2.1 Direction in Space

A microwave oven is a machine for grabbing molecules and shaking them, and it only works on molecules that give it something to grab. Water heats spectacularly; the nitrogen and oxygen of the air in the oven barely notice. The difference is not mass, not bond strength, not anything a single number captures: a water molecule points. Its electron density is shifted toward the oxygen, giving the molecule a positive end and a negative end — a dipole — and the oven’s oscillating field seizes that handle and swings it back and forth ten billion times a second. Nitrogen, perfectly symmetric, offers no handle at all. “Points toward the oxygen” is a statement no ordinary number can make. A temperature is 25 °C without any direction; a mass is 18 g pointing nowhere. To do chemistry on quantities that aim — dipoles, forces on ions, displacements of diffusing molecules, and above all bonds — we need a new kind of mathematical object, and this chapter builds it.

Magnitude and direction, welded together

Definition 2.1.1 (Vector). A vector is a quantity possessing both a magnitude and a direction in space. Two vectors are equal exactly when both agree — same length, same aim — regardless of where the vectors are drawn. Quantities fully described by a single number (temperature, mass, energy) are scalars by contrast. We write vectors in bold — \mathbf{v} — and draw them as arrows.

The clause “regardless of where drawn” deserves a pause, because it is a genuine modeling decision. An arrow of given length and aim represents the same vector whether its tail sits at the origin or across the room: a displacement “3 Å along +x” is one instruction, wherever it is executed. This is why a bond vector belongs to the bond and not to any coordinate origin — a fact that already bit us once, in the positions-versus-bond-vectors caution of Section 2.3, and that we now see was baked into the definition from the start.

To compute with vectors rather than draw them, fix a coordinate system and record how far the arrow extends along each axis: \mathbf{v} = (v_x, v_y, v_z), the components. An O–H bond vector might be (0.958, 0, 0) Å; the same physical bond described in a rotated coordinate system would carry different components — the vector is the physical arrow, the components are its shadow on a particular choice of axes, and Section 2.4 will make that distinction do real work.

Adding arrows

Why would anyone add two directions? Because nature does it constantly. Each O–H bond of water carries its own small dipole, and the molecule’s total dipole — the thing the microwave grabs — is the two bond contributions combined. A dust grain in solution takes one random diffusive step, then another; its net displacement is the two steps combined. The combination rule is forced by the displacement picture: to execute two displacements in succession, walk the first arrow, then walk the second starting from where the first ended. The net effect is the arrow from the original start to the final end — the tip-to-tail rule. In components the rule becomes arithmetic, and it is worth seeing that this is a small theorem rather than a notational reflex.

Theorem 2.1.2 (Vector addition is componentwise). If \mathbf{a} = (a_x, a_y, a_z) and \mathbf{b} = (b_x, b_y, b_z), the tip-to-tail combination is \mathbf{a} + \mathbf{b} = (a_x + b_x,\; a_y + b_y,\; a_z + b_z).

Proof. Walk \mathbf{a}: you move a_x along the x-axis, a_y along y, a_z along z. Now walk \mathbf{b} from there: a further b_x, b_y, b_z along the same axes. Motions along one axis simply accumulate, independently of the other axes — the eastward parts of a journey add up regardless of any northward wandering. Total motion along x: a_x + b_x; likewise for y and z. The net arrow therefore has exactly the claimed components.

Two corollaries come free. Since ordinary addition of each component commutes, \mathbf{a} + \mathbf{b} = \mathbf{b} + \mathbf{a} — geometrically, the two walking orders trace the two sides of a parallelogram to the same far corner. And scaling is equally componentwise: c\,\mathbf{v} = (c v_x, c v_y, c v_z) stretches the arrow by |c|, reversing it when c < 0. In particular -\mathbf{v} is \mathbf{v} exactly reversed, and \mathbf{v} + (-\mathbf{v}) = \mathbf{0}, the zero vector — the unique vector with no direction at all, whose special status was already flagged in §2.3’s cautions.

Example 2.1.3 (Two forces on one ion). In an electrophoresis cell, a chloride ion feels an electric force \mathbf{F}_1 = (3.2, -1.5, 0) pN and a viscous drag \mathbf{F}_2 = (-2.8, 0.9, 0) pN. Find the net force.

Setting it up. Forces on a single object combine by vector addition — nature runs the tip-to-tail rule with no help from us. Theorem 2.1.2 reduces the sum to arithmetic.

Solution. \mathbf{F}_1 + \mathbf{F}_2 = (3.2 - 2.8,\; -1.5 + 0.9,\; 0) = (0.4,\; -0.6,\; 0)\ \text{pN}.

Check. Units: pN + pN = pN ✓. Sign logic: the drag nearly cancels the electric push in x (0.4 is small compared to 3.2), which is exactly what drag does — a large residual would mean runaway acceleration. Magnitude sanity: each component of the sum lies between the sum and difference of the corresponding inputs ✓. A wrong but common answer, (6.0, -2.4, 0), comes from dropping the drag’s minus signs — the sign of a component is its direction along that axis.

The dipole of water, assembled

Now the section’s opening scene gets its calculation. Each O–H bond carries a bond dipole of magnitude \mu_b pointing from H toward O along the bond; the two bonds meet at water’s 104.5°. By symmetry the two components perpendicular to the bisector cancel, and the two components along the bisector reinforce: each bond contributes \mu_b \cos(\theta/2) along the bisector, so \mu_{\text{net}} = 2\,\mu_b \cos\!\left(\tfrac{\theta}{2}\right). Feeding in \mu_b = 1.52 D and \theta = 104.5°: \mu_{\text{net}} = 2 (1.52)(\cos 52.25°) = 2(1.52)(0.612) = 1.86 D — and the measured dipole moment of water is 1.85 D. The molecule performs Theorem 2.1.2 on itself, and the agreement to three figures is the theorem’s laboratory audit.

The formula also explains molecules other than water. Stretch the angle to 180° and \cos 90° = 0: the dipoles cancel exactly. That is CO₂ — two strongly polar C=O bonds, zero molecular dipole, a linear molecule invisible to the microwave for the same reason it is a greenhouse gas by a subtler mechanism. Close the angle toward 0° and the net dipole approaches 2\mu_b, the two arrows in lockstep. Geometry, not bond polarity alone, decides whether a molecule points.

Before touching the slider: water’s angle is 104.5° and its net dipole is 1.86 D. Predict the net dipole at \theta = 120° (ammonia-like planar geometry) — bigger or smaller than water’s? And at exactly what angle does the net dipole vanish?

The perpendicular components always cancel by symmetry; only the along-bisector parts survive, each worth \mu_b\cos(\theta/2). The net dipole dies only at \theta = 180° — molecular linearity, not weak bonds, is what silences CO₂. Notice the resultant’s direction never budges from the bisector while its magnitude runs from 2\mu_b down to zero: symmetry pins the direction, geometry tunes the size.

Remarks and cautions

Scalars and vectors do not mix. “Add 2 to the vector (1, 3)” is as meaningless as adding 5 mL to 30 °C. A scalar can multiply a vector (Theorem 2.1.2’s corollary); it cannot be added to one. If an expression sums a bold symbol with a plain one, a type error has already happened upstream.

The magnitude of a sum is not the sum of magnitudes. Two 1.52 D bond dipoles combined to 1.86 D, not 3.04 D — the angle stole the rest. In general |\mathbf{a} + \mathbf{b}| \le |\mathbf{a}| + |\mathbf{b}|, with equality only for perfectly parallel vectors (this triangle inequality is proved with the dot product in Exercise 9). Adding magnitudes as if they were masses is the most tempting wrong move in every dipole problem.

Distinguish position from displacement. “The hydrogen at (0.96, 0, 0)” names a position — origin-dependent bookkeeping. “The O→H bond vector (0.96, 0, 0)” names a displacement — physical and origin-free. The arithmetic is identical; the meaning is not, and §2.3’s bond-angle caution was exactly the price of confusing them.

Direction requires nonzero magnitude. The zero vector points nowhere; asking for its direction is undefined, and formulas dividing by a magnitude (unit vectors ahead in §2.2) must exclude it explicitly.

Summary

A vector welds magnitude to direction and is the native mathematical type of every chemical quantity that aims: bonds, dipoles, forces, displacements. Fixing axes turns a vector into components; Theorem 2.1.2 turns the physically mandated tip-to-tail combination rule into componentwise arithmetic, with commutativity and scalar multiples following at once. The payoff was quantitative chemistry from geometry alone: water’s measured 1.85 D dipole reconstructed as 2\mu_b\cos(\theta/2), and CO₂’s vanishing dipole exposed as the \theta = 180° limit of the same formula. What vectors cannot yet do is report their own length or compare two directions numerically — magnitude is next (§2.2), and the comparison machine, the dot product, is already waiting in §2.3.

Exercises

  1. Compute \mathbf{a} + \mathbf{b}, \mathbf{a} - \mathbf{b}, and 3\mathbf{a} - 2\mathbf{b} for \mathbf{a} = (2, -1, 4) and \mathbf{b} = (-3, 5, 1).

  2. A molecule diffusing in solution makes three successive displacements: (1.2, 0.5, -0.8), (-0.4, 1.1, 0.6), and (0.9, -1.3, 0.2) (µm). Find the net displacement, and explain in one sentence why the answer would be the same if the three steps happened in any other order.

  3. Which of these are vectors, which scalars? Bond energy; bond dipole; pH; the force on an ion in a field; the number of microstates W; the velocity of an effusing molecule. One-line justification each.

  4. Two forces act on a colloidal particle: (4.0, 2.0, 0) pN and one unknown \mathbf{F}. The particle is observed to feel zero net force. Find \mathbf{F}, and state the general rule your answer illustrates.

  5. Using bond dipole \mu_b = 1.52 D for O–H: what net dipole would “water” have if its angle were the tetrahedral 109.5°? Compare with the real 1.86 D and state the direction of the change in one sentence.

  6. H₂S has bond angle 92.1° and bond dipole 0.69 D. Predict its net dipole from the bisector formula and compare with the experimental 0.97 D. (The agreement is imperfect — the formula ignores the sulfur lone pairs’ contribution; say which direction the discrepancy runs.)

  7. ★ Boron trifluoride, BF₃, is planar with three equivalent B–F bonds at 120° to each other. Show by explicit components that the three bond dipoles sum to exactly zero. Hint: place one bond along +x and use \cos 120° = -\tfrac12, \sin 120° = \tfrac{\sqrt3}{2}.

  8. ★ Ammonia’s three N–H bonds (\mu_b = 1.31 D) make an angle of 67.9° with the molecular (C_3) axis. Argue from symmetry that the net dipole lies along the axis, then compute its magnitude 3\mu_b\cos(67.9°) and compare with the experimental 1.47 D.

  9. ★ Prove the triangle inequality |\mathbf{a} + \mathbf{b}| \le |\mathbf{a}| + |\mathbf{b}|: expand |\mathbf{a}+\mathbf{b}|^2 with the dot product (§2.3) and apply Cauchy–Schwarz. When exactly does equality hold, and what does that say about dipoles?

  10. ★★ A polymer chain is modeled as N bond vectors of equal length \ell, each pointing in an independent random direction. Explain why the average end-to-end vector is \mathbf{0}, yet the chain is almost never actually closed. (The resolution — averaging the squared length instead — is a §2.3 computation: \langle |\mathbf{R}|^2 \rangle = N\ell^2. Show this, using the vanishing of the average cross terms \langle \mathbf{b}_i \cdot \mathbf{b}_j \rangle = 0 for i \ne j.)

Practice until it sticks