File Name : supplementary figure 1 -  device fabrication sequence.tif

Caption : supplementary figure 1. cell processing device fabrication sequence. devices are batch fabricated with 32 devices on a 4” wafer; the sequence shown is for a cross-section of a single device. as the electrodes are deposited along the channel sides, they are not shown past step a10; see figure 2 and figure 5 for alternative views of the device

File Name : supplementary figure 2 - dispersion effects.tif

Caption : supplementary figure 2. taylor-aris dispersion modelling informs optimal flowrate, tube diameter, and tube lengths to maintain small concentrations of analytes resulting from the small lysate volume while balancing time and pressure drop considerations. (top) plot of effective diffusion coefficient vs flowrate for a 10 cm length of tubing shows that dispersion becomes dominate means of diffusive transport at higher flowrates.  (middle) plot of mean concentration (normalized cm/co where co is initial concentration) vs flowrate for 10 cm length of tubing shows that dilution curves collapse for microfluidic tube diameters (10-100 μm id) at nano-esi flowrates (1-10 nl/s). (bottom) plot of mean concentration vs flowrate for varying tube lengths for a 100 μm id tube highlights importance of minimizing tube length to mitigate dilution effects. reducing length also reduces transit time and pressure drop. for reference the experimental flowrate was 30 μl/hr (8.3 nl/s).

File Name : supplementary figure 3 - methionine and arginine analysis.tif

Caption : supplementary figure 3. both methionine and arginine show little variation in the protonated monoisotopic mass traces (top trace) but display distinct signal increases in multiple fragments reported in the massbank europe database for esi-ms. the traces are normalized by the maximum signal intensity for the displayed time range; the maximum signal intensity and corresponding mass are provided to the right of each trace.

File Name : supplementary figure 4 - phenylalanine analysis.tif

Caption : supplementary figure 4. the nh4+ adduct of phenylalanine (right) displayed a distinct signal increase compared to the protonated monoisotopic mass trace (left). the traces are normalized by the maximum signal intensity for the displayed time range.

File Name : supplementary figure 5 - representative spectra.tif

Caption : supplementary figure 5. representative spectra of both detected and undetected metabolites. for each analyte, the target m/z is shown for time periods before (top), during (middle), and after (bottom) lysate elution; the spectra are averaged over 15 seconds for each time point. signal to noise (sn) values are provided for each m/z marker with the detected signals circled in red; tryptophan was not detected. the target m/z value is listed below each sub-figure and denoted as either a protonated monoisotopic mass or fragment mass; instrument error was taken as 10 ppm for the analysis.

File Name : supplementary figure 6 - effect of cell number.tif

Caption : supplementary figure 6. traces of the protonated monoisotopic mass of each amino acid highlight dependence on number of cells loaded. the traces are normalized for each analyte by the maximum signal intensity (given in parenthesis) for the displayed time range.