Labs

Why do longer conjugated molecules absorb redder light? Tie the HOMO–LUMO gap to the UV-Vis spectrum.

orbitals (gap)UV-Vis (TD-DFT)

You’ll learn to: link the HOMO–LUMO gap to absorption wavelength via ΔE = hc/λ, and explain why extending conjugation red-shifts a molecule’s colour.

Background 1 — conjugation spreads the π system

When double bonds alternate with single bonds, their p orbitals overlap continuously and the π electrons delocalise over the whole chain. Each new conjugated double bond adds another π and another π^* level, so a set of evenly spaced bonding and antibonding orbitals builds up. Crucially, the orbitals spread apart in energy: the highest HOMO rises and the lowest LUMO falls. Ethylene has one isolated π/π^* pair; 1,3-butadiene has two of each; benzene has a fully cyclic, aromatic six-electron π system. More conjugation always means a smaller frontier gap.

Background 2 — particle-in-a-box intuition

A clean way to see why is the particle-in-a-box model: treat the delocalised π electrons as confined to a box of length L (the conjugated chain). The allowed energies are E_n = n²h²/8mL², so levels scale as 1/L². Lengthen the box and every gap, including the HOMO→LUMO gap, shrinks. This toy model captures the headline trend without any heavy computation: a longer conjugated chain → a smaller ΔE → a lower-energy, longer-wavelength absorption. The same scaling underlies the dyes and the carotene in carrots.

Background 3 — from gap to colour (TD-DFT)

The absorbed photon energy obeys ΔE = hc/λ, so λ is inversely proportional to the gap. A smaller gap pushes absorption to longer λ (a red-shift); once it reaches the visible (≈400–700 nm), the molecule shows the complementary colour. The bare orbital-energy difference is only a rough estimate — it ignores electron–hole attraction and orbital relaxation — so MoleBench uses TD-DFT for a proper excited-state energy. Treat both as qualitative: the ordering ethylene > butadiene > benzene-region gaps is robust, but absolute nm values from a single functional carry real error.

Molecules

Procedure

1. For each molecule, Compute orbitals (panel 4) and read the HOMO–LUMO gap in eV.
2. In panel 3, press UV-Vis (TD-DFT) and record the longest-wavelength (lowest-energy) absorption band.
3. Compare gap against wavelength across the series — as conjugation grows, which way does each move?

Questions

1. Rank ethylene, 1,3-butadiene and benzene by HOMO–LUMO gap, and explain in terms of how many π levels each builds and how far they spread.
2. Using ΔE = hc/λ, compute the wavelength absorbed by a molecule whose gap is 6.0 eV, and again for one whose gap is 2.5 eV. Which value lands in the visible, and what does that imply about colour? (Take hc ≈ 1240 eV·nm.)
3. The simple orbital-energy difference and the TD-DFT excitation energy disagree somewhat. Name two physical effects the bare gap leaves out that TD-DFT partly recovers.
4. In the particle-in-a-box picture, energy levels scale as 1/L². If you doubled the conjugated chain length, by roughly what factor would the HOMO–LUMO gap change, and which way would λ shift?
5. Predict & apply. β-carotene has eleven conjugated double bonds and looks orange. Using the trends from this lab, explain why it absorbs visible (blue) light while ethylene's absorption is far in the UV — and why we therefore see carotene as orange.
6. Benzene is cyclic and aromatic rather than a short open chain. Why is a simple linear box length only a loose proxy for its gap, and how does that caution your ranking in Q1?

How do an electron-donating vs electron-withdrawing group change where a ring is attacked?

ESP mapLUMO mapionization mapreactivity indices

You’ll learn to: read ESP and LUMO maps to predict where an aromatic ring is attacked, and classify substituents as activating (o/p) or deactivating (m).

Background 1 — electrophilic aromatic substitution follows electron density

In electrophilic aromatic substitution (EAS) an electron-poor electrophile attacks the ring's π cloud, so the reaction goes fastest where the ring is most electron-rich. A substituent that donates electron density — phenol's –OH, whose oxygen lone pair conjugates into the ring — is activating and steers the electrophile to the ortho/para positions, where the donated density piles up. A substituent that withdraws density — nitrobenzene's –NO₂, which pulls π density out by resonance and induction — is deactivating and meta-directing, because ortho/para attack would place positive charge next to the electron-hungry group.

Background 2 — reading the electrostatic-potential and frontier maps

An ESP map paints the molecular surface by electrostatic potential: red = electron-rich (negative, nucleophilic), blue = electron-poor. Phenol's ring face glows redder than benzene's — especially at ortho/para — while nitrobenzene's ring is drained toward blue. Because the electrophile is captured by the ring's filled orbitals, the HOMO and the local ionization energy map also flag the reactive sites: a lower ionization energy / higher HOMO amplitude marks where electrons are most available to share. These maps make the arrow-pushing "activating vs deactivating" rule something you can actually see.

Background 3 — putting a number on it: the electrophilicity index ω

Conceptual DFT condenses reactivity into indices from the frontier energies. The chemical potential μ ≈ ½(E_HOMO+E_LUMO) and hardness η ≈ (E_LUMO−E_HOMO) combine into the electrophilicity index ω = μ²/2η — how strongly a molecule wants electrons. Electron-withdrawing –NO₂ raises nitrobenzene's ω (a better electrophile, poorer at EAS), while electron-donating –OH lowers phenol's ω. These indices are derived from approximate orbital energies, so trust the ranking, not the absolute numbers.

Molecules

Procedure

1. Compute orbitals for each arene, then view the ESP map and the ionization-energy map — red marks the electron-rich sites an electrophile will attack.
2. Run DFT to get the conceptual-DFT reactivity indices, and note the electrophilicity ω and hardness η.
3. Decide which ring is activated and which is deactivated, and where each would react.

Questions

1. On the ESP maps, which ring face is reddest and which is bluest across benzene, phenol and nitrobenzene? Connect each colour to whether the substituent donates or withdraws electron density.
2. Why does –OH direct ortho/para while –NO₂ directs meta? Use resonance to explain where the donated (or withdrawn) density lands and why the electrophile follows.
3. Rank benzene, phenol and nitrobenzene by expected rate of EAS, and state which are activated and which deactivated relative to benzene.
4. The electrophilicity index is ω = μ²/2η, with μ ≈ ½(E_HOMO+E_LUMO) and η ≈ (E_LUMO−E_HOMO). Predict which molecule has the highest ω and explain what that says about its appetite for electrons versus its EAS reactivity.
5. EAS is governed by the ring's filled orbitals, yet a LUMO map is also offered. Why is the HOMO / ionization map the more relevant guide for predicting where an electrophile attacks?
6. Predict & apply. Aniline (–NH₂) and benzaldehyde (–CHO) aren't in the set. From the principles here, predict each one's directing effect and whether it activates or deactivates the ring.

Where are the bonding electrons and lone pairs really located? Count them for the first-row hydrides.

localized orbitalsorbital diagram

You’ll learn to: connect the delocalised orbitals quantum chemistry computes to the familiar Lewis bonds and lone pairs, and count them for the first-row hydrides.

Background 1 — canonical orbitals don't look like Lewis bonds

The molecular orbitals a quantum calculation hands you — the canonical orbitals, eigenfunctions of the Fock operator — are spread symmetrically over the whole molecule. Water's four occupied valence orbitals, for instance, are delocalised combinations with the molecule's full symmetry; none of them is a single "O–H bond" or a single "lone pair." This is correct physics, but it clashes with the intro-chemistry picture of two localized O–H bonds and two distinct lone pairs. The total electron density is identical either way — the question is only which orthogonal basis of orbitals you choose to describe it.

Background 2 — localization recovers the Lewis picture

Because any unitary mixing of the occupied orbitals leaves the total density and energy unchanged, you are free to rotate them into the most spatially compact, separated set. Localization schemes (Boys, Pipek–Mezey, Edmiston–Ruedenberg) do exactly this — Boys minimises the orbitals' spatial spread, Pipek–Mezey maximises atomic charge localization. The result is the familiar inventory: two-centre bonding orbitals sitting between bonded atoms and one-centre lone pairs on a single atom, plus an essentially untouched 1s core. Localized orbitals are the rigorous bridge from delocalised MOs to the dot diagrams of general chemistry.

Background 3 — counting pairs across the first-row hydrides

The central atom contributes its valence electrons to four electron pairs total — the octet. Carbon in methane has four bonding pairs and zero lone pairs; nitrogen in ammonia, three bonds and one lone pair; oxygen in water, two bonds and two lone pairs. That is precisely the VSEPR "four pairs around the central atom" that explains their shapes (tetrahedral → pyramidal → bent). Localization reproduces this exactly — though note real lone pairs are often two equivalent rabbit-ear orbitals rather than the textbook one-s/one-p split, and the count, not the precise shape, is the robust result.

Molecules

Procedure

1. Compute orbitals for each hydride, then click Localized (bonds) to transform them into bond and lone-pair orbitals.
2. Step through the localized orbitals and classify each as an X–H bond (spread over two atoms) or a lone pair (sitting on one atom).
3. Count the bonds and lone pairs and check they match the Lewis structure and VSEPR.

Questions

1. For methane, ammonia and water, state the number of bonding pairs and lone pairs the localized orbitals reveal on the central atom, and confirm each totals four valence pairs.
2. The canonical and localized orbitals describe the same molecule, yet look completely different. What is conserved when you switch from one set to the other, and why does that make the localization legitimate?
3. Methane, ammonia and water all have four electron pairs around the central atom but different shapes (tetrahedral, pyramidal, bent). Explain how the lone-pair count from this lab accounts for the shape change, linking back to VSEPR.
4. Each central atom also has a low-lying, nearly spherical orbital that doesn't participate in bonding. Identify it and explain why localization barely touches it.
5. The textbook draws water's two lone pairs as one s-rich and one p-type orbital, but a Boys localization often gives two equivalent "rabbit-ear" lone pairs. Which feature — the precise lone-pair shapes or the count of two — is the robust, method-independent result, and why?
6. Predict & apply. Using the same counting, predict the bonds and lone pairs a localization would give for hydrogen fluoride (HF) and for the hydronium ion (H₃O⁺), and the shape each implies.

How big is the energy barrier to rotating a single bond, and what does the curve look like?

dihedral scanchoose bondCSV download

You’ll learn to: map a single-bond rotational profile with a relaxed dihedral scan, and interpret its barrier heights and minima.

Background 1 — the torsional barrier and the relaxed scan

Rotating about a single bond traces a periodic energy curve, the torsional (rotational) potential, whose maxima are barriers between conformers. MoleBench maps it with a relaxed dihedral scan: the chosen dihedral ω is held fixed at each step while every other coordinate is re-optimized. This is essential — a rigid scan that froze the rest of the molecule would inflate barriers by forcing bad bond lengths and angles at the eclipsed points. The constrained minimisation lets the molecule relieve what strain it can, so the curve reflects the true rotation cost. The numbers come from GFN2-xTB, fast and qualitatively reliable for these small barriers.

Background 2 — ethane: hyperconjugation, not just sterics

Ethane gives the textbook case: a clean threefold (period 120°) curve with three equivalent staggered minima and three eclipsed maxima, a barrier of about 3 kcal·mol⁻¹. The minima aren't mainly about the tiny H···H steric clash; the dominant stabilisation is hyperconjugation — in the staggered geometry each C–H bonding orbital aligns to donate into the antiperiplanar C–H σ^* antibond, which it cannot do when eclipsed. The barrier is the cost of switching off that favourable σ→σ^* overlap.

Background 3 — butane and H₂O₂: richer curves

Butane's central C–C scan is no longer threefold-symmetric: the bulky methyls make a deep global anti minimum (180°), two shallower gauche minima (±60°) a bit higher, and an extra-tall syn-eclipsed barrier (~4–6 kcal·mol⁻¹) where the two methyls eclipse. Hydrogen peroxide breaks the pattern entirely: its minimum is non-planar (a skew O–O–H dihedral near 110–120°), a compromise between lone-pair–lone-pair repulsion (avoided at 0°) and favourable orbital interactions (lost at 180°). One tool, three strikingly different profiles — export the CSV to plot them.

Molecules

Procedure

1. Build and Optimize each molecule, then in panel 5 choose the bond to rotate (it highlights gold).
2. Scan it 0 → 360°; read the barrier height off the energy curve and count how many minima appear.
3. Click points on the curve to inspect the eclipsed and staggered geometries, and download the CSV.

Questions

1. Sketch the expected shape of ethane's scan: how many minima and maxima appear over a full 360°, what is its periodicity, and which geometries (staggered/eclipsed) correspond to each?
2. The standard ethane barrier is about 3 kcal·mol⁻¹. Explain why hyperconjugation — not steric H···H repulsion alone — is the leading reason the staggered form is lower.
3. Compare butane's central-bond curve to ethane's: identify the anti and gauche minima and the syn-eclipsed barrier, and explain why butane's curve loses ethane's threefold symmetry.
4. Hydrogen peroxide's minimum is neither planar-cis (0°) nor planar-trans (180°) but a skewed angle in between. Explain the two competing effects that push it off both planar geometries.
5. Why does MoleBench run a relaxed rather than a rigid scan, and how would the reported barriers differ if the rest of the molecule were frozen at each dihedral step?
6. Predict & apply. Using the staggered/anti minimum and ethane's ~3 kcal·mol⁻¹ barrier, estimate roughly how readily ethane interconverts between staggered forms at room temperature (thermal energy RT ≈ 0.6 kcal·mol⁻¹). Is the rotation effectively free?

Pull four properties together: how do substituents change the C=O bond, its π* orbital, its IR stretch and the dipole?

DFTLUMO mapIRdipole

You’ll learn to: cross-check four independent observables — geometry, π^* energy, IR stretch and dipole — on one functional group to build a consistent story.

Background 1 — one bond, four observables

The carbonyl is the same C=O group in formaldehyde, acetone and formic acid, yet four independent properties — the C=O bond length, the energy of the π^* (LUMO), the IR stretch and the molecular dipole — all shift when you change the substituents. They are not four separate facts: each reads out the same underlying change in how electron density sits on the carbon and oxygen. A capstone like this trains you to cross-check several numbers until one consistent electronic story explains them all, rather than trusting any single value in isolation.

Background 2 — how substituents push and pull

Acetone’s two methyls are weak electron donors (hyperconjugation + induction): they feed density toward the carbonyl carbon, easing the partial positive charge, lengthening C=O slightly and lowering the stretch. Formic acid’s –OH does two opposing things — an inductive σ-withdrawal of density and a resonance donation of an oxygen lone pair into the π^*. The lone-pair conjugation dominates the geometry, giving carboxylic acids a recognisably different carbonyl from a plain ketone or aldehyde, and a distinctly different dipole direction.

Background 3 — the π^*, the dipole and method caveats

The carbonyl LUMO is the π^* orbital, lobed largest on carbon — that is why nucleophiles attack carbon. A LUMO map paints this empty orbital onto the surface; reading it is qualitative, not a measured energy. The dipole points from the δ+ carbon toward the δ− oxygen and grows with C=O polarity. Remember these come from DFT (geometry, orbitals, dipole) and a harmonic IR treatment: absolute frequencies run ~5–10% high without scaling, and orbital energies are model-dependent, so trust the trends across the three molecules, not the raw digits.

Molecules

Procedure

1. For each carbonyl, run DFT (records the dipole and HOMO–LUMO gap).
2. Compute orbitals → LUMO map to see the electrophilic carbon, and Compute IR to read the C=O stretch.
3. Enter all four numbers in the table and look for correlations — does a bigger dipole track with a particular orbital or IR shift?

Questions

1. Cross-check the story: for each of the four observables (C=O length, π^*/LUMO energy, IR stretch, dipole), state whether you expect acetone to sit higher or lower than formaldehyde, and explain why a single electronic cause links all four.
2. The carbonyl LUMO is the π^* orbital. Why is its largest lobe on carbon rather than oxygen, and how does that explain why nucleophiles add to the carbonyl carbon?
3. Formic acid’s –OH both withdraws density inductively and donates a lone pair by resonance into the π^*. Which effect wins for the C=O geometry, and how would you tell from the bond length and stretch alone?
4. Acetone’s two methyls are electron-donating relative to formaldehyde’s two H’s. Predict the direction of the C=O stretch shift and the dipole change, and tie both to the same density change at the carbon.
5. Your computed IR stretch comes out near 1900 cm⁻¹ for formaldehyde, but the textbook aldehyde band is ~1730 cm⁻¹. What is the methodological reason for the gap, and why is the trend still trustworthy?
6. Predict/apply: acetyl chloride (CH₃COCl) has a strongly electron-withdrawing Cl with little π-donation. Relative to acetone, predict its C=O stretch, π^* energy and dipole, and say which lab molecule it should most resemble.

Acid strength is set by how stable the conjugate base is. Watch electron-withdrawing groups spread the negative charge.

charge −1ESP mappartial chargesconjugate base

You’ll learn to: explain acid strength through conjugate-base stability, and use ESP/charge maps to see an electron-withdrawing group delocalise the negative charge.

Background 1 — acidity lives in the anion

The most powerful shortcut for predicting acid strength is to stop looking at the acid and look instead at the conjugate base it leaves behind. A more stable anion means the proton leaves more readily, so the parent acid is stronger. All three molecules here are carboxylates, R–COO⁻; resonance already spreads the −1 charge equally over the two carboxylate oxygens. The question this lab visualises is what the R group does on top of that — whether it helps the molecule shoulder the remaining negative charge, and you watch that directly in the electrostatic-potential (ESP) picture.

Background 2 — the inductive effect, made visible

Fluorine and chlorine are highly electronegative, so each C–X bond is a dipole that pulls electron density through the σ framework toward itself — the inductive effect. In trifluoroacetate three C–F bonds tug density off the carboxylate and out onto the CF₃ end, draining negative charge away from the oxygens. Chloroacetate has one weaker puller; acetate’s methyl pulls nothing (it is a mild donor). The effect is purely electrostatic and falls off sharply with distance — this is through-bond polarisation, not π-conjugation onto the C–X bonds.

Background 3 — reading the ESP and partial-charge maps

An ESP map colours the molecular surface by the potential a point positive test charge would feel: deep red = most negative (electron-rich), fading to blue. A more delocalised anion shows its red spread thinner and over more of the surface; a poorly stabilised one shows an intense red puddle concentrated on the oxygens. Partial charges put a number on the same idea. Treat both as qualitative and method-dependent — atomic charges are not observables and depend on the partitioning scheme (Mulliken, Hirshfeld, …); compare the trend across the three, not absolute values.

Conjugate-base anions

Procedure

1. Build each conjugate-base anion — the charge box auto-fills −1 from the SMILES. Press Optimize & compute.
2. View the ESP map (orbitals → ESP) and judge how widely the negative (red) region is spread; record the most-negative partial charge.
3. Rank the anions by how delocalised the charge is, and predict the order of acid strength.

Questions

1. Read the picture: on the three ESP maps, where is the deepest red, and how does the spread of that red change from acetate to chloroacetate to trifluoroacetate? Connect what you see to which parent acid is strongest.
2. All three are carboxylates whose −1 charge is already shared over two oxygens by resonance. So what, specifically, do the halogens add that resonance alone does not — and why is that an inductive rather than a resonance effect?
3. From the partial charges, does the carboxylate oxygen carry more or less negative charge in trifluoroacetate than in acetate? Explain how draining charge off the oxygens stabilises the anion.
4. The inductive effect falls off with distance. Predict how the ESP and acidity of 3-fluoropropanoate (F two carbons further from the COO⁻) would compare with fluoroacetate, and why.
5. Quantitative: trifluoroacetic acid has pK_a ≈ 0.23 and acetic acid pK_a ≈ 4.76. Using ΔΔG = 1.36 · ΔpK_a (kcal/mol), find how much more the trifluoroacetate anion is stabilised, and convert the pK_a gap into a ratio of K_a.
6. Predict/apply: rank acetate, chloroacetate and trifluoroacetate by anion stability and by parent-acid strength, and state where you would place bromoacetate (Br less electronegative than Cl) in the ladder.

Why does a tertiary carbocation form so much more easily than a methyl one? See the empty p-orbital and where the + charge sits.

charge +1LUMO mapESP maphyperconjugation

You’ll learn to: rank carbocation stability, and explain the 3°>2°>1°>methyl ordering with hyperconjugation and the empty p-orbital.

Background 1 — the empty p-orbital that needs help

A carbocation is an sp², trigonal-planar carbon with only six valence electrons and an empty p-orbital perpendicular to the plane. That electron-deficient orbital is the cation’s liability — and its lifeline. Anything that can feed electron density into it lowers the energy and stabilises the ion. Across this series, methyl (CH₃⁺) has nothing but C–H bonds pointing the wrong way; each step toward tert-butyl swaps an H on the cationic carbon for a methyl group that brings new neighbouring bonds aligned to donate. More donors, more stable cation.

Background 2 — hyperconjugation and induction

Hyperconjugation is the donation of a filled adjacent σ_C–H or σ_C–C bond into the empty p-orbital: the bonding electrons partly delocalise onto the cationic carbon, sharing the positive charge over the whole alkyl framework. A 3° cation has nine β C–H bonds plus three C–C bonds positioned to overlap; methyl has none. A smaller, purely electrostatic inductive donation from the more-polarisable alkyl groups adds to this. Together they give the famous ordering 3° > 2° > 1° > methyl, with each substitution worth a sizeable stabilisation.

Background 3 — seeing it in the LUMO and ESP maps

For a cation the LUMO is essentially that empty p-orbital, and the LUMO map shows where an incoming nucleophile would attack. In methyl it is a clean lobe on one carbon; in tert-butyl it visibly spreads onto the methyl groups, the fingerprint of hyperconjugative delocalisation. The ESP map tells the complementary story: the most-positive (blue) region is intense and pinned on methyl’s carbon, but smeared and softened over the periphery in tert-butyl. Both maps are qualitative, orbital-shape and partial-charge pictures — method-dependent, so read the trend down the series.

Carbocations

Procedure

1. Build each cation (the charge auto-fills +1), press Optimize, then Compute orbitals.
2. View the LUMO map (the empty p-orbital) and the ESP map; note how much positive charge sits on the central carbon and whether the LUMO spreads onto neighbouring groups.
3. Rank the cations by stability and connect it to the number of adjacent C–H / C–C bonds.

Questions

1. Watch the orbital spread: compare the LUMO maps of methyl and tert-butyl cations. Where is the empty orbital localised in each, and how does that visual difference encode the 3° > 2° > 1° > methyl stability order?
2. Define hyperconjugation precisely. Why must the donating σ bond be on a carbon adjacent to the cationic centre and aligned with the empty p-orbital, rather than on the cationic carbon itself?
3. Count the number of β C–H (and C–C) bonds able to hyperconjugate in the ethyl, isopropyl and tert-butyl cations. How does the count track the stability ranking?
4. On the ESP maps, describe how the most-positive region changes from methyl to tert-butyl. Why does spreading the positive charge over more atoms lower the ion’s energy?
5. Apply to mechanism: SN1, E1 and Markovnikov addition all route through the most stable carbocation. Use the ordering to predict the major product when HBr adds to 2-methyl-2-butene, and explain via the intermediate.
6. Predict/apply: the benzyl cation (PhCH₂⁺) is a primary carbon yet remarkably stable. What stabilisation is available to it that none of these four aliphatic cations can use, and how would its LUMO map differ?

When you draw several resonance structures, are the "different" bonds actually identical? Measure them.

charge📏 measureπ orbitalsESP map

You’ll learn to: measure equalised bond lengths to demonstrate delocalisation, and dispel the misconception that resonance means the molecule “flips”.

Background 1 — resonance is one structure, not many

The single most misread idea in chemistry is that a molecule with several resonance structures flips between them. It does not. The Lewis structures you draw are incomplete bookkeeping; the real molecule is one fixed, averaged structure that exists all the time. Benzene never has alternating long and short bonds that swap; carbonate is not a fast equilibrium of one double and two single C–O bonds. The honest test is to measure the bonds your drawings label “single” and “double” — if delocalisation is real, they come out identical, an average of the limiting forms.

Background 2 — a delocalised π system and equalised bonds

Behind the equal bond lengths is a single delocalised π system: p-orbitals overlap continuously around the whole framework so the π electrons are shared evenly, not pinned in one location. Benzene’s six C–C bonds land at ~1.39 Å — squarely between a single (~1.54 Å) and a double (~1.33 Å) bond. Carbonate’s three C–O bonds are all equal, as are nitrate’s three N–O bonds. Each is a bond order of about 1⅓ (carbonate/nitrate) or 1½ (benzene), the exact average the resonance picture predicts — concrete proof that delocalisation, not flipping, is at work.

Background 3 — why delocalisation pays: stability and ESP

Spreading electrons over more atoms lowers their energy — this resonance (delocalisation) stabilisation is why benzene resists addition and why oxoanions like carbonate and nitrate are so stable and unreactive. The ESP map shows the complementary charge story: in carbonate the −2 and in nitrate the −1 charge is smeared evenly over all the oxygens, with no single oxygen redder than the rest — symmetry made visible. Bear in mind the geometry comes from DFT/GFN2-xTB and ESP from method-dependent partial charges, so weigh the equality of the bonds and the symmetry of the map, not the last decimal.

Molecules

Procedure

1. Build benzene, carbonate (charge auto-fills −2) and nitrate (−1), and Optimize each.
2. Use 📏 measure to compare the bonds that resonance structures draw as single vs double — are they actually equal?
3. Compute orbitals and view the delocalised π system and the ESP map to see the charge smeared evenly over the molecule.

Questions

1. Measure to settle it: after optimising, what do you find for benzene’s six C–C lengths, carbonate’s three C–O lengths and nitrate’s three N–O lengths? Explain how equal bonds disprove the idea that the molecule flips between resonance structures.
2. Benzene’s C–C bond is ~1.39 Å, between a single (~1.54 Å) and a double (~1.33 Å). What effective bond order does that imply, and how does it match the two main Kekulé structures?
3. What does the π-orbital view add that the bond-length measurement alone cannot? Describe how a delocalised π system over benzene (or over carbonate’s O–C–O) gives rise to the equal bonds you measured.
4. Read the ESP: on the carbonate and nitrate maps, is any one oxygen more negative than the others? Connect what you see to the equal bond lengths and to resonance stabilisation.
5. Carbonate carries −2 and nitrate −1, both over three oxygens. Estimate the average formal-charge-per-oxygen in each, and explain why spreading charge this way stabilises the anion (relate to the acid/conjugate-base reasoning of the carboxylate lab).
6. Predict/apply: the carboxylate ion (RCOO⁻) has two oxygens related by resonance but one alkyl-bearing carbon. Predict the relationship between its two C–O bond lengths, and contrast with a neutral ester where one C–O is clearly double and the other single.

What does a reaction actually look like at the top of its energy hill — and how do different reaction types compare?

reaction pathtransition stateactivation barriersaddle point + freqscrub the geometry

Five reactions: SN2 · E2 · Menshutkin · Diels-Alder · electrocyclic (Studio → panel 6, Reactions & transition states)

You’ll learn to: locate a transition state, verify it from its single imaginary frequency, and read off the activation barrier.

Background 1 — the saddle point and transition-state theory

On a reaction's potential-energy surface the reactants and products sit in valleys and are joined by a mountain pass. The transition state is the highest point along the lowest pass — a first-order saddle point: a maximum along the reaction coordinate but a minimum in every direction perpendicular to it. Transition-state theory says the rate depends exponentially on the height of that pass, the activation barrier ΔG^‡. A barrier a few kcal/mol higher slows a reaction by orders of magnitude, which is why locating and characterising the saddle point is the heart of this lab.

Background 2 — one imaginary frequency is the signature

How do you prove a stationary point really is a transition state and not just a minimum? Diagonalise the Hessian (the matrix of second derivatives of energy) to get the vibrational frequencies. A minimum has all real, positive frequencies; a genuine TS has exactly one imaginary frequency — a negative curvature whose normal mode is the bond-breaking/forming motion itself. Scrubbing along that mode literally animates the reaction: the S_N2 umbrella flip (≈328i cm⁻¹), the Diels–Alder bond formation (≈818i), the electrocyclic unzip (≈899i). Two or more imaginary frequencies means you found a higher-order saddle, not a real TS.

Background 3 — concerted mechanisms and method caveats

All three reactions here are concerted: bonds break and form in one step over a single barrier, with no intermediate. S_N2 inverts the carbon (Walden inversion, trigonal-bipyramidal-like TS); E2 demands an anti-periplanar H and leaving group so the developing π bond overlaps cleanly; the Diels–Alder forms two σ bonds at once (TS C···C ≈ 2.3 Å) and is strongly exothermic. Honest caveat: the fast GFN2-xTB scan gives an approximate TS quickly; the Psi4 refinement (HF/3‑21G or HF/6‑31G*) makes it a rigorous saddle point with the right shape and one imaginary mode — but these are learning-grade geometries, not benchmark barrier heights.

Procedure

1. Open panel 6, pick a reaction and press Run reaction (~30–90 s). Record the activation barrier and the reaction energy ΔE.
2. Press Jump to transition state ‡, then drag the slider back and forth to watch the bonds break and form — note exactly what's happening at the very top of the hill.
3. Press ✓ Refine TS (Psi4) — optionally choose HF/3-21G or HF/6-31G* — to rigorously optimize the saddle point and run a frequency check; confirm it has exactly one imaginary frequency, then Animate the reaction mode to see the atoms move along the reaction coordinate.
4. Now make your own: pick ✎ Custom in the reaction list, build a molecule, and click two atoms to mark a bond to form and/or break — try a ring-opening (break a C–C bond) or an intramolecular cyclization, then find and refine its TS just like the presets.

Questions

1. What geometric and mathematical features distinguish a transition state from a stable minimum on the potential-energy surface?
2. The diagnostic test: why must a genuine transition state have exactly one imaginary vibrational frequency — and what does the motion of that single imaginary mode physically represent?
3. At the S_N2 saddle point the central carbon is trigonal (planar three H's) with the nucleophile and leaving group both half-bonded. Explain how scrubbing through the imaginary mode demonstrates Walden inversion.
4. Why does the E2 transition state require an anti-periplanar arrangement of the β-H and the leaving group, and what would happen to the barrier if they were forced syn?
5. The Menshutkin reaction (two neutral molecules forming ions) climbs uphill the whole way in the gas phase, yet runs readily in solution. From transition-state and solvation reasoning, why does solvent rescue it?
6. Quantitative: using the Eyring equation k = (k_BT/h)·e^−ΔG‡/RT, estimate roughly how much the rate changes at 298 K if a refined barrier is 2 kcal/mol higher than a crude estimate. (RT ≈ 0.593 kcal/mol at 298 K.)

Molecules almost never react in a vacuum — how does putting them in a solvent change their properties and their reactions?

implicit solvationdipoleDFT (PCM)reaction barrier

You’ll learn to: predict how an implicit solvent changes a molecule’s dipole and a reaction’s barrier through the solvent reaction field.

Background 1 — the reaction field and continuum solvation

Place a polar solute in a polar liquid and the solvent molecules reorient so their dipoles oppose the solute's field. This polarised solvent generates a reaction field that points back at the solute, stabilising its charges and, in turn, pulling them further apart. Rather than track thousands of explicit solvent molecules, an implicit (continuum) solvation model replaces the solvent with a structureless dielectric of permittivity ε (water ≈ 78, acetonitrile ≈ 37). MoleBench uses xtb ALPB and Psi4 PCM; the higher ε, the stronger the reaction field and the larger the effect.

Background 2 — why dipoles grow in solution

The reaction field acts on a polar molecule like an external field, inducing extra polarisation: the molecule's electrons shift to deepen its charge separation, so the computed dipole moment is larger in solvent than in vacuum. For acetone the gas-phase dipole (≈2.8 D) swells toward ≈3.3 D in water. This is self-consistent — the solute polarises the dielectric, which polarises the solute back — and it is the same physics, scaled up, behind Born solvation of an ion: charge is stabilised in proportion to (1 − 1/ε).

Background 3 — differential solvation raises the S_N2 barrier

For an anionic S_N2 (e.g. Cl⁻ attacking), solvation does not stabilise reactant and transition state equally. The reactant has a compact, charge-concentrated chloride that a polar solvent solvates very strongly; at the TS that charge is smeared over a larger, half-bonded framework and is solvated more weakly. Stabilising the bottom of the hill more than the top raises the barrier — from ≈6 kcal/mol (gas) to ≈17 kcal/mol in water. Caveat: a continuum dielectric captures this bulk electrostatic trend but omits explicit hydrogen bonds, which matter for protic solvents.

Polar molecules (part 1)

Reaction (part 2): SN2 — Cl⁻ + CH₃Br (Studio → panel 6)

Procedure — part 1: solvent & dipole

1. Build acetone. In panel 2 leave solvent = gas phase and press DFT (B3LYP); record the dipole.
2. Change solvent = water and press DFT again. The result is labelled "water (implicit)" — record the new dipole. Try DMSO too.
3. Repeat for the other polar molecules. Does a more polar solvent enhance the dipole more?

Procedure — part 2: solvent & an SN2 barrier

1. Open panel 6, choose the SN2 reaction, set Solvent = gas phase and Run reaction; record the activation barrier.
2. Set Solvent = water and run it again. Compare the two barriers.

Questions

1. In an implicit solvation model, what physically replaces the individual solvent molecules, and which single solvent property controls how strong the effect is?
2. Why dipoles grow: explain in terms of the reaction field why acetone's computed dipole rises from ≈2.8 D in the gas phase to ≈3.3 D in water, and predict whether the increase would be larger or smaller in acetonitrile (ε ≈ 37).
3. The anionic S_N2 barrier jumps from ≈6 kcal/mol (gas) to ≈17 kcal/mol in water. Using differential solvation of reactant vs transition state, explain why the barrier goes up, not down.
4. Polar aprotic solvents (DMSO, acetonitrile) speed anionic S_N2 reactions relative to polar protic ones (water, alcohols). What does the continuum model capture about this, and what does it miss?
5. Quantitative/predict: an 11 kcal/mol increase in ΔG^‡ on going to water — using a factor of e^{−ΔΔG‡/RT} with RT ≈ 0.593 kcal/mol — corresponds to roughly what change in rate at 298 K? Why is "solvent isn't a detail" a fair summary?
6. Why does a continuum dielectric handle a neutral dipolar solute (acetone, formaldehyde) more faithfully than it handles a small, hydrogen-bonded anion in water?

Can a computer predict the ¹H and ¹³C NMR of a molecule well enough to assign every peak?

GIAO NMR¹H / ¹³Cintegrationassignment

You’ll learn to: predict ¹H and ¹³C NMR spectra and assign every peak using chemical shift and integration.

Background 1 — nuclear shielding and the chemical shift

In a magnetic field B₀ the molecule's electrons circulate and generate a tiny opposing field, so each nucleus feels a slightly reduced effective field: it is shielded. The shielding constant σ measures that reduction, and it depends on the electron density and bonding around the nucleus. Electron-rich environments shield (low δ, upfield); electron-poor or deshielded environments (next to electronegative atoms, in π ring currents) push downfield (high δ). The chemical shift reports σ relative to a reference: δ = (σ_ref − σ_sample), which is why a CH₃ appears near 1–2 ppm but an aromatic H near 7 ppm.

Background 2 — GIAO and empirical scaling

Computing σ in a magnetic field has a notorious problem: the answer should not depend on where you place the coordinate origin, but a finite basis set makes it do so. GIAO (gauge-including atomic orbitals) attaches a field-dependent phase to each basis function, restoring origin independence and giving reliable shieldings. MoleBench runs GIAO with real B3LYP DFT. Raw shieldings are then converted to shifts by an empirical linear scaling against an experimental calibration set, which absorbs systematic method error — making predictions quantitative to about ¹H ≈ 0.2 ppm and ¹³C ≈ 8 ppm across the whole range.

Background 3 — assignment: position and integration

Assigning a spectrum joins two pieces of information. Position (δ) reports the electronic environment: ethanol's CH₂ at ≈3.7 ppm sits next to electronegative oxygen, its CH₃ at ≈1.2 ppm does not; toluene's aromatic H's near 7.1–7.3 are deshielded by the ring current. Integration counts equivalent nuclei — ethanol's 3H : 2H ratio fingerprints CH₃ vs CH₂. On ¹³C the same logic places benzene at ≈128.7 ppm (lit 128.4) and a ketone C=O near 206 ppm (lit 207). Honest caveat: very polar O–H/N–H protons scatter, because the gas-phase model omits the hydrogen bonding that shifts them in real solutions.

Molecules

Procedure

1. Build each molecule and press NMR (¹H / ¹³C) in panel 3. Record the predicted ¹H and ¹³C peaks and their integration (the number in parentheses / the stick height).
2. Assign every peak to an atom or group — e.g. for ethanol the 3H peak ≈ 1.2 ppm is the CH₃, the 2H ≈ 3.7 ppm is the CH₂.
3. For toluene, count how many distinct carbon environments the prediction finds and match them to the ring (ipso/ortho/meta/para) plus the methyl.

Questions

1. What is nuclear shielding, and why does a higher shielding constant σ correspond to a peak farther upfield (lower δ)?
2. The two clues: in assigning ethanol's ¹H spectrum, how do position (≈3.7 vs ≈1.2 ppm) and integration (2H vs 3H) together let you assign the CH₂ and CH₃ peaks unambiguously?
3. Toluene's aromatic protons appear near 7.1–7.3 ppm while its methyl sits at ≈2.3 ppm. What deshields the ring protons so strongly, and why is the methyl comparatively shielded?
4. Why is a coordinate-origin problem (the gauge problem) inherent to computing shieldings, and how does the GIAO approach fix it?
5. The raw computed shieldings are run through an empirical linear scaling before being reported as δ. What systematic errors does that absorb, and why does it let even hard cases like a ketone C=O (≈206 ppm) come out accurate?
6. Quantitative: on a 400 MHz spectrometer (¹H), how many Hz apart are ethanol's CH₂ (3.7 ppm) and CH₃ (1.2 ppm) signals? (Hint: 1 ppm = spectrometer frequency in MHz, in Hz.)

A 1D scan rotates one bond — but real flexible molecules have several. What does the full energy surface look like?

2D scanpotential-energy surfaceMMFF94click-to-view

You’ll learn to: read a two-dimensional energy surface — basins, ridges and saddle points — and recognise effects (like syn-pentane) that a 1D scan misses.

Background 1 — from a 1D scan to a 2D energy surface

Rotating one bond traces a 1D energy curve; let two dihedrals vary together and the energy becomes a surface over a plane of two torsion angles. That surface has basins (stable conformers at the bottom of valleys), ridges (rotational barriers), and saddle points that mark the lowest passes connecting one basin to another. Reading it as a contour or heat map shows not just where the minima are but how high the walls between them sit — the whole conformational landscape in one picture. MoleBench builds this with the MMFF94 force field, which is fast and qualitatively reliable for conformational energies.

Background 2 — the Ramachandran connection and 1D blind spots

This is exactly the logic of the Ramachandran plot, where a protein backbone's allowed shapes are mapped over its two torsions φ and ψ. The key lesson is that a 2D map reveals coupling a 1D scan cannot: two torsions that are each individually fine can clash when set simultaneously. In pentane both central C–C bonds matter; the deep minimum is anti–anti (both ≈180°), gauche basins sit a few kcal/mol up, and the corners are eclipsed, high-energy geometries.

Background 3 — syn-pentane interference and an H-bonded basin

The signature coupling effect is syn-pentane interference: a gauche turn on each torsion is individually low-energy, but two gauche turns of opposite sign force the terminal methyls into direct steric contact, producing a high-energy spot a 1D scan would never flag. Ethylene glycol shows the opposite surprise: a gauche O–C–C–O basin is stabilised — not by sterics but by an intramolecular O–H···O hydrogen bond, the same effect seen in the 1D conformer lab, now appearing as a distinct basin on the surface. Caveat: MMFF94 is a classical force field, fast and qualitatively right, not a quantum benchmark.

Molecules

Procedure

1. Build pentane. In panel 5 under 2D conformational map, the two central C–C–C–C torsions are pre-chosen as bond X and bond Y. Press Run 2D map.
2. Read the heatmap (blue = low, red = high). Find the anti–anti minimum (both ≈ 180°) and the high-energy regions. Click cells to load and inspect those conformers.
3. Now try ethylene glycol (drive the H–O–C–C and O–C–C–O torsions): look for a low-energy basin that corresponds to the internal O–H⋯O hydrogen bond.

Questions

1. On a 2D potential-energy surface, what conformational feature does each of a basin, a ridge, and a saddle point correspond to physically?
2. Why does pentane's deepest minimum sit at anti–anti (both central torsions ≈180°), and what kinds of geometries occupy the high-energy corners of the map?
3. Syn-pentane interference: each central bond can adopt a gauche turn at only a small energy cost, so why does combining two specific gauche turns produce a high-energy spot — and why is this effect invisible to a single 1D dihedral scan?
4. Ethylene glycol's surface has a stabilised gauche O–C–C–O basin instead of the simple anti minimum a pure-steric molecule would show. What stabilises it, and how does that echo the 1D conformer lab?
5. How is this 2D map the same idea as a protein's Ramachandran plot, and what general lesson about multi-bond molecules does that comparison teach?
6. Predict/apply: a basin sits ≈3 kcal/mol above the global minimum. Using a Boltzmann factor e^−ΔE/RT at 298 K (RT ≈ 0.593 kcal/mol), estimate its relative population, and comment on whether MMFF94 energies are reliable enough to trust that number quantitatively.

Acetic acid and trifluoroacetic acid differ by just three fluorine atoms, yet trifluoroacetic acid is roughly 30,000× stronger. Where does that factor come from — and can you see it in a predicted pK_a, in the molecule's descriptors, and in a correlated‑level energy?

pK_a (GFN2/ALPB)inductive vs resonanceQSAR descriptorsMP2 correlationHammett

You’ll learn to: estimate a pK_a, connect it to molecular descriptors and inductive/resonance effects, and interpret the MP2 correlation energy.

Background 1 — pK_a is really a free energy

An acid sits in equilibrium with its conjugate base, HA ⇌ H⁺ + A⁻, with K_a = [H⁺][A⁻]/[HA] and pK_a = −log₁₀K_a. Because the position of any equilibrium is fixed by its free energy, pK_a is a disguised thermodynamic quantity: ΔG° = 2.303 RT · pK_a. At 298 K that prefactor is 1.36 kcal/mol per pK_a unit — so an acid whose pK_a is 5 units lower has a conjugate base about 7 kcal/mol more stable. The golden rule follows: almost everything that makes an acid stronger does so by stabilising A⁻.

Background 2 — two ways a substituent stabilises the anion

Electron‑withdrawing groups stabilise A⁻ by two distinct routes. The inductive effect pulls density through the σ‑bond framework; it is strongest for the most electronegative atoms (F > Cl > Br) and falls off steeply with distance — which is exactly why α‑halogens on a carboxylic acid lower its pK_a so much, and three fluorines make trifluoroacetic acid one of the strongest simple organic acids. The resonance (mesomeric) effect delocalises charge through the π system and needs conjugation plus the right geometry: a para‑nitro group lets the phenolate lone pair spill onto the –NO₂ oxygens, something a meta nitro cannot do. Physical‑organic chemistry compresses all of this into Hammett σ constants (for –NO₂, σ_para = +0.78; more positive = more withdrawing).

Background 3 — how the pK_a button works (and its limits)

MoleBench estimates pK_a from a thermodynamic deprotonation cycle. For every ionisable X–H (X = O, N, S) it optimises the neutral acid and its conjugate base in implicit water (ALPB continuum solvent) at GFN2‑xTB and takes ΔE = E(A⁻) − E(HA), then maps it to pK_a = m·ΔE + b through a line fitted to 15 acids of known pK_a across 0–16 (R² = 0.91, mean error ≈ 1.3 units). It reports the most acidic site and lists every candidate. Treat the value as an estimate (±~1–2 units) — one global line cannot perfectly fit carboxylic acids, phenols, alcohols and thiols together, and it omits explicit waters and full entropy — but the trends are reliable, which is what chemical reasoning usually needs.

Background 4 — what the descriptors measure

cLogP is the calculated oil/water partition coefficient (lipophilicity), summed from Crippen atom contributions. Molar refractivity (MR) scales with molecular volume × polarisability — how big and how “squishy” the electron cloud is. TPSA (topological polar surface area) sums the surface of N/O/polar‑H atoms and is a strong proxy for H‑bonding and membrane permeability. QED rolls several properties into one 0–1 drug‑likeness score. Keep one thing in mind that this lab is designed to expose: polarisability (MR) and electronegativity are different atomic properties, and they will openly disagree below.

Background 5 — Hartree–Fock, correlation, and MP2

Hartree–Fock (HF) is a mean‑field method: each electron feels only the average field of the others, so it misses the way electrons instantaneously avoid one another. The energy it leaves out is the correlation energy, E_corr = E_exact − E_HF, which is always negative. MP2 (second‑order Møller–Plesset perturbation theory) adds most of it back on top of the HF wavefunction. Correlation is decisive for reaction energies, dispersion / π‑stacking and weak interactions. The catch: total energies are only comparable within the same method and basis — an MP2 number for one molecule and an HF number for another say nothing on their own.

An acidity ladder (click any to load it)

Procedure — part 1: build the pK_a ladder

1. Build acetic acid; in panel 2 press Estimate pK_a. Record the predicted pK_a and which site it chooses (it lists every O/N/S–H).
2. Repeat for chloroacetic then trifluoroacetic acid. Each extra electronegative atom pulls harder on the carboxylate — watch the pK_a fall, and note how much each halogen is worth.
3. Now compare phenol with p‑nitrophenol. Predict the direction first, then test it.

Procedure — part 2: read the descriptors

1. For each of the three carboxylic acids, click QSAR descriptors in panel 1. Record cLogP, molar refractivity, TPSA and QED.
2. Going acetic → chloro → trifluoro, mark which descriptors grow, which barely move, and — importantly — whether the descriptor that grows the most is the same property that drives the acidity.

Procedure — part 3: Hartree–Fock vs MP2

1. Build acetic acid. Open ⚙ Advanced in panel 2, choose Hartree–Fock, basis 6‑31G*, task Single‑point energy, Run; record the energy.
2. Switch only the method to MP2 (correlated), keep 6‑31G*, Run again. The difference is the correlation energy recovered by MP2.

Questions

1. Write the deprotonation equilibrium for acetic acid and draw its conjugate base. State the golden rule that links A⁻ stability to pK_a.
2. Using ΔG° = 1.36 · pK_a (kcal/mol, 298 K) and the experimental values, how much more stable is trifluoroacetate than acetate? Show the arithmetic.
3. Trifluoroacetic acid and p‑nitrophenol are both strengthened by an electron‑withdrawing group. Classify each effect as inductive or resonance and justify it from where the group sits relative to the negative charge. Why would meta‑nitrophenol be a weaker acid than the para isomer?
4. The key insight. Molar refractivity rises from acetic to chloroacetic but is almost unchanged for trifluoroacetic — even though fluorination makes the acid far stronger. Explain why acidity and MR track different atomic properties.
5. All three carboxylic acids share TPSA = 37.3 Å², yet phenol is 20.2 and p‑nitrophenol 63.4. Account for all three numbers.
6. From your part‑3 values, give E_corr for acetic acid in Eh and kcal/mol, and state why it must be negative. Why can you not compare acetic acid's MP2 energy with phenol's HF energy to decide which molecule is more stable?
7. Apply it. A drug candidate's –COOH is so acidic it is fully ionised in the gut, hurting permeability. Suggest one structural change that would raise its pK_a, and predict the sign of the change in cLogP and TPSA.

Going further: walk fluoroacetic → difluoro → trifluoroacetic acid and test whether each fluorine lowers pK_a by a similar step (inductive additivity); compare ortho/meta/para‑nitrophenol; or build the conjugate base directly (set charge −1, drop the acidic H) and run descriptors on the anion.

A single hemoglobin molecule is four protein chains wrapped around four iron-bearing hemes — so how do biologists organize the journey from a flat string of amino acids to a working oxygen carrier?

protein structuresecondary structurequaternary structurehemoglobinbeginner

You’ll learn to: read a one-letter sequence in the Sequence panel, identify α-helices and β-sheets by coloring a structure, isolate individual chains, recognize a multi-subunit (quaternary) assembly versus a single-domain protein, and record a side-by-side comparison of two real structures.

Background 1 — the four levels of structure

Protein architecture is described in a strict hierarchy. Primary structure is the linear sequence of amino acids joined by peptide bonds — the order encoded by the gene. Secondary structure is local backbone folding stabilized by backbone hydrogen bonds, giving repeating α-helices and β-sheets. Tertiary structure is the full three-dimensional fold of one chain, packing those elements together. Quaternary structure is the assembly of two or more folded chains (subunits) into one functional unit. Each higher level is built from the one below.

Background 2 — why hemoglobin is a tetramer

Human hemoglobin is a tetramer: two identical α-subunits and two identical β-subunits (an α₂β₂ arrangement), each cradling one heme group whose iron binds O₂. The subunits are not just decorative company — they communicate. Oxygen binding at one heme nudges the others into a higher-affinity shape, a quaternary phenomenon called cooperativity that gives hemoglobin its efficient sigmoidal O₂ curve. A lone subunit (like the related protein myoglobin) cannot cooperate. By contrast, crambin (1CRN) is a tiny 46-residue plant protein: a single chain, a single domain, no quaternary structure at all.

Background 3 — how a viewer maps levels onto colors and panels

A structure viewer lets you see each level by switching representation and color. Color by structure assigns one color to α-helix, another to β-strand, and a third to loops, so secondary structure becomes visible at a glance. Color by chain gives each subunit its own color, exposing the quaternary arrangement directly. The Sequence panel is the primary structure made clickable, and isolating one chain strips the view down to a single tertiary fold. A key habit is matching the question to the right tool: ask “which secondary structure?” with color by structure, but ask “how many subunits?” with color by chain. Resolution and atom counts in the Structure metadata tell you how much detail the crystallography actually resolved.

Structures — open in the workbench ↗

Procedure

1. Open 4HHB in the workbench. Read the Structure (i) panel and note the method, resolution (Å), organism, and the counts of chains / residues / atoms / ligands. How many chains? How many ligands (these include the hemes)? Record these in the worksheet.
2. In Representation (1), confirm Show as → cartoon. This is the standard way to read tertiary structure — the ribbon traces the folded backbone path.
3. Still in Representation, set Color by → structure. Helices, sheets, and loops now get distinct colors. Spin the view (toolbar ◌ spin, then click ◌ again to stop and drag manually) and survey the fold. Is hemoglobin mostly helical, mostly sheet, or mixed? Record it.
4. Open the Sequence (5) panel. This one-letter string is the primary structure. Click any residue to highlight and zoom to it in the 3D view — watch where it lands on the cartoon.
5. In Chains (2), click isolate on the first chain (A). You now see one folded subunit on its own — this is the tertiary structure of a single α-subunit. Note its compact globin fold.
6. Click Show all in Chains, then in Representation set Color by → chain. The four subunits now get four colors — this is the quaternary structure. Identify the two α and two β chains by how they pair across the assembly.
7. In Ligands & binding sites (3), click focus on one heme to zoom to it, then contacts to see how it is held in its pocket. Note that there are four hemes — one per subunit.
8. Optionally, in Display (4) use 📏 measure: click an iron atom in one heme, then an iron in a neighboring subunit, to read the inter-heme distance (Å) and appreciate how far apart the cooperating sites are.
9. Now open 1CRN in the workbench. Read Structure (i): how many chains and residues? Set Color by → structure and compare crambin’s size and secondary-structure content to a single hemoglobin chain. Fill in the 1CRN column of the worksheet.
10. Use Saved views (6) to Save the structure-colored hemoglobin tetramer so you can restore it for comparison.

Record what you observe in both structures. Reference values are pre-filled; fill the shaded cells from the Structure (i) panel and your colored views.

Questions

1. After Color by → structure, which secondary-structure type dominates hemoglobin, and which is essentially absent? What is the name of the all-α fold each subunit adopts?
2. The Sequence panel shows a string of letters. Which level of structure is this, and what kind of bond links each letter to the next?
3. When you isolated chain A, which structural level were you looking at? When you colored all four chains differently, which level?
4. From your worksheet: how many chains, hemes, and total residues did 4HHB report versus 1CRN, and what secondary-structure type did each show? Which structural level does that residue/chain count let each protein reach?
5. Hemoglobin carries four hemes and four chains, while myoglobin has one of each. Why does the tetrameric arrangement matter functionally — what can hemoglobin do that a single subunit cannot?
6. Predict/apply: Sickle-cell anemia comes from a single Glu→Val change in the β-chain. Which structural level is altered directly, and explain how a change at that lowest level can still wreck the protein’s higher-level behavior.

A protein chain could fold a near-infinite number of ways — so what physical force reliably pushes the same sequence into the same compact shape every single time?

protein foldinghydrophobic effectthermodynamicsAnfinsencrambin

You’ll learn to: color a structure by hydrophobicity, distinguish buried core residues from surface residues, predict residue location from side-chain chemistry and confirm it in 3D, and connect the hydrophobic effect to the entropy of water as the driving force of folding.

Background 1 — the hydrophobic effect drives folding

Folding is not pulled together by the nonpolar side chains “liking” each other — it is pushed by water. Around an exposed nonpolar group, water molecules cannot hydrogen-bond to it, so they form an ordered clathrate-like cage, losing entropy. When the protein buries its nonpolar (hydrophobic) side chains together in a core, that caging water is released to the bulk, and its entropy rises sharply. This favorable increase in water entropy (−TΔS) is the main negative-ΔG term of folding — the hydrophobic effect. Polar and charged residues, which water can solvate happily, stay on the surface.

Background 2 — Anfinsen: sequence encodes structure

Christian Anfinsen showed that a denatured protein can spontaneously refold to its correct, active shape with no outside help — so all the folding information lives in the primary sequence. The folded state is the thermodynamic minimum: the most stable arrangement reachable for that sequence in water. The pattern of hydrophobic versus polar residues along the chain therefore acts like a code — it tells the chain which residues belong inside and which belong outside. Crambin (1CRN), a 46-residue plant seed protein with a well-resolved structure, is a classic miniature example of a clean hydrophobic core, with charged residues neatly displayed on its surface.

Background 3 — reading hydrophobicity from side-chain chemistry

You can predict where a residue lives before ever opening the structure. Aliphatic side chains (Leu, Ile, Val, Ala) and aromatic ones (Phe, Trp) are pure hydrocarbon — nonpolar, so they bury in the core. Charged residues (Lys, Arg positive; Asp, Glu negative) and polar ones (Ser, Thr, Asn, Gln) carry hydrogen-bonding or ionizable groups, so they favor the water-facing surface. Some residues are amphipathic and sit at the boundary. Hydrophobicity scales (Kyte–Doolittle, for example) rank these quantitatively, and a viewer’s color-by-hydrophobic mode applies essentially the same scale — so your chemical prediction and the colored 3D view should agree. Disulfide bonds (crambin has three) also help lock the small fold.

Structures — open in the workbench ↗

Procedure

1. Open 1CRN in the workbench. Read Structure (i): note that crambin is a single chain of about 46 residues — small enough to inspect every side chain.
2. In Representation (1), set Show as → cartoon and Color by → hydrophobic. Orange marks hydrophobic residues; blue marks polar/charged residues.
3. Make sure the toolbar ◌ spin is off, then drag to rotate by hand. Slowly turn the molecule and watch where the orange clusters — toward the middle or the outside?
4. In Representation, switch Show as → surface. You now see the solvent-facing outside of the protein. Note whether the visible surface is mostly orange or mostly blue.
5. Switch Show as → sticks (or ribbon+lig) while keeping Color by → hydrophobic. Rotate and look into the molecule: the orange side chains you now see packed in the interior are the buried core.
6. Open the Sequence (5) panel. Before clicking, fill the worksheet’s prediction columns from side-chain chemistry. Then click each listed residue (a Leu, Ile, Phe, Val; then a Lys, Arg, Asp, Glu, Ser) to zoom to it, and confirm in 3D whether it sits buried or on the surface.
7. In Display (4), toggle 💧 water off if any crystal waters are shown, so you can see the protein surface cleanly. Optionally toggle residue labels to confirm residue identities, and use 📏 measure to gauge how deep a core residue sits.
8. Use Saved views (6) to Save the surface view and the sticks/core view, so you can flip between “outside” and “inside” for the discussion.

For each residue, write your prediction first (from Background 3), then confirm it using Color by → hydrophobic with the surface and sticks views. Fill the shaded cells.

Questions

1. After Color by → hydrophobic, where do the orange (hydrophobic) residues concentrate — the interior or the surface? Where do the blue (polar/charged) residues sit?
2. In the surface representation, is the solvent-exposed outside mostly orange or mostly blue? Why does that make thermodynamic sense?
3. The hydrophobic effect is often called “entropy-driven.” Whose entropy increases when the core forms — the protein’s or the water’s — and why?
4. From your worksheet: did every prediction match what you saw in the viewer? List which of Leu/Ile/Phe/Val you confirmed as buried and which of Lys/Arg/Asp/Glu/Ser you confirmed as surface, and note any residue that sat at the boundary instead.
5. How does this experiment illustrate Anfinsen’s principle that the sequence encodes the structure?
6. Predict/apply: A mutation buries a charged Asp deep in the hydrophobic core, or exposes a big hydrophobic Phe on the surface. For each case, predict whether the protein becomes more or less stable, and why.

How does an enzyme “know” exactly where to cut a protein — and why does trypsin always slice right after an arginine or lysine?

enzymesactive siteserine protease

You’ll learn to: identify the residues lining an enzyme’s binding pocket, recognise a salt bridge that drives specificity, measure a key contact distance in 3D, and tabulate the network of contacts that holds an inhibitor in place.

Background 1 — The catalytic triad

Trypsin is a serine protease: it hydrolyses peptide bonds using a catalytic triad of three cooperating residues — Ser195, His57 and Asp102. His57 acts as a base, pulling a proton off the Ser195 hydroxyl so its oxygen becomes a powerful nucleophile that attacks the substrate’s carbonyl carbon. Asp102 stabilises the protonated histidine, sharpening its catalytic power. This charge-relay system is the chemistry of the cut — but on its own it says nothing about where trypsin cuts. That job belongs to a separate pocket.

Background 2 — The S1 specificity pocket

Next to the triad sits the S1 specificity pocket, a deep slot that grips the side chain just before the bond to be cleaved. At its floor lies Asp189, a negatively-charged aspartate. Only a long, positively-charged side chain — arginine or lysine — can reach the bottom and form a stabilising salt bridge with Asp189. That single buried charge dictates trypsin’s “cut after Arg/Lys” rule. Here we use the inhibitor benzamidine (ligand BEN): its flat, positively-charged amidinium group mimics an Arg/Lys side chain, so it lodges in S1 and freezes the enzyme.

Background 3 — The oxyanion hole and tetrahedral intermediate

Recognition and the triad are only part of the story. As Ser195 attacks the carbonyl carbon, the substrate passes through a high-energy tetrahedral intermediate whose carbonyl oxygen carries a developing negative charge — an oxyanion. Trypsin stabilises it in an oxyanion hole, a cradle of backbone amide N−H groups from Gly193 and Ser195 that donate hydrogen bonds to the oxyanion. This transition-state stabilisation is the true source of catalytic rate enhancement: the enzyme binds the fleeting intermediate far more tightly than the substrate, lowering the activation barrier by orders of magnitude. Substrate-mimicking subsites (S1′, S2) flank the scissile bond and help register the polypeptide for an exact cut.

Structures — open in the workbench ↗

Procedure

1. Open 3PTB in the workbench. In Structure (i), note the method, resolution, organism, and that there is one protein chain plus the ligand BEN.
2. In Representation (1) set Show as to cartoon and Color by to structure to see the two β-barrel domains typical of this enzyme family. Switch Color by to element later when you inspect contacts.
3. In Ligands & binding sites (3), click focus on BEN to zoom in on benzamidine, then click pocket. The residues within 5 Å appear as labeled sticks — this is the lining of the S1 specificity pocket. Find the label for Asp189 at the floor of the slot, and note Ser190 and Gly219 nearby.
4. Click contacts on BEN. Polar contacts are drawn as dashed lines with distances. Identify the salt bridge running from the benzamidine amidinium nitrogen to the carboxylate oxygen of Asp189, plus any backbone H-bonds to Gly219 or Ser190.
5. In Display & selection (4) turn on residue labels, then enable 📏 measure. Click one amidinium N atom of benzamidine, then the nearest carboxylate O of Asp189; record the distance (expect ~2.8–3.0 Å). Repeat for an amidinium N → Gly219 backbone carbonyl O and for an amidinium N → Ser190 contact.
6. Use the Sequence (5) panel to click residues 57, 102 and 195; watch them highlight and zoom. Note how close His57 and Ser195 sit to the bound inhibitor — the catalytic machinery is poised right beside S1. Toggle 💧 water in Display (4) to see ordered waters in the pocket, then save a Saved view (6).

Questions

1. The S1 pocket floor carries Asp189. Explain in terms of charge why this makes trypsin cleave specifically after Arg and Lys, and not after a negatively-charged residue like Glu.
2. From your worksheet, what distance did you measure between the benzamidine amidinium N and the Asp189 oxygen? Is this consistent with a hydrogen-bonded salt bridge rather than a covalent bond, and how does it compare to your Gly219/Ser190 contacts?
3. Benzamidine is described as a competitive inhibitor. Using what the pocket view showed, explain how a small molecule that is not a peptide can still block trypsin.
4. The catalytic triad, the S1 pocket and the oxyanion hole are physically separate features. What distinct job does each one perform, and why does an enzyme need all three?
5. The amidinium group makes more than one polar contact to the pocket (you recorded several). Why does a network of contacts — rather than a single bond — give such high binding specificity and affinity?
6. If you mutated Asp189 to lysine, predict how the enzyme’s substrate preference would change.

How do you design a drug that has never existed — by staring at the very enzyme you need to stop?

drug designHIV proteaseinhibitor

You’ll learn to: recognise a symmetric enzyme dimer, see how an inhibitor grips a catalytic cleft, measure the bridge to the catalytic dyad, and assess a drug’s drug-likeness against Lipinski’s rule with a quantitative worksheet.

Background 1 — A symmetric aspartic protease

HIV-1 protease is the molecular scissors the virus uses to chop its long polyprotein precursors into functional proteins — an essential maturation step. It is an aspartic protease: catalysis is performed by two Asp25 residues that activate a water molecule to hydrolyse the peptide bond. Crucially, the enzyme is a homodimer — two identical chains related by C2 symmetry, each contributing one Asp25 to a single shared active site at the dimer interface. Block that one cleft and the virus cannot mature. That single, druggable site made the protease a prime target.

Background 2 — Structure-based drug design

Structure-based drug design (SBDD) means using a target’s 3D structure to guide the shape, charge and hydrogen-bonding of a candidate molecule. Indinavir (ligand MK1) is a peptidomimetic: it imitates the transition state of the natural substrate but resists being cut, so it lodges permanently in the cleft. A central hydroxyl reaches between the two catalytic aspartates while bulky arms fill the symmetric subsites, gripping the enzyme far more tightly than substrate. Because indinavir was tailored to a real structure, it became a clinical antiretroviral — though, as you’ll see, it sits at the edge of conventional drug-likeness.

Background 3 — The flaps, the bridging water, and resistance

Two glycine-rich β-hairpin flaps fold down over the cleft and clamp the inhibitor in place, sealing it from solvent. In substrate complexes a single structural water bridges the inhibitor’s carbonyls to the flap Ile50 backbone amides — a tetrahedrally-coordinated water unique to retroviral proteases that later inhibitors were designed to displace. Because the virus mutates rapidly, residues lining the pocket (e.g. near positions 82, 84, 90) drift to weaken inhibitor contacts, producing drug resistance. SBDD therefore aims for contacts to conserved main-chain atoms and the catalytic Asp25 dyad, which the virus cannot easily mutate without crippling its own enzyme — a durable grip on an unchangeable feature.

Structures — open in the workbench ↗

Procedure

1. Open 1HSG in the workbench. Read Structure (i): method, resolution, and the residue/chain counts. Confirm there are two protein chains plus the ligand MK1 (indinavir) and waters.
2. In Representation (1) choose Show as → cartoon and Color by → chain. The two chains are drawn in different colours — this is the homodimer. In Chains (2), click isolate on each chain in turn to see they are identical halves, then click Show all.
3. In Ligands & binding sites (3), click focus on MK1 to zoom to indinavir buried at the centre, then click pocket to reveal the lining residues as labeled sticks. Note how residues from both chains contribute — the drug straddles the dimer interface. Look for the Ile50 flap residues capping the cleft.
4. Click contacts on MK1. Observe the dashed polar contacts and distances showing how indinavir hydrogen-bonds to the cleft, including near the catalytic Asp25 pair. Use the Sequence (5) panel to click residue 25 in each chain and watch both catalytic aspartates highlight.
5. Turn on 📏 measure in Display & selection (4) and measure from indinavir’s central hydroxyl O to a nearby Asp25 carboxylate O to confirm it bridges the catalytic dyad (record this; expect ~2.7–3.2 Å). Toggle 💧 water to look for an ordered flap-bridging water near Ile50.
6. Click ⚛ Studio on MK1 to send indinavir into the molecule Studio (its SMILES loads). Read off MW, cLogP, H-bond donors, H-bond acceptors and TPSA, plus the Lipinski drug-likeness summary. Enter each into the worksheet below and mark Pass/Fail. Back in the workbench, Save view (6).

Questions

1. HIV protease is a homodimer with one active site at the interface. Why is dimerisation essential for catalysis, and why does that make dimer-disruption an alternative drug strategy?
2. From the contacts view and your measured hydroxyl–Asp25 distance, describe how indinavir physically blocks catalysis. Why does the enzyme fail to cut a molecule that fills its own active site?
3. Using your completed Lipinski worksheet, evaluate indinavir against the rule of five (MW < 500, cLogP < 5, donors ≤ 5, acceptors ≤ 10, TPSA ≤ ~140 Ų). Which row(s) fail, and by roughly how much does MW exceed the limit?
4. Indinavir is a large peptidomimetic that bends Lipinski’s guidelines yet still works as an oral drug. What does that tell you about treating drug-likeness rules as guides rather than laws?
5. Background 3 noted the rapidly-mutating virus. Why is targeting contacts to the conserved Asp25 dyad and main-chain atoms a smarter long-term strategy than gripping variable side chains?
6. How did having the 1HSG structure in hand speed the design of an inhibitor compared with blind screening?

A protein’s amino-acid sequence is just a string of letters — so how does a computer fold it into a working 3D machine, and how much should you trust the answer?

structural predictiondeep learningconfidence

You’ll learn to: interpret a pLDDT confidence map from the B-factor column, locate high- and low-confidence regions, compare a predicted fold to an experimental structure, and judge what an AlphaFold model can and cannot tell you.

Background 1 — Anfinsen and the folding problem

In the 1960s Christian Anfinsen showed that a denatured enzyme refolds spontaneously into its active shape — proving that the amino-acid sequence alone encodes the native 3D structure. This is Anfinsen’s dogma. Yet knowing that the answer is hidden in the sequence did not mean we could compute it. A chain can adopt astronomically many conformations (Levinthal’s paradox), and physics-based simulation was far too slow. For five decades, the protein-folding problem — predicting structure from sequence — remained one of biology’s hardest open challenges, tested every two years at the blind CASP competition.

Background 2 — AlphaFold, CASP14 and pLDDT

At CASP14 in 2020, DeepMind’s AlphaFold achieved near-experimental accuracy, effectively solving the problem. Its deep-learning network reads patterns of co-evolution across related sequences to infer which residues sit close in space. Crucially, each model carries a per-residue confidence score, pLDDT (0–100), stored in the B-factor column. In MoleBench, Color by → B-factor therefore paints confidence: blue ≈ high (>90) in well-folded cores, red/orange ≈ low in flexible loops, chain termini and intrinsically disordered regions. Low pLDDT often flags genuine floppiness, not just model error.

Background 3 — Reading the pLDDT scale and its limits

The pLDDT scale has four agreed bands: >90 very high (backbone and most side chains reliable), 70–90 confident (backbone reliable), 50–70 low (treat with caution) and <50 very low (often a sign of intrinsic disorder, not just error). Because pLDDT lives in the B-factor slot, a blue→red ramp reads as high→low confidence on an AlphaFold model — the opposite meaning of B-factor on an X-ray structure, where red marks mobile atoms. Remember the score is the network’s self-assessment of local accuracy; it does not certify the relative orientation of distant domains, which needs the separate PAE map.

Structures — open in the workbench ↗

Procedure

1. Open the AlphaFold model P69905 (human haemoglobin α-subunit, a prediction). In the Structure (i) panel, note the method reads as a computed/predicted model, with one chain.
2. Set Representation (1) → Show as → cartoon so you can see the fold of α-helices.
3. In Representation (1) → Color by, choose B-factor. Because this is an AlphaFold model, the colours now read pLDDT confidence: blue = confident, red/orange = uncertain.
4. Use ⟲ fit to frame the whole chain. Look at the helix cores versus the N- and C-termini and tight loops — spin with ◌ to inspect them.
5. Open the Sequence (5) panel. Click a residue near a terminus, then one in the middle of a long helix, and watch the viewer zoom to each — relate its colour to its sequence position.
6. For each region in the worksheet, record the dominant colour (blue/green/orange/red) you see under Color by → B-factor, then translate it into a confidence band (high/med/low) using Background 3.
7. Now open the experimental 4HHB (X-ray hemoglobin). In Chains (2), isolate a single α-chain (A or C) so you compare like with like.
8. Set 4HHB to cartoon and Color by → structure, then visually compare its helix layout to the P69905 prediction and complete the final worksheet row.

Questions

1. What does a pLDDT value of 95 versus 45 tell you about a residue? Which colour is which under Color by → B-factor, and which confidence band does each fall in?
2. From your worksheet, which region scored the lowest confidence? Give two physical reasons why those regions score poorly.
3. Why does a red region mean the opposite thing on this AlphaFold model than it would on the experimental 4HHB structure coloured by B-factor?
4. How does the single predicted α-chain compare to one isolated α-subunit of experimental 4HHB — same overall fold, or different? Cite your final worksheet row.
5. Why is fast, accurate structure prediction so valuable for biology and medicine? Give a concrete example.
6. Name two things an AlphaFold model does not directly give you that the experimental 4HHB structure does.

Two strands, four letters, one of the most famous shapes in science — what holds the DNA double helix together, and why does its sequence change how easily it comes apart?

nucleic acidsbase pairingB-form DNA

You’ll learn to: identify the two antiparallel strands and the sugar-phosphate backbone, measure Watson–Crick hydrogen-bond distances, distinguish the major and minor grooves, and reason about how GC content sets DNA stability.

Background 1 — Watson–Crick base pairing

DNA stores information as a sequence of four bases: adenine (A), thymine (T), guanine (G) and cytosine (C). They pair with strict complementarity by hydrogen bonds: A–T forms 2 H-bonds and G–C forms 3 H-bonds. Each pair links a purine to a pyrimidine, giving every rung the same width so the helix stays uniform. Because G–C has an extra hydrogen bond (and stronger base-stacking), GC-rich DNA is more thermally stable than AT-rich DNA. This complementary pairing is what lets each strand serve as a template to rebuild the other during replication.

Background 2 — The antiparallel B-form helix

The two strands are antiparallel: one runs 5′→3′ while its partner runs 3′→5′ alongside it. The outside is a sugar-phosphate backbone (deoxyribose linked by phosphates), and the bases point inward. The Dickerson dodecamer (1BNA) is the textbook B-form: a right-handed helix with about 10 base pairs per turn, a rise of ~3.4 Å per pair and a twist near 34°. Because the two backbones are not diametrically opposite, the helix surface has a wide major groove and a narrow minor groove — the channels DNA-binding proteins reach into to read the sequence.

Background 3 — Geometry, grooves and melting

The regular geometry is what makes DNA both stable and readable. With ~10 bp per turn and a 34° twist between adjacent rungs, the bases stack like coins, and overlapping π-systems supply the base-stacking energy that — together with H-bonds — holds the duplex shut. The 3.4 Å rise sets the helix pitch at ~34 Å per turn. Heating breaks H-bonds and unstacks the bases; the midpoint of this denaturation is the melting temperature (T_m). Because G–C rungs carry an extra bond and stack better, GC content raises T_m — the rule behind PCR primer design and probe selection.

Structures — open in the workbench ↗

Procedure

1. Open 1BNA (the Dickerson dodecamer). In the Structure (i) panel, note the resolution and that it contains 2 chains — in Chains (2) you should see both strands listed.
2. Set Representation (1) → Show as → sticks so every atom of the backbone and bases is visible.
3. Choose Color by → chain to give each strand its own colour — trace one backbone and confirm the two strands run in opposite (antiparallel) directions.
4. Switch Color by → element to pick out the N (blue), O (red) and backbone P (phosphorus) atoms; the H-bonds of a base pair run between N and O atoms.
5. Open Sequence (5), click a G on one strand to zoom to it, and locate its paired C on the opposite strand; then find an A–T pair the same way.
6. Use 📏 measure (click 2 atoms → distance): on the A–T pair, click an N or O donor on one base, then its partner N/O across the rung — a Watson–Crick H-bond reads ~2.8–3.0 Å. Record it in the worksheet, then repeat on the G–C pair.
7. With 📏 measure, click an atom on one base pair and the equivalent atom on the next pair up the helix to estimate the rise (~3.4 Å); record it.
8. Count the rungs through one full turn of the helix to confirm ~10 base pairs per turn, and enter it.
9. Use ◌ spin (and ⟲ fit) to rotate the helix and identify the wide major groove and the narrow minor groove. Save the best angle in Saved views (6).

Questions

1. Why is a G–C pair more stable than an A–T pair? Refer to the H-bond counts pre-filled in your worksheet.
2. What does antiparallel mean, and how did Color by → chain help you confirm it?
3. What donor–acceptor distance did you record for the A–T and G–C N–N / N–O contacts, and what does that distance represent physically (versus a covalent bond)?
4. The rise and base-pairs-per-turn you measured (~3.4 Å and ~10) identify which DNA form, and what helical pitch (Å per turn) do they imply?
5. How can you tell the major groove from the minor groove, and why do many DNA-binding proteins prefer the major groove?
6. Predict: two DNA samples of equal length, one ~70% GC and one ~30% GC. Which has the higher melting temperature (T_m), and why?