The Electric Dipole & Dipole Moment

Tear a piece of paper into flakes, rub a plastic scale through your hair, and hold it close: the flakes leap up and stick. Stop and look at what just happened. The paper was neutral — equal positive and negative charge, net zero — yet it felt an electric force strong enough to beat gravity. Neutral does not mean electrically boring, and the concept that explains why is the electric dipole: two equal and opposite charges held a small distance apart. Nature manufactures dipoles by the mole. A water molecule is one — its oxygen end slightly negative, its hydrogen end slightly positive — and that tiny built-in lopsidedness is why water dissolves salt, why a charged comb bends a thin stream from the tap, and why a microwave oven can cook dinner by grabbing water molecules and twisting them billions of times a second. Your board exam cares for a more immediate reason: the dipole is the second character of the electrostatics chapter, and CBSE, ISC, JEE and NEET all test the same short list — the dipole moment $\vec{p}$ , its magnitude and direction, its units, and which molecules have one. This article builds that foundation properly: what a dipole is, why the single vector $\vec{p} = q\vec{d}$ captures everything about it, the notation trap between $d$ and $2a$ that quietly costs a mark every year, and how chemists read molecular shapes straight off measured dipole moments. The dipole's field pattern and its behaviour inside an external field get their own modules next — this one makes sure the object itself is solid.

Key takeaways

An electric dipole is a pair of equal and opposite charges $+q$ and $-q$ separated by a small fixed distance.
Its strength and orientation are packed into one vector, the electric dipole moment: magnitude = charge × separation, direction from $-q$ to $+q$ . Unit: coulomb-metre (C·m).
Notation trap: if the charges sit a distance $d$ apart, $p = qd$ ; NCERT writes the separation as $2a$ , so $p = q \times 2a$ — the separation is always the full distance between the charges.
For a neutral pair the dipole moment is a genuine property of the system — it does not depend on where you put your origin.
Polar molecules (H₂O, HCl, NH₃) have permanent dipole moments of order $10^{-30}$ C·m, measured in debye (1 D $= 3.336\times10^{-30}$ C·m); symmetric molecules (CO₂, CH₄) have zero.
Key idea: a dipole is how a neutral object still speaks electrically — one vector $\vec{p}$ , from minus to plus, says how loudly.

The core insight

An electric dipole — charges $+q$ and $-q$ a distance $d$ apart — is completely characterised by its dipole moment $\vec{p} = q\vec{d}$ , a vector of magnitude (charge × full separation) pointing from the negative charge to the positive charge.

Where this lives today

The dipole moment is quietly one of the most measured quantities in science. Every microwave oven is a $2.45$ -GHz electric field grabbing the $1.85$ -D handle on each water molecule and twisting — the friction of a hundred trillion trillion tiny dipoles being wrenched back and forth is what heats your food. (Contrary to the popular factoid, that frequency is not a "resonance of water" — it is deliberately set below water's absorption peak so the waves penetrate centimetres deep instead of scorching the surface.) In pharmaceutical and materials labs, computed and measured dipole moments screen molecules by the thousand: a drug candidate's moment helps predict solubility (polar dissolves polar — water's dipole is why it is chemistry's default solvent), and solar-cell designers tune molecular dipoles to steer charge across interfaces. At the coldest frontier, physicists laser-cool polar molecules like CaF (moment near 3 D) to microkelvin temperatures precisely because the dipole moment gives an electric-field handle for steering them — candidate hardware for quantum computers built on tiny, permanent versions of the two-charge object defined in this module.

The physical picture

Take a charge $+q$ and a charge $-q$ and glue them to the ends of an imaginary rod of length $d$ . From very far away, the pair looks like nothing: $+q + (-q) = 0$ , so a distant observer measuring the total charge gets zero. But "far away" is never infinitely far, and the two charges are never in exactly the same place. Whichever side of the pair you stand on, one charge is slightly closer to you than the other, so its influence wins by a whisker. That whisker of imbalance — a leftover that never quite cancels — is the dipole's entire electrical personality. How should we measure the size of the leftover? Two knobs control it. Crank up the charge $q$ and each end shouts louder, so the imbalance grows. Pull the charges farther apart (bigger $d$ ) and the difference between "near charge" and "far charge" grows, so again the imbalance grows. The natural strength-meter multiplies the two knobs together, and because the imbalance is lopsided along a direction — the line from minus to plus — the meter must be a vector. That vector is the dipole moment.

Figure

Anatomy of an electric dipole — two charges, one vector

Keep one mental contrast ready for exams. A point charge is a monopole: its total charge $q$ is the headline, and everything it does scales with $q$ . A dipole has a zero headline — the interesting information starts one level deeper, at how the charge is arranged. That is why the dipole gets its own vector, its own field pattern (next module), and its own behaviour in external fields (the module after). Whenever a question says "a system has zero net charge but experiences/produces an electric effect," your first thought should be: dipole.

The dipole moment: one vector to rule the pair

Everything measurable about an ideal dipole — the field it creates, the torque it feels, the energy it stores in an external field — turns out to depend on $q$ and $d$ only through their product. Physics is telling us what the right currency is:

\vec{p} = q\,\vec{d} \qquad (1)

where: $\vec{p}$ — the electric dipole moment, in coulomb-metres (C·m); $q$ — the magnitude of either charge (a positive number); $\vec{d}$ — the displacement vector from the negative charge to the positive charge, whose length is the full separation between them. Read the direction rule twice: $\vec{p}$ points from $-q$ toward $+q$ . It is a definition, not something you derive, and examiners test it as a one-marker with reversed options waiting. The magnitude is simply $p = qd$ . If the charges are $\pm 5.0$ nC and they sit $2.0$ mm apart, then $p = (5.0\times10^{-9})(2.0\times10^{-3}) = 1.0\times10^{-11}$ C·m — small numbers multiplied by small numbers, which is why real dipole moments live far down the powers of ten. Now the notation trap, the single most common mark-loser on this topic. NCERT (and every paper set from it — CBSE, JEE, NEET) draws the dipole with charges at $-a$ and $+a$ , so the separation is $2a$ , and writes

p = q \times 2a \qquad (2)

where: $2a$ — the full minus-to-plus distance; $a$ — only half of it, the distance of each charge from the centre. Equations (1) and (2) are the same statement in two costumes: $d = 2a$ . The error pattern is grabbing whatever letter the question offers and multiplying blindly — writing $p = qa$ when the question gave the half-separation, or $p = q(2d)$ when it gave the full one. Anchor the physics, not the letter: dipole moment = charge × (full distance between the charges), always. Is $\vec{p}$ a property of the dipole itself, or does it secretly depend on where we set up our coordinates? For a neutral pair, it is genuinely intrinsic — and the proof is two lines of algebra worth owning.

Derivation AWhy the dipole moment does not care where your origin is

We want to show that for a system with zero net charge, the dipole moment defined charge-by-charge comes out the same no matter which point of the room you call the origin — so $\vec{p}$ is a real physical property of the dipole, as objective as its mass. The strategy: write the general recipe for $\vec{p}$ (each charge times its position vector, summed), apply it to $+q$ and $-q$ at arbitrary positions, and watch the origin's contribution cancel because the charges are equal and opposite. Watch that cancellation — it is exactly where neutrality earns its keep.

Step 1. The general recipe. For any set of charges $q_1, q_2, \ldots$ at position vectors $\vec{r}_1, \vec{r}_2, \ldots$ measured from some chosen origin, the dipole moment is the charge-weighted sum of positions: $\vec{p} = q_1\vec{r}_1 + q_2\vec{r}_2 + \cdots$ (Each term is "how much charge" times "where it sits" — a lopsidedness meter for the whole arrangement.)

Step 2. Apply it to our pair: charge $+q$ at $\vec{r}_+$ and charge $-q$ at $\vec{r}_-$ . Then

\vec{p} = (+q)\,\vec{r}_+ + (-q)\,\vec{r}_- = q\,(\vec{r}_+ - \vec{r}_-)

Why this is progress: the difference $\vec{r}_+ - \vec{r}_-$ is the arrow from the negative charge to the positive charge — the origin has vanished from the answer before we even asked it to.

Step 3. Name that arrow: $\vec{r}_+ - \vec{r}_- = \vec{d}$ , the minus-to-plus displacement. So $\vec{p} = q\vec{d}$ — definition (1) recovered from the general recipe, with the direction convention emerging automatically rather than by decree.

Step 4. Now move the origin deliberately, by any shift $\vec{s}$ : every position vector changes, $\vec{r}_+ \to \vec{r}_+ - \vec{s}$ and $\vec{r}_- \to \vec{r}_- - \vec{s}$ . Recompute:

\vec{p}\,' = q\,\big[(\vec{r}_+ - \vec{s}) - (\vec{r}_- - \vec{s})\big] = q\,(\vec{r}_+ - \vec{r}_-) = \vec{p}

Why it cancelled: the shift $\vec{s}$ entered once with $+q$ and once with $-q$ ; equal and opposite charges paid it in and took it back out. If the charges did NOT balance — say $+2q$ and $-q$ — the leftover $-q\vec{s}$ would survive, and the "dipole moment" would change with your choice of origin. Neutrality is precisely what makes $\vec{p}$ objective. *Verified: for $+q$ at $(a, 0)$ and $-q$ at $(-a, 0)$ , Step 2 gives $\vec{p} = q(2a)\hat{x}$ — magnitude $q \times 2a$ along minus-to-plus, matching equation (2); shifting the origin to $(5, 0)$ changes both position vectors but Step 4's difference still returns $2a\hat{x}$ .*

The Step-1 recipe is also your tool for systems of several charges — JEE occasionally plants three or four charges and asks for the net dipole moment. Treat each pair-up honestly: multiply each charge by its position vector, add vectorially. And because dipole moments are vectors, dipoles inside a molecule combine by vector addition — which is the key that unlocks the chemistry connection below.

Units, the debye, and the molecular scale

The SI unit of $p$ is the coulomb-metre. It is an honest unit and an awkwardly enormous one: a full coulomb is already a monstrous charge, and a metre of separation is laboratory-absurd. Real molecular dipoles involve fractions of the electron charge across fractions of a nanometre, so their moments sit near $10^{-30}$ C·m. Rather than haul that exponent through every table, chemists use a tailored unit, the debye (D):

1\ \text{D} = 3.336\times10^{-30}\ \text{C}\,\cdot\,\text{m} \qquad (3)

where: the odd conversion factor is historical (the unit predates SI), but the scale is the useful part — one debye is a typical molecular dipole moment, so molecules score pleasant numbers like 0.1 to 10 D. A helpful anchor: an electron charge displaced by one ångström ( $10^{-10}$ m, about one atomic radius) gives $p = (1.6\times10^{-19})(10^{-10}) = 1.6\times10^{-29}$ C·m $= 4.80$ D. Any molecular moment you meet is some fraction of that, which tells you immediately that molecular charge separations are partial electron transfers over sub-atomic-spacing distances. For board exams you compute in C·m; the debye appears in reasoning questions ("which molecule has the larger moment and why") and in chemistry crossover questions. Know the conversion exists, know water's value, and you are covered.

Polar and non-polar: molecules as ready-made dipoles

Here is where the dipole stops being a textbook toy. Molecules are neutral, but the electrons inside them are not shared evenly: oxygen pulls electron cloud away from hydrogen, chlorine hogs it from hydrogen, and the result is a molecule with a slightly negative end and a slightly positive end — a permanent, built-in dipole. Such molecules are called polar. Molecules whose charge centres coincide — by symmetry or by even sharing — have $p = 0$ and are non-polar. The gallery of standard examples, with measured moments:

molecule	shape	dipole moment	verdict
H₂O	bent (104.5°)	$1.85$ D $= 6.2\times10^{-30}$ C·m	polar — the exam's favourite
HCl	linear, lopsided	$1.08$ D	polar
NH₃	pyramid	$1.47$ D	polar
CO₂	linear, symmetric	$0$	non-polar — bond dipoles cancel
CH₄	tetrahedral	$0$	non-polar — four bond dipoles cancel

The CO₂ row is the one to understand rather than memorise. Each C=O bond is polar — oxygen tugs charge from carbon — so the molecule contains two genuine bond dipoles. But the molecule is a straight line, O=C=O, with the two bond-dipole vectors pointing in exactly opposite directions with equal magnitudes. Vector sum: zero. The molecule is non-polar despite having polar bonds, purely by geometry. Water escapes this fate because it is bent: its two bond dipoles sit at $104.5°$ to each other, and vectors at an angle never fully cancel.

Figure

Water versus carbon dioxide — geometry decides polarity

Go deeper · Derivation B · 2 minThe dipole moment of a bent molecule — water by vector addition

Derivation BThe dipole moment of a bent molecule — water by vector addition

The goal is the net dipole moment of a bent molecule from its two bond dipoles — the calculation that explains why bent H₂O is strongly polar while straight CO₂ scores zero, and a live demonstration that dipole moments obey the vector addition you learned in mechanics. The strategy: place the two bond-dipole vectors at the known bond angle, resolve them along and across the molecule's axis of symmetry, and watch the across-components cancel while the along-components add. Keep an eye on the half-angle — using the full bond angle where the half-angle belongs is this calculation's classic slip.

Step 1. Set up the geometry. Water's two O–H bond dipoles, each of magnitude $p_b$ , point from each H toward the O region along the bonds (each bond dipole runs H → O in the charge sense; what matters here is that the two vectors are symmetric about the molecule's bisector). The angle between the two bond directions is $2\phi = 104.5°$ , so each bond makes angle $\phi = 52.25°$ with the bisector axis.

Step 2. Resolve each bond dipole into components. Along the bisector: $p_b\cos\phi$ each. Perpendicular to the bisector: $p_b\sin\phi$ , but the two bonds lean to opposite sides, so these components are equal and opposite. Why the perpendicular parts must cancel: the molecule is mirror-symmetric about its bisector, and a net sideways dipole would break that symmetry — the algebra is just the mirror talking.

Step 3. Add. The perpendicular components cancel; the parallel components reinforce:

p_{\text{net}} = 2\,p_b\cos\phi \qquad (4)

Step 4. Run it with real numbers, backwards — that is how it is actually used. Water's measured net moment is $p_{\text{net}} = 1.85$ D and $\phi = 52.25°$ , so each O–H bond carries $p_b = \dfrac{1.85}{2\cos 52.25°} = \dfrac{1.85}{2(0.612)} = 1.51$ D. Chemists really do infer bond dipoles this way.

Step 5. Read the two limits of equation (4). If the molecule straightens out, $2\phi \to 180°$ , then $\phi \to 90°$ , $\cos\phi \to 0$ , and $p_{\text{net}} \to 0$ — that is CO₂, the straight-line cancellation. If the two bonds fold onto each other, $\phi \to 0$ and $p_{\text{net}} \to 2p_b$ — full reinforcement. Bent molecules live between the limits, polar by geometry. Verified: $\cos 52.25° = 0.6122$ , so $2 \times 1.51 \times 0.6122 = 1.85$ D — the round trip closes; the CO₂ limit reproduces the table's zero.

Limiting cases

case	result	meaning
$d \to 0$ (charges merge)	$p \to 0$	overlapping + and − is just nothing, electrically
$q \to 0$	$p \to 0$	no charge, no dipole — both knobs must be nonzero
charges unequal ( $+2q, -q$ )	not a dipole; $\vec{p}$ origin-dependent	the "dipole" label needs $Q_{\text{net}} = 0$ ; net charge dominates from afar
linear symmetric molecule (CO₂)	bond dipoles cancel, $p = 0$	polar bonds can still make a non-polar molecule
bond angle $2\phi \to 180°$ in (4)	$p_{\text{net}} \to 0$	straightening a molecule kills its polarity
ideal (point) dipole: $d \to 0$ with $qd$ fixed	$p$ stays finite	the theorist's dipole — used for the field formulas next module

The last row deserves a sentence, because the next module leans on it. Physicists like to shrink the dipole — imagine $d$ getting tiny while $q$ grows so the product $qd$ stays fixed. The result, called an ideal or point dipole, keeps a finite $\vec{p}$ with no size at all. Real molecules are close enough to ideal whenever you observe them from distances much larger than their size — which, for a $0.1$ -nm molecule viewed from even a few nanometres, is nearly always.

Common trap“The dipole moment points from plus to minus, like the field lines”

It points the other way — from $-q$ to $+q$ — and this reversed-direction option appears on the answer key of some board or entrance paper virtually every year. The confusion is natural: electric field lines outside a dipole do run from the positive charge around to the negative one, so "plus to minus" feels right. But $\vec{p}$ is defined through the displacement $\vec{d}$ of the positive charge relative to the negative one, so it runs minus → plus. (The field between the charges runs plus → minus — opposite to $\vec{p}$ — which is exactly why reasoning from field lines misleads here.) Treat the direction as a hard convention: $\vec{p}$ : from tail $-q$ to head $+q$ . Chemistry adds one more twist: some chemistry textbooks draw the molecular-dipole arrow toward the negative end (the old "crossed arrow" convention). In physics exams, physics wins: minus to plus, always.

Worked examples

Board level — dipole moment of a charge pair (CBSE/ISC pattern)

Two point charges $+5.0\ \mu\text{C}$ and $-5.0\ \mu\text{C}$ are placed $4.0$ mm apart. Find the magnitude and direction of the electric dipole moment of the system. State the SI unit.

Step 1. Identify the two knobs. Charge magnitude $q = 5.0\times10^{-6}$ C. Separation: the full distance between the charges, $d = 4.0\times10^{-3}$ m. (If the question had said "each charge is 2.0 mm from the centre," you would first double it — that is the $2a$ trap.)

Step 2. Multiply: $p = qd = (5.0\times10^{-6})(4.0\times10^{-3}) = \boxed{2.0\times10^{-8}\ \text{C}\,\cdot\,\text{m}}$ .

Step 3. Direction: along the line joining the charges, from the $-5.0\ \mu$ C charge toward the $+5.0\ \mu$ C charge. Say it in words — the direction statement carries its own mark.

Step 4. Unit sentence for the mark scheme: the SI unit of dipole moment is the coulomb-metre (C·m); it is a vector quantity.

Competitive level — how ionic is HCl? (JEE-Main/NEET crossover pattern)

The HCl molecule has a measured dipole moment of $1.08$ D and a bond length of $1.27\ \text{Å}$ . If the molecule were a perfect ion pair — a full electron transferred from H to Cl — what dipole moment would it have? Hence find the "fractional ionic character" of the bond. (1 D $= 3.336\times10^{-30}$ C·m, $e = 1.6\times10^{-19}$ C.)

Step 1. Model the perfect ion pair as an ideal dipole: charge $e$ separated by the bond length. $p_{\text{ionic}} = e \times d = (1.6\times10^{-19})(1.27\times10^{-10}) = 2.03\times10^{-29}$ C·m.

Step 2. Convert the measured moment to SI: $p_{\text{meas}} = 1.08 \times 3.336\times10^{-30} = 3.60\times10^{-30}$ C·m.

Step 3. The fraction of "full transfer" actually realised:

\text{ionic character} = \frac{p_{\text{meas}}}{p_{\text{ionic}}} = \frac{3.60\times10^{-30}}{2.03\times10^{-29}} = \boxed{0.18 \approx 18\%}

Step 4. Read the physics off the number: the H–Cl bond behaves as if only about a fifth of an electron has moved across — a partial charge $\delta \approx 0.18e$ on each end. This is exactly the kind of multi-concept item entrance papers like: the dipole definition (this module), unit conversion, and a modelling judgement, all in ninety seconds. Sanity checks: $p_{\text{ionic}} = 6.1$ D is the ceiling, the measured $1.08$ D sits well under it, and a fraction must land between 0 and 1 — if yours doesn't, you inverted the ratio. Timing note: 60–90 seconds. The only decisions are (i) ionic ceiling $= e \times$ bond length and (ii) which number goes on top — measured over ideal. Convert to SI once, at the start, and the arithmetic is two multiplications and one division.

Exam traps

The $2a$ trap (the big one). NCERT's dipole has charges at $\pm a$ : separation $2a$ , moment $p = 2qa$ . Questions freely switch between giving the full separation and the half-separation. Before multiplying anything, write down: "full minus-to-plus distance $= \_\_$ ." One second of bookkeeping, one mark saved.

Direction reversal. $\vec{p}$ runs from $-q$ to $+q$ . Options stating "from positive to negative charge" are planted. See the warning box — and remember chemistry's crossed-arrow convention does not apply in physics papers.

"Neutral ⟹ no electric effect." False, and the examiner knows you half-believe it. A dipole is neutral yet produces a field and feels torque. If an MCQ hinges on whether a neutral object can experience electrostatic force, think dipole (and induced dipoles — the charged-comb-and-paper effect) before answering "no force."

Vector addition skipped. For systems of charges or molecular bond dipoles, moments add as vectors. In symmetric arrangements expect cancellations (CO₂, CH₄); at angles use components or the parallelogram law — cashing the same skills as force addition. Adding magnitudes blindly is the planted wrong option.

Unit slips. Dipole moment is C·m — not C/m, not C·m². In the debye conversion, multiply by $3.336\times10^{-30}$ going D → C·m (a debye is tiny in SI; if your converted number got bigger, you divided).

Calling any two charges "a dipole." The word demands equal and opposite. For $+2q$ and $-q$ , the net charge $+q$ dominates the distant field, and the "dipole moment" changes with your origin choice — the derivation showed why. If a question's charges don't cancel, do not reach for dipole formulas.

Constants discipline. Where a calculation needs $k = 1/4\pi\varepsilon_0$ : CBSE/ISC and NEET accept $9\times10^9$ N·m²/C² unless the paper states otherwise; JEE answer grids are built on the same value. Use $e = 1.6\times10^{-19}$ C. No dipole-moment formula in this module needs $g$ — if you find yourself writing one, you've wandered into the wrong chapter.

Origins · 1 minPeter Debye and how this was discovered

Peter Debye

1884–1966 · Maastricht, Zurich, Berlin, Cornell

Debye began as an electrical engineer and became the man who taught chemistry to read molecular shapes with a voltmeter. His 1912 insight: molecules can carry permanent dipole moments, and heat makes them tumble — so a substance's response to an electric field falls with temperature in a way you can measure and invert. Run the logic backwards and dielectric measurements hand you $p$ ; the value of $p$ then betrays the geometry. Water's large moment means bent; CO₂'s zero means linear — molecular architecture deduced decades before anyone could image a molecule. The unit of molecular dipole moment carries his name for exactly this program, which earned him the 1936 Nobel Prize in Chemistry. Pressed by the Nazi government while directing a Berlin institute, he left for an American lecture tour in 1940 and simply did not return, finishing his career at Cornell.

History

1785

— Coulomb establishes the inverse-square force law: the raw material from which all dipole field results are built.

1912

— Debye proposes that molecules carry permanent dipole moments and shows how thermal tumbling makes polarity measurable.

1929

— Debye's monograph "Polar Molecules" codifies the field; the debye unit fixes the molecular scale of charge separation.

1936

— Nobel Prize in Chemistry to Debye for reading molecular structure from dipole moments and diffraction.

1945

— Microwave heating of food is patented: an oscillating field gripping water's permanent dipole, industrialised.

Going further

Texts. NCERT Physics Class XII Part 1, Ch. 1 (Electric Charges and Fields, §1.11) — the CBSE/JEE canonical treatment and the source of the $2a$ notation · HC Verma, Concepts of Physics Vol 2, Ch. 29 · Halliday–Resnick–Walker, Ch. 22 · For the university-level story of why $\vec{p}$ rules the far field of any neutral object, see the deep dive below.

Provenance. Definitions follow NCERT Class XII §1.11 and standard references. A verifier should recompute: $p = (5.0\times10^{-6})(4.0\times10^{-3}) = 2.0\times10^{-8}$ C·m; water $6.17\times10^{-30}/3.336\times10^{-30} = 1.85$ D; electron-ångström $= 1.6\times10^{-29}$ C·m $= 4.80$ D; HCl: $p_{\text{ionic}} = 2.03\times10^{-29}$ C·m, $3.60\times10^{-30}/2.03\times10^{-29} = 0.177$ ; water bond dipole $1.85/(2\cos 52.25°) = 1.51$ D.

Prerequisites. The Electric Field · Adding Vectors: Graphical & Analytical. Leads to. Field of a Dipole: Axial & Equatorial · Dipole in a Uniform Field: Torque & Energy · Electric Dipole Moment (UG deep dive).

Why does the dipole moment multiply q and d instead of, say, adding them or using q times d squared?

Because everything a dipole does depends on exactly that product. When you compute the field of a dipole (next module) or the torque on it in an external field, the answers come out proportional to $qd$ — double the charge at half the separation and every measurable effect is unchanged. Physics is telling you that $q$ and $d$ are not separately visible from outside; only their product is. Adding them would be dimensionally illegal (coulombs plus metres), and $qd^2$ belongs to a different, weaker character (the quadrupole, which you'll meet at university). The definition looks arbitrary only until you watch $qd$ pop out of every calculation, at which point it becomes obvious.

Is a dipole's total charge really zero? Then how can it attract anything at all?

Total charge: genuinely zero. But force depends on where the charges are, not just their sum. Bring a dipole near a positive charge: the dipole's negative end is (say) closer, so its attraction beats the positive end's repulsion — net attraction. The cancellation of the two effects is only perfect at infinite distance; up close, the separation $d$ buys a real leftover force. That leftover is weaker and dies off faster with distance than a point charge's force ( $1/r^3$ for the field instead of $1/r^2$ — derived next module), but it is absolutely real: it is why neutral paper flakes jump onto a charged comb.

My chemistry textbook draws the dipole arrow pointing toward the negative end. Who is right?

Both, in their own kingdoms — it is purely a convention clash. Physics (and every physics board: CBSE, ISC, JEE, NEET) defines $\vec{p}$ from $-q$ to $+q$ . Traditional chemistry used a crossed arrow drawn from the positive end toward the negative end (the "plus tail" marking δ⁺); IUPAC has since recommended aligning with the physics convention, but older chemistry books and teachers still use the crossed arrow. In a physics exam, minus-to-plus is the only defensible answer. If you take both subjects, write the convention you are using once at the top of your answer — examiners respect the disambiguation.

Water's bond angle is 104.5° — why isn't it 90° or 109.5°, and does the exact angle matter for p?

The angle comes from chemistry: oxygen's four electron pairs (two bonds, two lone pairs) arrange roughly tetrahedrally ( $109.5°$ ), and the lone pairs squeeze the H–O–H angle down to $104.5°$ . For physics purposes the exact value matters only through $\cos\phi$ in $p_{\text{net}} = 2p_b\cos(52.25°)$ : a slightly wider angle would give a slightly smaller moment. What your exam actually needs: bent means the bond dipoles cannot cancel, so water is polar; the specific number is for showing off in the calculation, not for memorising beyond "about 104.5°."

If I move a dipole around or rotate it, does its dipole moment change?

Translating it: never — the derivation showed $\vec{p}$ depends only on the relative displacement of the charges, so carrying the dipole across the room leaves $\vec{p}$ untouched. Rotating it: the magnitude stays fixed but the direction rotates with the dipole, because $\vec{p}$ is glued to the minus-to-plus line. That direction change is exactly what an external field exploits to exert a torque and swing the dipole into alignment — the whole subject of the "dipole in a uniform field" module coming up. What can change $p$ 's magnitude is deforming the dipole itself: stretch a molecule and its moment grows.

Do I need the ideal (point) dipole for exams, or is it just theory decoration?

You need the idea, one line of it: an ideal dipole is the limit of $d$ shrinking to zero while $q$ grows to keep $p = qd$ fixed. Its payoff arrives in the next module, where the field formulas $2kp/r^3$ (axial) and $kp/r^3$ (equatorial) are exact for an ideal dipole but only far-distance approximations for a real one. Exam questions occasionally say "short dipole" or "point dipole" — that word is your licence to use the simple $1/r^3$ formulas without correction terms. No question will ask you to philosophise about the limit; they ask you to recognise when it applies.