Graphene sheets and nanotube structure

A sheet and a cylinder share a network motif

Graphene is a one-atom-thick carbon sheet with a hexagonal bonding network. A carbon nanotube has a cylindrical related network. Tubes can have one wall or multiple concentric walls; diameter, arrangement and defects affect response.

Two carbon specimens can look like similarly sized cylinders yet have different atomic arrangements. Start with the bonded surface. In an ideal graphene interior, each carbon has three nearest carbon neighbours. Three directional σ bonds connect that network; the remaining p orbitals contribute to extended π-electron states. An orbital describes an allowed atomic electronic state and its spatial distribution. Here σ (sigma) bonding involves overlap along the atom-to-atom directions in the sheet; the remaining p orbitals have lobes above and below the sheet, whose sideways overlap forms the π (pi) states. The hexagons are not separate six-atom molecules. Edges, missing atoms and other defects need additional description.

A single-wall nanotube is one continuous cylindrical network, not a solid carbon rod. Describing it as a rolled graphene sheet is a geometric construction: it identifies which sheet sites become the same position around a closed circumference. It is not a claim that a growing nanotube was manufactured by rolling a loose sheet, or that the ideal wall contains a glued seam. Curving the network also changes bond directions and orbital overlaps.

Choose which sheet repeat becomes the circumference

Mark two repeat vectors a₁ and a₂ that connect equivalent sites in the sheet. They have equal length a and are 60° apart. A vector has both a length and a direction; a₁ and a₂ are not nearest-neighbour bond vectors. Use the stipulated ideal lattice value a = 0.246 nm throughout the calculations here. Its nearest-neighbour bond length is a/√3 ≈ 0.142 nm. These rounded reference values describe the geometric model, not every strained or curved specimen.

Choose integers n and m and form the wrapping vector C = n a₁ + m a₂. Identify its endpoints when forming the cylinder. The length |C| becomes the circumference, so the ideal diameter is d = |C|/π. The tube axis is perpendicular to C in the original sheet. The index pair describes atomic wrapping, not the number of walls, a tube’s length or the number of tubes in a bundle.

For a₁ along the sheet’s x direction and a₂ inclined at 60°, the components are Cₓ = a(n + m/2) and Cᵧ = a√3 m/2. Pythagoras then gives |C| = a√(n² + nm + m²). The plus nm term comes from the 60° angle: treating the two repeat directions as perpendicular would give the wrong circumference.

The chiral angle θ is the angle from a₁ to C in this convention, with tan θ = √3 m/(2n + m). Choose n ≥ m ≥ 0 with n > 0 for the examples. Then θ is between 0° and 30°. The names zigzag, armchair and chiral describe how the network runs around the cylinder: m = 0 gives zigzag, n = m gives armchair, and 0 < m < n gives a chiral tube. A chiral arrangement cannot be superposed on its mirror image using only translations and rotations. Armchair and zigzag are achiral special cases. Naming every tube’s index pair “chirality” does not mean every tube is structurally handed.

A closed circumference also constrains electronic states

An electronic wave on a closed tube must be consistent after one trip around the circumference. That boundary condition selects allowed circumferential states from those of the sheet. In a simple ideal π-electron model of a long, isolated, unstrained tube, neglecting curvature corrections and defects, the resulting bands meet with zero gap when n − m is divisible by three; other index pairs have a finite gap. A band gap is an energy interval separating the relevant valence and conduction states. The classification is a model result, not a measurement of a sample’s resistance.

This model explains why “all carbon” and “almost the same diameter” are insufficient electronic descriptions. The (10,0) and (6,6) examples have diameters differing by only about 3.9%, but the simple model puts them in different gap classes. Conversely, scaling both indices can preserve the wrapping angle while changing the circumference. An angle is not a size.

The zero-gap label needs care. Curvature can open small gaps in nominally metallic non-armchair tubes, and distortions or neighbouring tubes can alter electronic states. Atomically resolved spectroscopy has found small gaps in particular nominally metallic zigzag tubes and changes when armchair tubes are bundled. Those observations establish a limit of transferring the isolated ideal classification; they do not give a universal gap or conductivity for a new specimen.

Contacts, carrier population, temperature, length, disorder and the measurement arrangement also affect electrical response. A finite gap does not mean no current can ever flow, and a zero-gap model does not promise zero resistance. Charge carriers and electronic energy bands provides the preparation for interpreting states and carriers; this lesson uses only the structural distinction.

Geometry changes electronic and mechanical questions

The way a sheet wraps into a tube is described by its chirality. Electronic character depends on structure, and curvature and multiple walls add interactions. A graphene result is not automatically a measurement of every tube or bundle.

A multiwall tube contains nested cylindrical networks with different circumferences. Each wall has its own structure; there is no single index pair that describes every wall. A bundle instead contains several separate tubes in contact. Contacts between walls or between tubes are not simply the same in-plane covalent bonds continued through a thicker solid. They can transfer force and electronic interactions, but the amount depends on the actual interfaces and constraints.

The schematic distinguishes a sheet network from concentric wall cross-sections. It does not specify wrapping direction, lattice spacings, wall-to-wall separation or a mechanical test. Use it to identify the structural level; use the labelled construction to answer the wrapping question.

Count walls, then ask how the load reaches them

More walls provide more possible load paths, but wall count alone does not establish that they carry the same extension or force. End attachments can grip only an outer wall, or may engage several walls. Interwall interactions or crosslinks can transfer load; sliding and incomplete engagement can limit it. Experiments and simulations of particular multiwall specimens show that changing interwall coupling can change load sharing. This does not mean every added crosslink improves every property.

Use a separate invented small-extension model to isolate the boundary condition. Two equal-length walls have supplied axial stiffnesses kₒ = 80 N/m and kᵢ = 50 N/m. Stiffness here means force divided by axial end-to-end extension in a linear, quasistatic interval; it is not a modulus or a failure load. Measure extensions from an initially unloaded state, without pre-tension or other axial forces. Ignore changes of geometry and stiffness, and take interwall axial force transfer as zero in both cases. No elastic constants or attachment performance of a real tube are claimed.

If both walls are directly attached at both ends to the same ideal rigid plates, they share an extension δ. Their forces add: F = (kₒ + kᵢ)δ. At δ = 0.20 nm = 2.0 × 10⁻¹⁰ m, the outer wall carries 16 nN and the inner 10 nN, giving 26 nN. Here the common extension is imposed by the plates, not supplied by an assumed strong interwall interface.

If the plates grip only the outer wall while the inner wall has free ends and no interwall transfer, the inner wall carries no axial force in this model. At the same outer extension, F = kₒδ = 16 nN. The inner wall still exists. The change from 26 to 16 nN concerns which paths are engaged, not a change of carbon chemistry or a measured loss of tensile strength. Real partial-transfer cases need a suitable interface/contact model and observations; these two ideal boundaries do not identify them.

Keep structure and the reported response together

A useful structural description separates sheet or tube, single or multiple walls, each relevant diameter and wrapping assignment, length, defects and whether the object is isolated or assembled. Keep the supports, contacts and loading or electrical measurement direction as well. A microscopy outline alone may identify a tube’s dimensions without resolving its atomic index pair. All-carbon composition supplies still less structural information.

For a mechanical comparison, state whether the quantity is a measured force, component stiffness, nominal stress under an area convention or a fracture endpoint. The two-wall example needs no arbitrary wall thickness because its given quantities are force and extension. Converting a real tube result to stress introduces a separate choice about the loaded wall area. Cross-sectional area, gauge length and test conditions develops that issue; the wrapping formula does not settle it.

A sheet result, an individual tube result and an assembled bundle result are three different evidence levels. A bundle adds tube lengths, orientation, contacts and sliding or other load-transfer paths. Preserve the tested description and mechanism before asking whether a property can be transferred. Matching an elemental composition or one geometric dimension is not that evidence.

Worked examples

Try the reasoning

1. Does all-carbon composition identify nanotube chirality?

   Show the answer and reasoning

   No.

   Atomic arrangement around the cylinder is additional information.

2. Is a bundle one perfect enlarged tube?

   Show the answer and reasoning

   No.

   It adds interfaces and load-transfer pathways.

3. Replace (8,4) by (16,8) at the same ideal a. What happens to circumference, diameter, wrapping angle and the simple gap class?

   Show the answer and reasoning

   Circumference and diameter double to 5.20684 nm and 1.65739 nm. The angle stays 19.1066°. Both examples have a finite gap in the stated simple classification.

   Both vector components double, leaving their ratio unchanged. The first n − m is 4 and the second 8; neither is divisible by three. Equal angle does not imply equal diameter or equal numerical band gap.

4. Compare the ideal pairs (9,1) and (6,5). Calculate n² + nm + m² for each. Can an exact ideal diameter by itself distinguish their wrapping angles?

   Show the answer and reasoning

   Both sums are 91, giving d = 0.746975 nm. Their angles are about 5.2087° and 26.9955°, so diameter alone does not distinguish them.

   For (9,1), the sum is 81 + 9 + 1; for (6,5), it is 36 + 30 + 25. The directions differ even though the lengths agree. Both have finite gap in the simple rule, but that does not establish identical full bands or measured transport.

5. A calculation replaces the repeat length a = 0.246 nm by the nearest-neighbour bond length 0.142 nm in d = a√(n² + nm + m²)/π. For (8,4), what error results?

   Show the answer and reasoning

   It gives about 0.47835 nm instead of 0.828694 nm, approximately 42.3% too small.

   The indices count a₁/a₂ repeat vectors, not bond vectors. The mistaken result is scaled by 0.142/0.246 ≈ 0.57724. Units alone would not catch this error because both input lengths are in nm.

6. In the same no-transfer linear model, keep kₒ = 80 N/m but use kᵢ = 20 N/m and δ = 0.30 nm. Compare the two gripping conditions.

   Show the answer and reasoning

   Both-wall gripping gives 24 nN outer plus 6 nN inner, total 30 nN. Outer-only gripping gives 24 nN.

   The changed SI extension is 3.0 × 10⁻¹⁰ m. Direct plate engagement determines whether the inner stiffness contributes; neither force is a fracture prediction.

7. A theoretical model cuts the closed tube along its axis and opens it into a strip while preserving its other local bonds. Can it automatically retain the closed-tube circumferential-state rule?

   Show the answer and reasoning

   No. The periodic closed circumference has been replaced by edges, so the electronic boundary problem has changed.

   Keeping carbon composition and many local bonds is insufficient to keep the same allowed states. Edge structure and the new geometry need their own model; the original (n,m) gap rule is not a rule for every carbon strip.

8. Rotating an intact isolated tube in the laboratory changes the direction of its axis relative to the apparatus. Does this rotate its lattice into a new (n,m) tube?

   Show the answer and reasoning

   No. A rigid rotation changes the external orientation, not which lattice repeat closes around the tube.

   The measurement or loading geometry may change, but the wrapping indices are structural information internal to the tube. A bond rearrangement or a changed specimen would be a different situation.

What to carry forward

• Specify sheet, wall structure and tube geometry.
• Retain the tested structural level.
• The wrapping vector is a non-perpendicular lattice sum around the circumference; it is perpendicular to the tube axis.
• Diameter, wrapping angle, wall count and contacts are distinct information.
• Use an electronic classification only within its stated model, and distinguish a gap class from measured resistance.
• More visible walls do not by themselves establish equal load sharing or a fracture result.