Fig 1.
(A) Diagram of the device. Air is distributed between 8 electrovalves (EV), each of which directs it either through an empty vial (via normally open exit, NO) or an odorized vial (via normally closed exit, NC). All vial outlet tubes are collectively connected to a glass tube that transports the stimuli to the preparation. (B) Diagram of the mass transfer model representing odorant transport from one vial to the antenna. The odorant is diluted at a concentration Cl in a volume Vl of mineral oil. The odorant diffuses through the air-solvent interface of area A with a mass transfer coefficient Kglob that depends on the air-solvent partition coefficient Khl. kl and kh are the mass transfer coefficients from bulk liquid to interface and from interface to bulk headspace, respectively. With the headspace of volume Vh, the odor concentration has concentration Ch. During stimulation, an airflow Qs passes through the odor source, transports odor molecules, and merges with a non-odorized airflow Qc-Qs within the glass tube (transport compartment). Between stimuli, no airflow enters the odor source. The delivered airflow Qc is constant throughout the experiment. Vc is the volume of the glass tube and Cc is the odor concentration at the outlet of this compartment. The 3 black arrows in (B) indicate the locations where odorant concentration were measured with a PID.
Table 1.
List of variables and parameters of the model.
Table 2.
Partition coefficient values measured for the panel of tested VPCs.
Fig 2.
Results of stimulation system modelling.
Predicted relationship between the partition coefficient (Khl) and (A) the global mass transfer coefficient from source liquid to source headspace (kglob); and (B) the dimensionless concentration at dynamic pseudo-stationary regime inside the source’s headspace (Yh,psr). Different line styles and colors indicate alternative couples of kh and kl values. Predicted time-course of the odorant concentration (C) inside the source headspace and (D) at the source outlet for 4 Khl values. Simulation parameters: in A-D: = 1.38 cm² and
= 200 mL.min-1, i.e. values of the odor delivery device. In C,
= 1 mL and
= 3 mL, settings of the odor delivery device, kl = 10−5 and kh = 10−2. In D, simulations are for a transport compartment with the same dimensions as a PID inlet needle (
= 0.026 mL,
= 750 mL/min) located at source outlet.
Table 3.
Values of air and liquid transfer coefficients (kh and kl) estimated from the fit of equation (13) on PID measured values of Yh,psr.
Fig 3.
Dimensionless concentration delivered by odorant sources at dynamic pseudo-stationary regime (Yh,psr), as a function of the partition coefficient (Khl) of odorants.
Black circles: observed Yh,psr values, measured at source exit using the calibrated PID. Red lines: fit of equation (13) to the observed Yh,psr data (RSS = 0.146, thick: estimate, thin: confidence interval, computed from the limits of 95% confidence intervals for kh and kl). Black crosses: predicted values for the partition coefficients corresponding to the panel of VPCs (Table 2).
Fig 4.
Comparison of observed (green circles and crosses) and predicted (red crosses) concentration at pseudo-stationary regime.
(A) Comparison of predicted values of pseudo-stationary concentration inside source headspace (Ch,psr, red crosses) with pseudo-stationary concentrations measured with the calibrated PID at source outlet (green circles) and at Teflon tubing outlet (green crosses). is included for comparison. (B) Comparison of predicted values of pseudo-stationary concentrations in the glass tube (Cc,psr, red crosses) with plateau concentrations measured with the calibrated PID at glass tube outlet (green circles).
divided by 8 (blue bars) is given for comparison.
Fig 5.
PID signal recorded at source outlet in response to 60-s puffs of 11 VPCs sorted vertically from top to bottom by increasing air-mineral oil partition coefficient.
Response amplitudes were normalized relative to the plateau. For clarity, from the bottom (isoprene) to the top (β-caryophyllene), each recording is shifted up and to the right.
Fig 6.
Expected effect of source diameter and airflow on delivered concentrations.
(A) Theoretical relationship between source diameter (d) and airflow through the odor vial (Qs) and the mass transfer coefficient from interface to headspace (kh), as predicted by equation (12). The point corresponding to the experimental source diameter and flowrate is indicated by a square. (B) Effect of source diameter (d) on the expected relationship between Khl and Yh,psr, as predicted by equations (11), (12) and (14). Computed for Qs = 200 mL.min-1, kl = 10−7, and d varying between 10 and 300 mm (headspace volumes varying between 0.8 mL and 21 L). (C) Effect of source airflow Qs on the expected relationship between Khl and Yh,psr, as predicted by equations (11), (12) and (14). Computed for source diameter d = 1.32 cm, kl = 10−7 and Qs varying between 10 mL min-1 and 5 L min-1. Points and curves are color-coded by Log10 of kh value.
Fig 7.
Simulated evolution of source lifetime.
(A) Relationship between source lifetime and Khl (red). The relationship between Yh,psr and Khl (blue) is given as a reference. Source lifetime is defined as the time t after valve opening at which Yh(t) = 0.95 × Yh,psr. Source dimension, airflow and transfer coefficients correspond to those of the odor delivery device used in this work. Points correspond to Khl values from to the panel of tested VPCs. (B) Source lifetime as a function of source diameter (x axis) and airflow (line shading). Calculated for Khl = 6.3 × 10−4 (value of (E)-2-hexenal), kl = 10−7, and for a source liquid volume = 1/3 of the source headspace volume.
Fig 8.
EAG dose-response of male and female A. ipsilon to (Z)-3-hexenyl acetate (Z3HA).
Response intensity to control stimuli was subtracted from the raw response intensities. Stars indicate significant differences to control (one sample t tests) or between males and females (ANOVAs). N = 10 for males and females.
Fig 9.
Response of pheromone-sensitive ORNs from male A. ipsilon to pheromone in a background of (Z)-3-hexenyl acetate (Z3HA).
(A) Time course of the firing response to Z3HA background (controls = no Z3HA background were tested before (control 1), and after (control 2) the series of Z3HA backgrounds) and to pheromone puffs (10 ng, 0.2 s). Green box and red line at the bottom indicate the timing of background and pheromone delivery, respectively. Patterns of neuronal activity under the five lowest tested background concentrations are identical to control. (B) Responses to Z3HA and (C) to pheromone in Z3HA background. Z3HA backgrounds are expressed as both a dilution in mineral oil and a concentration delivered at dynamic pseudo-stationary regime (ppbv). Stars indicate background concentrations under which average firing response differs significantly from that observed under the first control background (Anova followed by Dunnett’s test; * = P < 0.05; ** = P < 0.01; *** = P < 0.001). N = 16-20.