Scientific Writing

Scientific Writing

Whitesides take: Youtube; Advanced Materials.

STAGE I: The skeleton – a framework to guide the execution of an experimental study where it is recognized a priori that said study and the results which could reasonably be expected of that study are of sufficient significance to warrant dissemination to a broader audience. The skeleton should consist of the following components:

  • Title: A tentative title. It should be short, but descriptive of the key point or focus of the work.
    • E.g. “Ordering kinetics of diblock copolymers during morphological transitions induced by a magnetic field”
  • Hypothesis: A description of the hypothesis guiding the work. Ideally all work is driven by a well-formed hypothesis. It may be very precise, or quite broad.
    • E.g. (broad): The kinetics of field-induced order-order transformations depends on the field strength.
    • E.g. (broad): The kinetics of field-induced order-order transformations depends on the MW of the polymer.
    • E.g. (less broad): The characteristic time describing the kinetics of field-induced order-order transformations should scale with field strength squared. This is based on theory describing the balance of magnetostatic energy and enthalpy of phase transitions.
  • Justification: a description of why the work could be significant, and to what audience it would be significant. This will help guide the choice of journal to which the work may eventually be submitted and the writing of the introduction.
    • E.g. the field-strength dependence of the order-order transition is important for guiding the design of field-anneling treatments for processing block copolymers using magnetic fields. Since this is a study of polymer physics and not necessarily some functional property/device, the audience would be folks interested in macromolecular science, or soft matter physics.
  • Schematic: A description of a schematic which would convey the relevant essential physical or chemical information about the study.
    • E.g. Schematic (Figure 1) should show transformation from spherical microdomains in a disordered matrix to  hexagonally packed cylindrical microdomains in a smectic matrix. Schematic should be 2D and cartoon style, with illustration of chain-level continuity in the transformation from spheres to cylinders. Spheres merge along the <111> axis to form cylinders, with crystallographic axes and field direction indicated.
  • Data/Experiments: Descriptions of data sets.
    • Figure 2: Time-resolved, isothermal 1D (I(q)) SAXS data spanning tranformation, with appropriately selected timepoints
    • Figure 3: Panel showing 2D SAXS diffractograms at start, mid-point and end of transformation
    • Figure 4: Panel showing TEM of system at start and end, and if possible, TEM of sample in mid-point of transformation by quenching from the high temperature state.
    • Figure 5: Analysis of kinetics – integrated intensity of (001)_bcc and (001)_hcp as a function of time. Is there a particular growth law describing the kinetics? Are there models out there to describe the kinetics?
    • Figure S1: Supporting info: DSC data from sample along with GPC analysis confirming MW

STAGE II: The draft outline – an outline to guide the preparation of a draft manuscript. The outline incorporates properly formatted figures and descriptions of the data. The outline should have a tentative title, a paragraph encapsulating the findings and significance, a schematic illustrating the relevant physics, chemistry or mechanisms at play, and properly formatted raw and analyzed data.

  • Title: A tentative title. It should be short, but descriptive of the key point or focus of the work.
  • Findings and Significance: A short paragraph of no more than 200 words describing in clear, logically connected sentences the findings of the work and its significance/context. This is the precursor to the abstract and helps set the vector or direction taken for the introduction.
  • Figures – lay out all figures. Write a short paragraph for each figure describing the data or illustrations in the figure. Each figure should have a point associated with it, i.e. something that it, the figure, is to communicate specifically. The point of the figure must be justified by the overall hypothesis or point of the paper. As such, the description must establish a connection to the main finding &/or to another figure or data set. For example, Figure 1 may be a schematic showing a transformation from BCC spheres to HCP cylinders with an applied field indicated. As such it is connected to the central thesis of the paper, i.e. order-order transition under field. Figure 2 could be SAXS data showing that the system does indeed exist as BCC spheres at t=0, and then HCP cylinders at some later time after the transformation. So the main point of Figure 2 is to show that a field-induced transition occurs. Figure 3 may be extracted data of the intensities, so it is connected to Figure 2, and its main point is to provide the raw data of the kinetics of the BCC-HCP transformation.

STAGE III: The Introduction.

  • The first draft of the introduction and abstract are added to the draft outline.
  • The figure descriptions and figures themselves are refined.
  • Abstract: Provide a short abstract of no more than 200 words, and ideally, ~ 150 words. The primary purpose of the abstract is to communicate succinctly what was done or accomplished, and/or the findings of the study. Often, the findings can be put into context by using 1 or 2 carefully constructed sentences that convey some information about prior work and the significance of the results. For example, a sentence at the start may provide some contextual information, and a sentence at the end may continue this by helping to establish significance. Be careful however that the abstract does not start to become a “background and introduction” section. It is meant to provide a very brief summary of what was done and the findings of the study
  • Introduction: The precise length and nature of the introduction depends on the nature of the article. Nevertheless, all introductions must set the stage for the data to be presented. A good introduction provides the reader with the necessary context in which to appreciate the need for the current work and its potential impact. It is not possible to write a good introduction without being cognizant of the relevant literature. Even if the introduction does not provide a short review of associated work, the ‘vector to the story’ can only be effective is the author is aware of the current state of the field and why the work is justified/needed. In general, the composition of the paper should be logical, such that content of paragraphs and even sentences follow one from the other. This requires effort, but this effort is particularly needed and justified in the introduction. In many cases, an effective introduction includes some preview of the conclusions. The old maxim of “Tell them what you will tell them, tell them, then tell them what you told them” holds true. References are required where you make an appeal to an argument/condition that is especially important for an argument/interpretation/justification in the paper, and thus where you may benefit from substantiation by external sources. References are also required to illustrate specific cases that are mentioned and which the reader may reasonably be interested in following up on.

STAGE IV: The draft – Complete the draft by writing the experimental, results/discussion and conclusions, along with any supporting information. Remember that the writing should be logically structured. A common pitfall is to describe experiments and results chronologically, but this if not a priori the most effective way to communicate the underlying science. Unless there is a specific benefit that derives from communicating data chronologically, data should be communicated in the sequence that is most effective in delivering the narrative, given the context of the work, i.e. the hypothesis and justifications for the work.

STAGE V: Versioning – The draft is improved though successive rounds of review and revision. The final version is optimized in terms of layout and content for submission to the chosen journal. You must read and re-read your paper several times, with a very critical eye. Be on the lookout especially for logical fallacies and factual errors – any sentences which don’t sound right when you read them represent red flags – if they don’t sound right to you, it is likely that they will not sound right to a reviewer, and more importantly, to readers of your work over the many years (eternity!) to come.

Letter to the editor

To be completed.

Suggested referees

To be completed.

Citations: Avoid citing a group of papers at the end of a sentence that in principle is referring to several distinct topics. For example, instead of “The physical chemistry of metal oxide nanoparticles finds relevance in topics such as quantum confinement, sensing and lasing.[1-8]” you should write “The physical chemistry of metal oxide nanoparticles finds relevance in topics such as quantum confinement[1-3], sensing[3,4] and lasing.[5-8]”

Logic and structure: Effective scientific communication relies largely on the logical, clear, presentation of concepts and data. Presenting the concepts and data in sequences that lay bare the connections between/among individual elements is a key goal. Presenting grammatically correct text that is nonetheless ambiguous in meaning is a common mistake that detracts from the work. Some examples appear below. Can you spot the issues?


  • Pimco’s parent, the German insurer Allianz, is likely to be insulated from any challenges Pimco faces, given its wide array of business lines. But Pimco has long been run as a nearly independent company from its sunny office complex in Newport Beach in Southern California.
  • The police said they ordered him to put down the weapon and instead he refused to comply and raised the weapon toward the officers, at which point one of them, Jorge Romero, a seven-year veteran of the Ridgefield police, began firing.
  • The electron and hole pairs were generated within silicon and ZnO due to the transparency of ZnO shell, thereby enabling penetration of light into the silicon layer.
  • But the situation came to a head when Sharif fired Musharraf, a junior general whom Sharif himself had promoted through the ranks, while he was overseas.
  • Shear thinning is a common response displayed by suspensions of non-spherical colloids such as fumed silica, because the particles have a preferential orientation under flow.
  • Englert, a Belgian theoretical physicist, and Higgs, a British physicist, earned the prestigious prize for “the theoretical discovery of a mechanism that contributes to our understanding of the origin of mass subatomic particles” which was confirmed through the discovery of the predicted fundamental particle by the ATLAS and CMS experiments at CERN’s Large Hadron Collider, the Nobel Prize Twitter feed reported.
  • This approach is known as “grafting-to” and has been used to functionalize various metal oxide nanorods such as TiO2, ZnO, SnO2 and CdTe.
  • Nevertheless, the power in molecular recognition as a means of hierarchical assembly is justifiable in its practical simplicity and, with the proper mechanistic understanding, applicability to a wide variety of inorganic/organic systems through simple chemical functionalization.
  • Chinese medicine almost never uses individual plants or minerals. Instead, it relies on diagnoses based solely on the doctor’s questions, observations and the skillful taking of the pulse.
  • The mantra, these days, is that there are only two types of companies left in this country: those that have been hacked by China, and those that do not know they have been hacked by China.
  • Millions are without power and cold. Some are dying.
  • Obesity, walking pace and risk of severe COVID-19 and mortality: analysis of UK Biobank.

Tense: It was once the case that all scientific writing employed the past tense, and passive voice. For example, one might write:

Somewhere in the introduction: “Here, the morphology of samples was examined as a function of annealing conditions.”

Somewhere in the experimental section: “Samples were subjected to various annealing conditions ranHere, the morphology of samples was examined as a function of annealing conditions.”

Somewhere in the data/discussion section: “The morphology was found to be strongly dependent on the annealing conditions, as shown in Figure X.”

Nowadays that has largely given way to the use of active voice, and the use of first person narration (use of “we” in writing). There is no longer a strict adherence to past tense which can often make for a confusing mixture of past and present tense in the narrative. A good guideline to overcome this is to use past tense when describing distinct actions that were taken, or things that were done in performing an experiment or study. By contrast, present tense is more appropriate when describing data that is being presented in the paper.

Somewhere in the introduction: Here, we examine the dependence of morphology as a function of annealing conditions.”

Somewhere in the experimental section: “Samples were subjected to various annealing conditions ranHere, the morphology of samples was examined as a function of annealing conditions.”

Somewhere in the data/discussion section: “We observe that the morphology is strong dependent on the annealing conditions, as shown in Figure X” or “As shown in Figure X, the sample morphology depends strongly on the annealing conditions used during sample preparation.”

“Here, we examine the dependence of morphology as a function of annealing

Convoluted sentences: Avoid long convoluted sentences, especially ones which use the passive voice. Try not to place qualifiers before the main point of a sentence, particularly when/if the qualifiers are long!

For example:
“In the event that the sample was not first cooled before being subjected to UV irradiation, the morphology was found to be independent of the molecular weight.”

Here, the main point is that the morphology was independent of the molecular weight, so this should be stated first, then the qualifier given.

“The morphology is independent of molecular weight for cases where the sample was not first cooled before being subjected to UV irradiation”

If the qualifier is very long, it should be broken up into separate sentences:

“As shown in Figure X, the morphology is independence of the molecular weight under certain conditions. This is the case for samples which were not first cooled before being subjected to UV irradiation”


Resolution matters! There is nothing you can do to generate higher resolution in your data. Vector graphics/drawings can always be re-exported from the original source at a different resolution, but for actual data, micrographs etc, try to capture as many pixels as you can when you first take your data, and then resize later. A good rule of thumb is to presume that 300dpi is the resolution used for printing articles. A typical figure in single column display is no more than 3.5″ wide. That means you *need* at least 300×3.5=1050 pixels of data to be able to resize your image without throwing away information. If your original micrograph/image is only 525 pixels wide, then to print it at 300dpi and fill 3.5″ of space required *inventing* an additional 525 pixels of data. Image processing applications are happy to do this for you by interpolation etc., but the results are usually not satisfactory. It’s better to start with more pixels and resize downwards. Now how does all this translate for digital production, i.e. many manuscripts are viewed entirely digitally, on a monitor or tablet or whatever. The situation here is a bit more nuanced. In theory the size of the image on the screen is determined only by the number of pixels constituting the image and the resolution of the screen. If you display your 1050 pixel image ‘natively’ on a 1024×768 monitor it will take up all of the screen/the screen won’t be able to display all of the image (26 pixels will be off screen). However when you look at your 1050 pixel image at 300dpi in MS Word or in a PDF document, it clearly does not occupy the entire screen, so what is going on? The applications perform some resizing/translation to account for the difference between the desired display size, the actual number of pixels in the image and the native resolution of the device. This also happens with physical printers of course, and it’s the reason that your data looks better on higher resolution screens, i.e. those that have more physical display pixels for a given physical width, can actually *show* data that’s there when the image is reproduced at a given size.

Designing for web display is another issue entirely and fortunately for scientific publishing we don’t have to consider it too much. Briefly however, web display is usually standardized on 72 or 96 dpi. So if you post your 1050 pixel image directly into a webpage without any specification of the size at which it should be displayed, it will appear at either 14.6 or 10.9″ in width. There’s some good reading here on this topic.

Regarding image resizing for manuscript preparation – please avoid resizing in word processing or presentation applications, i.e. MS Word or Powerpoint. You are *far* better off laying out your images and figures in desktop publishing or image processing applications such as Adobe Illustrator, Fireworks, Corel Draw etc. These are excellent applications that provide image processing, vector drawing and layout. Avoid drawing figures in Word and Powerpoint – the results are often less than satisfactory.

Reviewing Papers

The process of reading, distilling and critiquing papers will go a long way in helping one improve as an author in one’s own scientific writing. A good referee report describes and considers the novelty of the research, the technical quality of the experiments, data and analysis and whether the conclusions are adequately supported by the data presented. If possible, the report should help the authors to produce a better paper, so if a revision is recommended, the report should communicate specific ways in which the manuscript can be improved. Above all, the goal of the reviewer is to enforce the criteria of the journal. A referee report most typically contains two sections. The first is a short paragraph (a few lines) that summarizes the paper to the editor, details the referee’s overall thoughts about the manuscript, makes a recommendation to publish/reject/submit to another journal/publish after minor revisions etc. The recommendation to publish/reject/revise should be justified with a discussion in the following section of the report. The discussion should describe, point by point, the items that the authors of the paper need to address in any revision, or places where the paper has fallen short, or where it stands out. At the end, the review should be something that if one received would make one feel that the referee had read, understood and carefully considered the work of the paper, regardless of the final recommendation.

ACS provides a good resource with their peer review training in ACS Reviewer Lab.