while – I’ve been busy with other stuff. But this week’s issue has a LOT of good articles that caught my attention. It’s the story of my life; everything always gets lumped together, rather than being spaced out in regular intervals.
Now that I’m learning how to use the Pandas package in Python, all I can think of is this:
I mentioned before my reasons for not wanting to do a postdoc after completing my PhD. I will freely admit that when I started as a young, naive, doe-eyed first-year graduate student, I initially wanted to go into academics – I was even told multiple times by people from within and outside my department that I had the “mentality” and “intelligence” for academia. After having my soul properly crushed a few years into the program, my goals readjusted to something more realistic – that is, getting an industry job, a goal that was considered by many “selling out”, “settling”, or “selling yourself short”. I didn’t do a postdoc because I wanted to get an industry job, but now it appears that a postdoc is necessary – and this is information that I only found out after graduating.
Now, the problem is that there’s no clear-cut advice as to what PhD’s should do in order to get industry jobs these days. As far as academia is concerned, a postdoc or two is mandatory in order to broaden your knowledge base, make your CV more competitive, and get additional network contacts and letters of recommendation. However, if industry is the goal, then you will hear things from all over the spectrum, such as “doing a postdoc lowers your eligibility for industry positions since it means you’re too focused on academics”, to “we throw the resumes of applicants without postdocs in the trash”, to illustrate the two extremes.
SeeArrOh wrote about this situation a month ago, but it is still valid, and I think the situation is going to get worse with time, as the saturation of scientists at the PhD level keeps increasing year after year. My experience tracks with SeeArrOh’s observations. I think that my inability to get job after completing my PhD could be attributed in part to not doing a postdoc after graduating. That being said, doing a postdoc does not guarantee getting a job either! It’s still a very risky gamble.
One big problem is that these employment issues are very opaque to graduate students, and it is only recently, thanks to the efforts of truth-tellers like Chemjobber, that these issues are coming out into the open, and students/postdocs are able to read about employment and unemployment in the chemistry job market (largely anecdotal, but these are better than no information at all). It also works to the advantage of PI’s to keep their students in the dark regarding employment after graduation; PI’s can promise the (nonexistent) big payoff in order to keep their students working hard 80-90 hours a week, sacrificing their lives at the altar of science.
Unfortunately, the issue “do you need a postdoc if you want to get an industry job?” has not been resolved, and this is something that incoming students need to be aware of. If the answer is yes (a postdoc is necessary), then you need to be prepared for the long haul; an additional 7-10 years in school after undergrad (PhD + postdoc) in order to get a job. That’s why I tell people science is a lousy career path these days. People used to criticize medicine for taking too much training before being able to start one’s career, but I think science has safely beaten that now. According to the 2014 NSF Survey of Earned Doctorates, the mean time to PhD in the physical sciences is 6.5 years (5.7 years in chemistry), and is slowly increasing every year. The question of “how long is the average postdoc?” is more difficult to answer, but SeeArrOh did a back-of-the-envelope calculation for chemistry, and the mean postdoctoral stay (for those who went to academia) was 3.7 years. 5.7 + 3.7 = 9.4 years in school. Granted, these numbers were only derived from those going to academia, but they at least give some sense of the situation. Compare this with medicine, which is strictly 4 + 4 (8 years, 4 for medical school, 4 for residency – or 3+4 in some universities!). Suddenly, medicine seems like a smart choice, when one factors in the opportunity cost of time, the fact that residents on average get paid more than postdocs (for similar hours of work), and the fat doctor salaries at the end (the big payoff!) thanks to the AMA.
Finally, and this is something that will hit most people the hardest: Unfortunately, society sends PhD students mixed messages. On one hand, there are people who say “wow, doing a PhD is great, you’ll be able to change the world!”. But once you graduate, you see the real value of the degree, which is…less than toilet paper, due to insane market saturation in both academia and industry. Another issue is that it is very difficult to find employment statistics of graduates of PhD programs – this data is crucial to being able to assess the relative strength of a program, because after all, you get a degree in order to get a job and make money, right? But most universities do not care about what happens to their graduates after getting a PhD, which is very unfortunate.
This needs to change. If departments properly tracked career outcomes of their graduates, then maybe the equilibrium salary of PhD scientists would properly reflect the amount of training involved, rather than being depressed due to an artificial flooding of the market.
Wow, we’re on our fifth post in this series!
This next paper is a classic by E. J. Corey that all serious students of organic chemistry should have read at least once. It’s not terribly complicated, but at the same time is iconic enough that it is worth mentioning in this context.
A key concept when doing total synthesis (or almost any type of organic synthesis) is “functional group protection”. During the course of a synthesis, you may want to conduct a reaction to change one part of the molecule, but the reaction conditions employed for that transformation may affect other sensitive groups in other parts of the molecule. Thus, you will need to “protect” the vulnerable functional group(s) in the molecule so that they are not unnecessarily changed by the reaction.
When I was younger, I used to joke that a lot of total synthesis was basically “protection, coupling, deprotection”, making it like sex. If you do use this dirty chemistry joke later, remember to credit me (or not…do I really want to be associated with such hopeless humor?).
Anyway, a particularly sensitive functional group in organic chemistry is the hydroxyl (-OH) group. The hydroxyl group is what makes compounds alcohols. Ethanol, the alcohol we all know and love, is simply an ethyl group with an -OH.
Alcohols can also be thought of as organic derivatives of water. Water’s formula is H2O, and so alcohols can be thought of as water molecules with one of the hydrogen atoms replaced by an organic group (giving rise to the generic formula R-OH, where R represents any organic structure). Alcohols also undergo reactions on their own, and as such, are sensitive to acidic and basic reaction conditions. In basic media, the hydroxyl can be deprotonated, making it a stronger nucleophile (R-O– vs. R-OH), and in acidic media, E1 and E2 eliminations are possible, depending on the structure of the carbon backbone. Therefore, if you want to do a reaction on a molecule containing hydroxyl groups, you have to be aware of the reaction conditions, and if the alcohol proves to be reactive, it may be necessary to temporarily “protect” it or “mask” it. A common way of doing so is by converting it to an ether, such as a THP (tetrahydropyranyl) ether.
Another alternative is the use of a silyl ether (-OSiR3) as a protecting group. Silyl ethers are easily synthesized by adding a chlorosilane and a weak base (to mop up the HCl produced) to an alcohol. As with everything in chemistry, while the concept is simple, the situation is much more nuanced than might seem on the surface. Standard TMS (trimethylsilyl) ethers are not useful, because as mentioned in Corey’s paper, “Trimethylsilyl ethers are too susceptible to solvolysis in protic media (either in the presence of acid or base) to be broadly useful in synthesis“. One solution is to increase the steric bulk of the groups attached to the silicon atom. Replacing one of the methyl groups with an isopropyl, while promising, still did not give the stability desired, especially in “Grignard reagent formation, Wittig reaction, or Jones (CrO3) oxidation”. Thus, Corey decided to increase the steric bulk further, opting for a tert-butyl group on the silicon. This is called the “t-butyl dimethylsilyl” group, and is abbreviated -TBS or -TBDMS.
However, another roadblock ensued – TBSCl proved to be too unreactive with most alcohols under conventional or even forcing conditions, “for example, with excess silyl chloride, excess dry pyridine in tetrahydrofuran at temperatures from 20 to 60° for many hours”. This was solved by the discovery that the use of TBSCl with catalytic imidazole in DMF turned out to be appropriate conditions for the activation of TBSCl. These conditions quickly became known as Corey’s classic conditions for TBS protection. Even though a detailed experimental section is not included in this communication, Corey demonstrates the versatility of this method in the total synthesis of several prostaglandins. The TBS group is much more stable to hydrolysis than less bulky silyl ethers, and as such became widely used in synthesis. Like all silyl groups, it can be quantitatively removed by the use of fluoride (either a solution of KF or TBAF (tetra n-butyl ammonium fluoride)).
This begs the question – how does this method of activation work? My guess is that imidazole is able to initially bond to the silicon, forming a hypervalent pentavalent species, which is then transferred to the hydroxyl oxygen atom. Unfortunately, I cannot find the source right now, but imidazole is known to be able to activate silicon groups too, although not as strongly as fluoride does. DMF is also known to be able to activate silicon compounds such as TMSCN and TMSCF3.
More in the next. I need to get some coffee!
EDIT (7/28/2016): Prof. Andrei Yudin has a very nice discussion about this reaction on his blog.