Ion Implantation 1970s Accelerate Ions Energy Not Temperature

# Step 2 — Accelerate Ions to a Precise Energy, Not a Precise Temperature: Two Orthogonal Knobs That Decouple Depth From Dose

## 1. Why Kinetic Energy and Electric Charge Replace Temperature and Time

By extracting ionized dopant atoms into an electrostatic accelerator, ion implantation splits the coupled thermal schedule of diffusion into two completely independent, electronically metered physical controls: acceleration voltage sets the depth where the ions stop, while integrated beam current counts the exact number of dopants delivered. In Step 1, this series demonstrated that every diffusion furnace since 1954 forced depth and surface concentration onto an inflexible thermodynamic curve because both quantities answered to the same temperature $T$ and drive time $t$. Ion implantation abandons thermodynamic equilibrium entirely. In this apparatus, dopant gas is stripped of electrons in an arc discharge to create positive ions ($B^+$, $P^+$, or $As^+$), filtered through a magnetic mass analyzer to reject all isotopic impurities, and accelerated down an electrostatic column across a calibrated potential difference $V_{\text{acc}}$. The depth the ions reach inside the silicon crystal depends strictly on their kinetic energy ($E = q V_{\text{acc}}$), while the total dose delivered ($\Phi$) is measured by integrating the electric beam current striking the wafer in a Faraday cup—two physical mechanisms governed by entirely separate laws of physics.

$$R_p = f(q V_{\text{acc}}), \qquad \Phi = \frac{1}{q A}\int_0^{t_{\text{implant}}} I_{\text{beam}}(t)\,dt$$

where $R_p$ is the mean projected range beneath the surface, $V_{\text{acc}}$ is the acceleration voltage (typically $10\text{ to }200\ \text{keV}$), $I_{\text{beam}}$ is the measured ion beam current, $q$ is the elementary charge ($1.602 \times 10^{-19}\ \text{C}$), $A$ is the scanned wafer area, and $\Phi$ is the resulting dopant dose in ions per square centimeter. Changing the acceleration voltage moves the dopants deeper into the silicon without adding a single extra atom to the dose. Conversely, adjusting the beam current or exposure time multiplies the dose across five orders of magnitude without shifting the peak depth by a single angstrom.

The Ion Implantation Column: Two Orthogonal Controls high voltage sets penetration depth; current integrator counts delivered dose ION SOURCE Plasma Arc ANALYZING MAGNET Mass Filter (q/m) ACCELERATION TUBE KNOB 1: VOLTAGE (Vacc) 10 keV — 200 keV → Depth Scan Plates FARADAY CUP / TARGET Wafer KNOB 2: CURRENT (∫I dt) Electrometer → Precise Dose THE PHYSICAL PRINCIPLE OF ORTHOGONALITY ✓ Kinetic Energy Knob: E = q · Vacc fixes stopping range Rp without introducing a single extra atom ✓ Charge Counting Knob: Dose Φ = Q / (q A) counts total atoms to 1% accuracy without changing depth A process room temperature physics: the wafer stays cold, and thermal equilibrium no longer dictates the doping.

## 2. Real Diagram: Depth Control by Voltage vs. Dose Control by Beam Charge

The parametric charts below illustrate the complete independence of penetration depth and total concentration under electrostatic beam control.

Decoupled Physical Knobs: Voltage Sets Depth, Current Sets Dose demonstrating true orthogonality: changing one variable leaves the other strictly constant KNOB 1: PROJECTED RANGE VS. ENERGY Acceleration Energy E = q Vacc (keV) → Projected Range Rp (µm) Boron (B+) Phosphorus (P+) 50 keV ≈ 0.16 µm 100 keV ≈ 0.30 µm → Range depends purely on nuclear & electronic stopping KNOB 2: DELIVERED DOSE VS. BEAM CHARGE Integrated Charge Q = ∫ Ibeam dt (µC) → Implanted Dose Φ (ions/cm²) Φ = Q / (q A) 10¹¹ cm⁻² (Threshold Adjust) 10¹⁵ cm⁻² (Source/Drain) → 5 orders of magnitude dynamic range at constant depth Diffusion forced depth and dose along a single thermal trajectory. Ion implantation spans the full 2D parameter space: any depth can receive any dose with electrometer precision.

## 3. The Structural Mechanics of Stopping: Why the Peak Is Buried

When dopant atoms enter the silicon substrate by diffusion, thermal random walks ensure that the highest concentration is always at the surface ($x = 0$), decaying monotonically into the bulk according to the complementary error function:

$$N(x) = N_0\,\text{erfc}\!\left(\frac{x}{2\sqrt{Dt}}\right)$$

Ion implantation completely upends this boundary condition because the ions enter the wafer as a directed projectile with initial kinetic energy $E_0$. As each ion penetrates the silicon lattice, it experiences continuous decelerating drag:
1. Electronic Stopping ($S_e$): Inelastic drag against the sea of electrons in the silicon crystal, dominating at high velocities ($E > 100\ \text{keV}$) like viscous friction without significant angular deflection.
2. Nuclear Stopping ($S_n$): Elastic Coulomb collisions directly with the heavy silicon atomic nuclei, dominating at lower velocities as the ion slows down, which violently deflects the ion and brings it to rest.

Because the ion must lose its kinetic energy before it halts, it comes to rest around a mean projected depth $R_p$, creating an approximately Gaussian distribution centered beneath the surface:

$$n(x) = \frac{\Phi}{\sqrt{2\pi}\,\Delta R_p} \exp\left(-\frac{(x - R_p)^2}{2\,\Delta R_p^2}\right)$$

where $\Delta R_p$ is the projected straggle (the statistical dispersion caused by stochastic collisions). For the first time in semiconductor fabrication, the dopant peak is buried below the surface, while the concentration drops off on *both* sides of $R_p$.

Step 2 establishes the physical foundation of ion implantation: by accelerating charged particles rather than heating a furnace tube, semiconductor manufacturing acquires two independent, orthogonal, and metrologically exact knobs—turning doping from an uncontrolled thermodynamic diffusion into an electrostatically mastered beam process.

Take ion implantation 1970s accelerate ions energy not temperature further

Ask the copilot about this term, or have our engineers assess it against your process.