How to be a resourceful problem solver:
Tips and insights for achieving great research

These are key ways to doing research efficiently:

1.  Educate yourself

No one starts a scientific career by knowing everything about a research project.  The true excitement of science is when one discovers how little is known, and how one can make a difference by contributing to “human knowledge”.  A “bad” scientist is one who doesn’t get excited through new knowledge.  It is a common characteristic of great scientists to have their eyes and mind excitedly open to their environment, picking up on differences, similarities, and subtleties, while staying critical of uncertain facts.  Each student, postdoc, or faculty needs to continually expand their knowledge by reading or by attending seminars (in a broad sense, not just what touches a project), and by discussing their own or their peers’ work, or published papers, with lab mates and the advisor.

2.  Work ethics

A key aspect of being a great researcher is to possess a high work ethic, for example by constantly trying to convince oneself (and not the boss) that every possibility for carrying out a reaction or research project has been followed.  This is a quality easily seen in people who always make sure that every detail is correct, checked and rechecked (being “anal” in other words…) Being a researcher, especially at the stage of graduate or postgraduate school, is more about living an experience than just doing a job. After all, why would one want to spend so many years of a youthful life to achieve little in the end?

A great researcher spends as many hours as needed to finish an experiment with precision, care, and attention to details.  Carelessness is a sign of poor work ethics, and accordingly, it is unsafe and unpredictable.  This attitude tends to rapidly bring down the morale of other group members, which should not be tolerated by peers or advisor.

3.   Budget your time

Learn to divide tasks by planning ahead.  A reaction can be setup when you come in the morning, before doing any other work.  As a result, you will be able to workup most reactions after you have taught or kept up with literature, as well as purify and characterize synthesized compounds by the mid to late afternoon.

Your Research Project

1.  Background literature (get informed on what you probably don’t know)

If one tells you that swimming is easy, and that you just need to jump into the river since you will learn it “cold turkey”, you would be stupid to follow such advice.  The same goes with starting a research project.  You need to know as much as possible about your field to get an appreciation for the concepts lying at the heart of your project. This cannot be learned all at once, so you will need time and sometimes the years of your graduate studies to gain a full appreciation of all the work that has been done in the area you are interested in.  Thus, you should first learn the basics, and later enhance your knowledge by reading literature as widely and regularly as possible.

Thanks to the convenience of online journals, one of the best ways to appreciate literature is by browsing entire table of contents of journal issues to get an appreciation of other work in the scientific community. Just by locating articles relevant to your work with a few keywords is often useful, but does not enhance the breadth of your knowledge to the same extent.  This may be quite overwhelming at first, but you will learn so much this way and in much less time than you would think. Also, once you have located an abstract that is related to your compound or subject, go to citing articles to find who else is working in this area. You can also do a more general search of cited references with ISI Web of Science.  After all, others have already done the literature search before you, so there is no reason not to use it to your advantage.  The authors of published paper will (or should) list the most relevant papers to their work in the introduction and discussion sections.

2.  Planning experiments

It is of course critical that you understand what you will be doing before starting any type of experiment.  You should find reaction procedures that are similar to yours (use “model compounds” as often as possible). Either try the known reaction conditions first, or if you are more experienced, apply judiciously thought out modified reaction conditions to your molecules.  Several great sources for methods are available in the Rubin labs (Larock, Vogel, March, and Organikum).

Importantly, a common mistake that most beginners do is to follow a reaction “recipe” to the letter.  If one uses 100 mL of a solvent M for this many Y moles of compound, it doesn’t mean that you have to do the same. Also, if a particular solvent is used, it can sometimes be for trivial reasons (for example, everyone else has done it this way before, so other groups just follow the same trend).  Don’t forget about elementary principles of chemistry in choosing the polarity of your solvents (e.g. SN2 or other ionic reactions work best in polar aprotic solvents like DMF, DMSO, even carcinogenic HMPA).  Also, fancy reagents can look “sexy” (e.g. RedAl), but a number of older reagents (H2, LiAlH4) have been time-tested.  They can work better than the fancy ones…

You want also to try first-time reactions on a small scale.  No need to waste 25 grams, or even ½ gram of compound, whether expensive or not, to find out that you are unable to get the desired reaction to go.  Every researcher (including the boss) has had to “learn” a reaction before it works reliably.  New chemistry may seem obvious on paper, but carrying it out is often a matter of learning a number of parameters.  Often, this seems to be achieved by getting a “feel” for the reaction, even it this sounds rather unscientific.  There is actually often more of an artisanship aspect to organic synthesis than a scientific one.  I have witnessed it a number of times, with certain “gifted hands” getting just any reaction to work, while “synthetically challenged” hands messed things up miserably in the most elementary transformations.

One last word:  it is rare that a reaction “does not work” A reaction can go differently than you wanted it to, but you have to be able to address the problems and make changes that will lead your project the way you want it to.  If a reaction really doesn’t proceed, find a way around by changing reagents, systems, or even overall target.  If there is something that you will learn during your thesis, it is that you have to be a resourceful problem solver.  This quality can be used anywhere in life, business, science, or family (this may save your marriage one day…)

3.  Keep a record of your results

It is of utmost importance to keep records on all the data that you are generating.  There is obviously a limit to what needs to be written down, but most of the time too little information and data is recorded in lab notebooks.  However, “standardized” procedures or workups can be written as such.  However, all your hard work will be worthless if no one can repeat it.  Should someone ask for experimental details (photocopies of a notebook), it is extremely embarrassing not to have something documenting your work. You can also save yourself an immense amount of time writing down your experimental procedures in a publication or thesis style right away, rather than using cryptic and unclear sentences containing very few details.  Two or three years down the line, you will not remember the details (even critical ones) to repeat or describe your work, and you may have to repeat it to get it in publication shape!

As a rule, the following points should be applied systematically:

–    Record all experiments you do, including failed ones, with dates, formulae shown, amounts of reagents, compounds, and solvents used, and a clear narrative statement of the procedures and purifications (someone else must be able to follow them, not just you).  It is not enough to write down an equation to indicate that you have done the reaction.  To another observer, it simply says what it is:  a reaction written down, but never carried out.  You might as well have saved the paper.

–    Write down important observations, including failures, and add a brief analysis of why this may have occurred (e.g. “The reaction of C60 with benzene under irradiation gave only recovered C60 in 97% yield.  Energy transfer to the solvent probably quenches the triplet excited state of C60”).  Of course, this is a rather silly example but it makes the point…

–    Keep enough sample material (20-200 mg) for all new compounds!  This concerns mostly intermediates and compounds that are stable, but many C60 compounds should also be preserved for subsequent uses such as spectroscopy, TLC reference, and so on.  Make sure that you label the vial with a chemical drawing (a notebook number is unacceptable!)  You will discover that you often have to re-synthesize a compound close to the end of your thesis, just to obtain complete characterization.  You’ll hate the boss for requiring this, but this is simply what needs to be done for publication. Then you will blame yourself for not doing it properly the first time around!  Thus, it is really a bad idea to leave certain characterization data (e.g. IR, m.p.) for “later”, when you think you’ll have plenty of time to do it (this will never happen…)

4.  How to determine that you have the correct compound (proper characterization)

An experiment that is not demonstrated unequivocally is worthless.  All new compounds have to be characterized as part of your historical contribution to scientific knowledge (however small your contribution may be).  Cost in this case is not a major issue.  It just has to be done!

–      Record 1H and 13C NMR, IR, UV (where applicable) and MS and HIRES spectra for all NEW compounds.  Keep only one example of the good spectra in a separate file (use our folder system) as reference for the entire group.  As you may have already found out, these files are extremely useful and you should be thankful to previous group members for their care in doing this right.

–      Once you have done that, you will also need to analyze your data.  Compare all the characteristics of data within a particular experiment so that they are consistent with all the available data.  For example, a carbonyl peak in an IR spectrum should also be found in its 13C NMR spectrum (unless it is lost in the noise in rare cases).  The mass spectra must be recorded as both LOW RES and HI RES.  An example of a typical experimental section is provided at the end of this note (in accord with JOC Guidelines, see attached).

5.  Presenting results at a group or public meeting

Another important quality of a great scientist is to be able to express and communicate his knowledge to others.  There is no point doing great work if it gets lost in the crowd.  This applies to both oral presentations and written reports or publications.  There are many available resources in the lab for you to learn and use.  Power Point presentations should take full advantage of your presentation skills.  Working on a presentation entails several aspects:

–      You need to put your data together carefully (even for a literature report) and make them clear to yourself.  Write down important points on separate lines so that you have a clear breakdown of your presentation.  You can use a title as an entry point for each slide.  Once you have the basic ideas in mind, start drawing schemes or writing down ideas (keep them brief so not to lose your audience).

–      Once finished, go through the slides on your own. Always try to “project” yourself in the other person’s mind who will be listening to you. This way, you will make sure that your audience knows what you are talking about (this tends to be a common problem, even with “famous” scientists; or should I say infamous?)  You need to make sure that your audience will understand what you are talking about.  If they don’t, then you will have worked for nothing, and your message will be lost (or worse, you may have made it boring or repulsive…)  Try to keep your message simple by using mostly illustrations in your slides, while supporting them by orally expressing a few brief points.  Slides with too much writing never work.  The ideas that you want to convey have to be orally presented and supported by gestures, and the figures on your slides.

–      Avoid going into too many details.  People know that you are an expert in a field, but you have to be able to convey the “big picture”.  Forget how much you have sweated through a particular set of reactions, or how much work you have put in the presentation.  Rather, present the essence of your work.  In a few years, the details will not matter, but the new important results that you will have achieved will stay forever in history.

 

6.  Publishing (this will be one of the most significant contributions to your career and science)

Reporting your results in a published form is one of the most important goals of your work.  Although publishing is not an end in itself, there are absolutely no results, how beautiful or original as they can be, that become part of the scientific community unless they are published.  No one will be able to use or credit your work if it goes unreported (presentations at scientific meeting do not really count as publications; they are a good way to disseminate preliminary results, and practice oral presentation skills).

Your chances to get a postdoctoral or job position will be decided on the quality of your work and publications.  It is therefore important not to get caught up in your project and forget that a publication has to be made regularly to publicize your work.  As you will soon find out (or already have), publishing takes a lot more work and finesse than you ever imagined. Even details such as writing down spectral data, writing a cover letter to the journal editor, final editing of all aspects of your manuscript, all these take a huge amount of time.  It is therefore imperative for each co-author of a publication to participate in the writing, editing, and finalization of the manuscript.  This starts by having all experimental data cleaned up and written down in ACS format well in advance of manuscript writing.

The publication guidelines for ACS journals should be followed as the best model for most of your manuscripts (see attached).  In particular, follow the text style and use templates available from those journals to do things right.  Following these formats is simply a matter of convention, but stylistic carelessness reflects a lack of professionalism, and it also exasperates referees who have to decide on acceptance of your manuscript. Thus, poor writing, typographical errors, misnumbered compounds, poorly referenced work, etc, are all to be avoided.  Make also sure that you give proper credits to people in your field who have done ground-breaking work “for you”, meaning that they have opened up opportunities for your project by establishing precedents.  No project starts in vacuo, and the people who have contributed to your “inspiration” should be recognized.

7.  “Getting out of here…”

Graduate Students

Once you start writing your thesis, you will probably fully appreciate all the above remarks.  You will have accumulated 4 to 5 years worth of data that have to be explained in a few chapters.  Don’t you wish you had done all the characterizations earlier?  What about having written down more details in your notebooks?  You could have spent more time working on your research reports all these years.  It would have helped so much!

My point is that a thesis is a culmination of your learning, and an “acceptance” by your peers and elders, that you are now a Doctor and fully deserve it!  Your research results will land you a great postdoctoral or industrial position.  When you apply for jobs, your accomplishments will speak for themselves.

Your thesis has to be thorough and start with an introductory chapter exposing the background of your project and the reasons for your work.  This is where you put forth the broad ideas underlying you thesis.  The following chapters span different projects you have worked on, and each one has to be consistent, including a short introduction, results and discussion, conclusion, experimental section, and references.  Your thesis should be checked carefully by other group members before it reaches me and your dissertation committee members.  I will do the final reviewing.  All three signing members of the Ph.D. committee need to read the thesis and approve it.  Plan to spend about 2-4 months to write the thesis so that you can also make arrangements for jobs or postdoctoral fellowships.  You will find that it always takes much longer to write the thesis than you anticipated.  Please keep in mind that, regretfully, I will not be able to sign a thesis that is not properly finished.

As to the matter of finishing up practical work in the lab, you should follow the guidelines provided below for the postdocs.

Postdocs

It is required that you turn in a thorough research report before you leave, no matter how constrained with time you are before leaving for your new position.  Once you have left the group, there is a little chance that you will turn in any report.  It is also important, out of fairness to subsequent group members, that you clean up your hood, drawers, and desk space.  Additionally, all chemical samples to be stored (those lying in desk drawers, fridges, underneath hood, etc.) must be transferred to vials and marked with a securely attached label (use scotch tape to secure the paper label; use chemical structures only, no one will be able to tell anything from a notebook number!)  You can discard most other chemical samples, including TLC samples.  All compounds should be kept neatly in a cardboard box and stored in our group chemical storage drawer. Do not keep relatively unstable or light-sensitive substances.

Primary (clean) spectra must be filed into individual folders for each new compound.  A clear chemical structure should be marked on the folder tag, and the different types of spectra listed on the front of the folder. Only the best spectra should be kept for each of these compounds.  Routine spectra can be thrown away, unless they are useful to have for comparison.

Make sure that you leave an address with the department, since often checks or mail come in after someone has departed.

PROPER  WRITING  OF  DATA  FOR  PUBLICATIONS

In other words, all commas, italics, subscripts, and other stylistic forms need be followed and carefully checked and rechecked…

 

Experimental Section Example

General: All reactions were performed under argon and, for ¹O₂ sensitive compounds, in absence of light.  All NMR data were acquired on Bruker ARX-500 or ARX-400 spectrometers at 298 K (25 ˚C) unless specified otherwise.  Me4Si was used as the internal reference for 1H NMR, and the deuterated solvent was used as a reference for 13C NMR.  Proton-proton coupling constants were obtained from resolution-enhanced spectra after Gaussian multiplication of the FID (LB = -1.0, GB = 0.3 – 0.5).  Two-dimensional T-ROESY[31] data were collected in phase-sensitive mode (States-TPPI)[52 ] with 512 complex points in the directly detected dimension and ≤256 complex points in the indirect dimension, zero-filled to 256 complex points.  A spectral window of 4500 Hz was used for both dimensions in the T-ROESY experiments with a spin lock time of 500 milliseconds, a relaxation delay time of 2 sec, and a number of scans equal to 16.  All T-ROESY spectra were analyzed unsymmetrized, and were symmetrized for the Figures.  The natural abundance 13C-HMQC spectrum for 18b was acquired with a BIRD pulse[53] with 1024 complex points in the directly detected dimension and 128 complex points in the indirect dimension.  A spectral window of 4500 Hz was used for the 1H channel and 26000 Hz for the 13C channel with a relaxation delay time of 0.22 sec, and a number of scans equal to 48.

The matrix used for FAB mass spectra was m-nitrobenzyl alcohol.  Column chromatography was performed on silica gel 70-230 mesh or 230-400 mesh (flash) from E. Merck or Scientific Absorbents; thin layer chromatography (TLC) was performed on glass plates coated with silica gel 60 F254 from E. Merck.  All yields for reactions with C60 as the starting material are indicated based both on isolated product and on consumed C60 since the latter was easily recovered in the chromatographic purifications.  For all other reactions, yields are based on isolated product.

Materials: All N,O-ketene N-1,3-butadienyl-N-alkyl-O-silyl acetals (1a-e) used in this work have been described elsewhere.[15]  They were either received from the Neuchâtel group or prepared according to the published work.[15]  The C60/C70 soluble extract and, subsequently, pure C60 (99.8%) were obtained from MER Corporation, Tucson, Arizona 85706 or from Southern Chemical Group, Stone Mountain, Georgia 30087.  Details on separations and solvent purification can be found elsewhere.[18a]

Synthesis:  1,2,3,4,1’,2’,3’,4’,4a’,5’,6’,8a’b-Dodecahydro-N-isopropyl-3’a-methyl-2’-oxoquino-lino[4’,4a’,5’:1,2,3]buckminsterfullerene (2a), Method A: To a refluxing solution of C60 (108 mg, 0.15 mmol) in 50 mL of toluene was added N,O-ketene acetal 1a (52.7 mg, 0.188 mmol) in 15 mL of toluene through a syringe pump over a period of 3h.  After the addition, the reaction mixture was maintained at reflux for another 3h, cooled to ambient temperature and purified by flash column chromatography on silica gel (toluene, then toluene/EtOAc, 98:2) to give 65.8 mg (49%; 61% with 20.0 mg of recovered C60) of the product as a dark brown solid:  1H NMR (500 MHz, Cl2CD-CDCl2) d (ppm) 0.90 (d, J = 6.8 Hz, 3H), 1.32 (d, J = 6.8 Hz, 3H), 1.99 (d, J = 6.6 Hz, 3H), 3.95 (dq, J = 15.0, 3.0 Hz, 1H), 4.01 (dd, J = 15.0, 6.8 Hz, 1H), 4.18 (q, J = 6.6 Hz, 1H), 4.97 (sept, J = 6.8 Hz, 1H), 5.02 (q, J = 3.0 Hz, 1H), 5.84 (s, 1H), 6.75 (dt, J = 9.1, 3.0 Hz, 1H), 6.93 (ddt, J = 9.1, 6.8, 3.0 Hz, 1H); 13C NMR (125.7 MHz, Cl2CD-CDCl2) d (ppm) 12.23, 20.05, 20.97, 41.28, 44.37, 49.46, 58.13, 58.25, 61.67, 63.33, 67.55, 129.54, 132.76, 135.24, 136.58, 137.26, 137.65, 137.74, 140.76, 141.09, 141.16, 141.63, 142.00, 142.41, 142.45, 142.53, 142.57, 142.65, 142.71, 142.86, 142.97, 143.39, 143.67, 143.89, 143.99, 144.04, 144.15, 144.23, 144.36, 144.42, 144.52, 144.70, 144.73, 144.79, 145.17, 145.19, 145.20, 145.37, 145.41, 145.45, 145.82, 145.89, 146.39, 146.45, 146.61, 147.01, 147.07, 147.25, 147.38, 148.04, 148.33, 148.46, 148.66, 148.83, 149.37, 149.55, 149.73, 152.74, 154.60, 169.84; FT-IR (KBr) n (cm-1) 1661 (vs), 1455 (s), 1433 (s), 1376 (s), 1262 (m), 1209 (s), 730 (s), 524 (s); UV-Vis (nm) 256 (e 119 400), 328 sh (32 900), 404 sh (7 020), 431 (5 920), 648 (440), 676 (330), 712 (290); FAB-MS m/z (rel intensity) 888 (30, MH+) (include other recognizable fragments), 720 (100, C60+); HRMS (FAB+) calcd for C70H17NO.H+:  888.1388; found:  888.1400.

NOTE:

In ¹H NMR spectra, please calculate all coupling values within the same groups of peaks (for example the four values in the smallest branches of a ddd); then average those values to the highest decimal.  Then, write down the average of the two coupling constants found for both spin partners, e.g. H_a and H_b, since the reader has to be able to recognize what spin systems are paired within your experimental description.  Giving differing values (e.g. 6.7 and 6.5 Hz), even though they are the experimentally determined numbers, is not very useful because other coupling constants may be present in the same range for your particular compound, and there is no way to distinguish which spin partners you have assigned without pairing these coupling values.

Please make also sure that you scan (or save graphic files) for all spectra. They need to be included into Supporting Material sections when submitting publications.