Skip to content

Complement SMILES

Complement SMILES is AutoPoly's format for defining polymer sequences with precise positional control. It extends standard SMILES notation with wildcard atoms ([*]) that mark where each monomer connects to its neighbors.

Why Complement SMILES?

In a polymer chain, monomers behave differently depending on their position:

  1. First monomer — has a starting end group and one connection point
  2. Middle monomers — have two connection points (to both neighbors)
  3. Last monomer — has one connection point and an ending end group

Standard SMILES cannot distinguish these positions. Complement SMILES solves this with one rule:

Number of wildcards = number of connections

Position Wildcards Example (ethylene) Role
First 1 (right) "CC[*]" Chain start
Middle 2 (both sides) "[*]CC[*]" Interior units
Last 1 (left) "[*]CC" Chain end

The Three Position Types

First position (chain start)

CC[*]
│ │
│ └─ Wildcard: connects to the next monomer
└─── Free end group (methyl here)

Middle positions

[*]CC[*]
 │  │  │
 │  │  └─ Wildcard: connects to the next monomer
 │  └──── Carbon backbone
 └─────── Wildcard: connects to the previous monomer

Last position (chain end)

[*]CC
 │  │
 │  └─ Free end group (methyl here)
 └──── Wildcard: connects to the previous monomer

Building Sequences

A chain is an explicit Python list — one complement SMILES per monomer, in order. The degree of polymerization is always len(sequence).

Step 1: Define the middle variant

middle = "[*]CC[*]"   # ethylene, interior

Step 2: Derive first and last

Remove the appropriate wildcard:

first = "CC[*]"       # left wildcard removed
last  = "[*]CC"       # right wildcard removed

Step 3: Build the sequence

dop = 100
sequence = [first] + [middle] * (dop - 2) + [last]

Step 4: Create the polymer

from AutoPoly import Polymer

polymer = Polymer(
    chain_num=10,
    sequence=sequence,
    topology="linear",
    tacticity="atactic",
)
print(polymer.dop)   # 100

A reusable helper:

def uniform_polymer(first: str, middle: str, last: str, dop: int) -> list:
    """Create a uniform homopolymer sequence."""
    if dop == 1:
        return [middle]
    if dop == 2:
        return [first, last]
    return [first] + [middle] * (dop - 2) + [last]

sequence = uniform_polymer("CC[*]", "[*]CC[*]", "[*]CC", 100)

Common Patterns

Uniform homopolymer

# Polyethylene, DOP=50
sequence = ["CC[*]"] + ["[*]CC[*]"] * 48 + ["[*]CC"]

Block copolymer

Contiguous blocks — just mix monomers in the list:

# PE(10)-PS(20)-PE(10) ABA triblock, DOP=40
sequence = (
    ["CC[*]"] + ["[*]CC[*]"] * 9 +                  # PE block (10)
    ["[*]CC([*])c1ccccc1"] * 20 +                   # PS block (20)
    ["[*]CC[*]"] * 9 + ["[*]CC"]                    # PE block (10)
)

Alternating copolymer

# Ethylene-styrene alternating, DOP=6
sequence = [
    "CC[*]",                      # A (first)
    "[*]CC([*])c1ccccc1",         # B
    "[*]CC[*]",                   # A
    "[*]CC([*])c1ccccc1",         # B
    "[*]CC[*]",                   # A
    "[*]CC(c1ccccc1)",            # B (last)
]

Comment complex sequences

For long block sequences, annotate block boundaries with inline comments — the sequence is data, and future-you will read it.

Ring vs Linear Polymers

Linear polymers

Need all three variants (first, middle, last):

sequence_linear = ["CC[*]"] + ["[*]CC[*]"] * 48 + ["[*]CC"]

polymer = Polymer(chain_num=10, sequence=sequence_linear, topology="linear")

Ring polymers

Use only the middle variant — every position has two connections:

sequence_ring = ["[*]CC[*]"] * 50    # all middle; no first or last

polymer = Polymer(chain_num=10, sequence=sequence_ring, topology="ring")

Why? Ring polymers have no end groups — every monomer connects to two neighbors. Using a first or last variant in a ring will fail.

Common Monomers Reference

Commodity polymers

Polyethylene (PE)

first  = "CC[*]"
middle = "[*]CC[*]"
last   = "[*]CC"

Polypropylene (PP)

first  = "CC(C)[*]"
middle = "[*]CC([*])C"
last   = "[*]CC(C)"

Polystyrene (PS)

first  = "CC(c1ccccc1)[*]"
middle = "[*]CC([*])c1ccccc1"
last   = "[*]CC(c1ccccc1)"

Poly(vinyl chloride) (PVC)

first  = "CC(Cl)[*]"
middle = "[*]CC([*])Cl"
last   = "[*]CC(Cl)"

Engineering polymers

Poly(methyl methacrylate) (PMMA)

first  = "CC(C)(C(=O)OC)[*]"
middle = "[*]CC([*])(C)C(=O)OC"
last   = "[*]CC(C)(C(=O)OC)"

Polyacrylonitrile (PAN)

first  = "CC(C#N)[*]"
middle = "[*]CC([*])C#N"
last   = "[*]CC(C#N)"

Poly(ethylene oxide) (PEO)

first  = "COC[*]"
middle = "[*]COC[*]"
last   = "[*]COC"

Specialty polymers

Poly(tetrafluoroethylene) (PTFE, Teflon)

first  = "C(F)(F)C(F)(F)[*]"
middle = "[*]C(F)(F)C([*])(F)F"
last   = "[*]C(F)(F)C(F)(F)"

Polyisoprene (natural rubber)

first  = "CC(=C)C[*]"
middle = "[*]CC([*])=CC"
last   = "[*]CC(=C)C"

Poly(lactic acid) (PLA) — condensation polymer with an ester backbone

first  = "OC(C)C(=O)[*]"
middle = "[*]OC(C)C(=O)[*]"
last   = "[*]OC(C)C(=O)O"

Best Practices

1. Always check the wildcard count

✓ Correct:

first  = "CC[*]"        # 1 wildcard
middle = "[*]CC[*]"     # 2 wildcards
last   = "[*]CC"        # 1 wildcard

❌ Incorrect:

first  = "[*]CC[*]"     # 2 wildcards — that's a middle monomer
middle = "CC[*]"        # 1 wildcard — that's a first monomer
last   = "CC"           # 0 wildcards — no connection point at all

2. Verify the sequence length

sequence = [...]
expected_dop = 50
assert len(sequence) == expected_dop, f"Expected DOP {expected_dop}, got {len(sequence)}"

3. Validate SMILES before a big run

from rdkit import Chem

def valid(smiles: str) -> bool:
    return Chem.MolFromSmiles(smiles.replace("[*]", "C")) is not None

assert all(valid(s) for s in sequence)

Advanced Topics

Custom end groups

Change the non-wildcard end of the first/last monomers:

first_OH = "OCC[*]"     # hydroxyl start
last_OH  = "[*]CCO"     # hydroxyl end

first_Br = "BrCC[*]"    # bromine start
last_Br  = "[*]CCBr"    # bromine end

Explicit stereochemistry

Use SMILES chirality markers for tacticity control at the monomer level:

isotactic_middle     = "[*]C[C@@H]([*])C"   # R-configuration
syndiotactic_middle  = "[*]C[C@H]([*])C"    # S-configuration

For whole-chain tacticity, prefer the tacticity parameter of Polymer ("atactic", "isotactic", "syndiotactic").

Appendix: SMILES Basics

Complement SMILES builds on standard SMILES. The essentials:

  • Atoms — C aliphatic carbon, c aromatic carbon, O, N, halogens F Cl Br I
  • Bonds — single (omitted or -), double =, triple # (e.g. C=C, C#N)
  • Branches — parentheses: CC(C)C is isobutane
  • Rings — matching digits close a ring: C1CCCCC1 cyclohexane; aromatic rings use lowercase: c1ccccc1 benzene
  • Wildcards — [*] (brackets required; a bare * is invalid)

To find a SMILES for a new monomer, look it up in PubChem or draw it in a chemical editor and export. Then replace the two connection-point hydrogens with [*] wildcards, one per side for a middle monomer.

Next Steps