State-to-State Rate Coefficients for NH3–NH3 Collisions from Pump–Probe Chirped Pulse Experiments

The kinetics of rotational inelastic NH3–NH3 collisions are recorded using pump–probe experiments, carried out with a K-band waveguide chirped pulse Fourier transform microwave spectrometer, in which the population of one inversion doublet is altered by the pump pulse. Due to self-collisions, the resulting deviation from equilibrium propagates to other states and, thus, can be interrogated by probe pulses as a function of the pump–probe delay time. A clear hierarchy of the state-to-state collision processes is found and subsequently translated into propensity rules. State-to-state rate coefficients are estimated, first via an analysis of the kinetics, and then more robustly and accurately derived from the pressure-dependent measurements using a global fitting procedure.


Chirped-Pulse Fourier Transform Spectrometer
The spectrometer at the Center for Astrochemical Studies at MPE is based on the chirpedpulse technique S1 coupled with a K u -band waveguide. It uses a 5 GHz, 2-channel arbitrary waveform generator (Keysight M8190A) as a source to generate probe (chirp) and pump pulses. Its I/Q signals are upconverted in an I/Q mixer (Marki Microwave MLIQ-1845L) by the power splitted output (Marki Microwave PD-0140) of a phase-locked 70 GHz signal generator (Keysight PSG E8257D). A 4 W solid-state amplier (Microsemi C1826-36-T964) with internal high-speed pulser, whose output is protected by a circulator (DiTom D3C1826K), amplies the signal before it is fed into one end of a 1.5 m long K u -band waveguide. The other end is connected to a low noise amplier (B&Z BZR-P0226500-351032-182525), whose input is protected by a pin diode switch (Kratos F9022). A second I/Q mixer downconverts the signal. Both outputs are again amplied (MiniCircuit ZX60-V62+) and connected via lters (Mini-Circuits SHP-100A+; 0.1-3.0 GHz) to the inputs of a 2-channel digitizer (Acqiris M9310A). The arbitrary waveform generator, digitizer and signal generator are all locked to a rubidium reference (SRS FS725), whose 1 PPS signal is used to trigger the pulse sequence.
The sample marker outputs of the arbitrary waveform generator trigger the acquisition in the digitizer as well as a delay generator which controls the protection switch and the amplier pulsar. Two pressure ports are mounted close to both ends of the waveguide. Capacitance gauges (Pfeier CMR 365, 0.1 hPa F.S.) are connected to both of them followed by needle valves that control the ow of the molecular sample through the waveguide towards the turbo molecular pump to guarantee a constant pressure in the waveguide.

Measurements
Measurements have been conducted at pressures ranging from 4 to 30 µbar by adjusting a slow stable ow in the order of 10 −5 mbar l/s of the ammonia sample (Air Liquide N38, 99,98%) through the waveguide. The pressure was continuously read during the experiment.
S-2 Its gradient between both ends was typically 1 µbar, and its stability about 0.5 µbar.
In contrast to previous, similar experiments (Oka S2 ), a rather low power has been se- The temporal evolution of the systems studied in this work has been simulated numerically.
The simulations are based on optical Bloch equations S3 in which the phenomenological rates T −1 1 are replaced by a set of rate equations (master equation). The global t uses one common set of parameters to calculate the temporal evolution of each system and to t the experimental data.
In total, 29 parameters are included in the global t: for each inversion doublet addressed by the pump pulse (J=1-6) the Rabi frequency (6) and a coherence loss rate (6) is included for each inversion doublet the rate coecient for parity-changing collsions within the doublet (6) is included for each collision induced transition with (∆J = 1), one rate coecient for paritychanging (5) and one rate coecient for parity-conserving (5) collsisions are included. one pressure-independent coherence loss rate (1) is included Rate coecients involving the inversion doublets (7, 1) are not included in the t, because this inversion transition has not been observed. Therefore, rate coecients determined in the t for higher J are not presented in Tab. 2 of the letter.