All Articles

A Study of Terpene Preservation in Child Resistant Rigid Packages




As the cannabis industry’s leading rigid packaging provider, eBottles endeavored to determine terpene preservation in the various child resistant rigid packages available to the market. Partnering with The Good Lab testing laboratory in Colorado Springs, CO, eBottles tested 11 different types of packages with tests completed at eight, sixteen and twenty-four week intervals. 

eBottles and The Good Lab filled 33 containers, each containing 3.5g of flower, from a homogenized batch of cannabis supplied by Rocky Road dispensary in Aurora, CO. 

  • We tested eBottles brand containers:

➜   Glass Jars made from clear glass, black glass and UV coated clear glass
      all with closures with foil liners. (3)

➜   Plastic Jars made from PET resin in clear, black and clear with a UV inhibitor
      all with closures with foil liners. (3)

 Total: 6 versions

  • We also tested Competitor packages including:

➜   Pop Top Vials - made from PP material
➜   Push ‘N Turn Vials - made from PP material
➜   Calyx brand “drams” - made from PP material with Calyx proprietary closures
➜   Calyx brand Glass Jars - with Calyx proprietary closures
➜   “Atid” brand Paper Tubes from MMC Depot

Total: 5 versions



The laboratory results were surprising.

The jars made from PET plastic performed as well as glass in preserving terpenes.

All of the containers made from PP (polypropylene) material performed poorly. This was surprising as it is well known that PP provides an excellent moisture barrier. Despite this fact, all of the PP containers tested lost terpenes at a rapid rate. Our conclusion is that the problem lies with the closure seal. Pop Top and Push ‘N Turn vials are meant to meet the USP standard of “moisture tight” but our tests revealed that effective seals were not being made. Along with a decline in moisture loss, terpene loss was dramatic - even over the shorter duration test periods.

We were also surprised to see that the Calyx brand square glass jars with press on lids also performed poorly in terms of terpene preservation. No doubt this finding was the result of the closure not making an effective seal on the jar.

The paperboard container was at the bottom of the range with dramatic terpene loss.

 eBottles remains committed to furthering research in the quest to find the absolute best packaging options for cannabis products.

With the growth of the legal cannabis market, the need for high quality, performance packaging for flower, which also meets various regulatory requirements has burgeoned. The need to gain a better understanding of the efficacy of various available packaging options to best preserve the integrity of the product is critically important.

eBottles, as a leader in the industry, set out to find the best combination of materials for cannabis packaging by testing nearly all of the different combinations commonly found in the retail cannabis market. 

The packages included in this test were made of various materials including Glass, Polypropylene (PP) and Polyethylene Terephthalate (PET) with foil liners; we also tested a paperboard package.  Samples were filled and stored in sets for four time periods: four, eight, sixteen and twenty-four week intervals.



Samples were filled with a measured 3.5g of cannabis flower from a single harvest batch which had cured for ten days prior to being used in this test.

Nineteen sample types were tested, divided into three categories.

Package Sample Test Types
Container Comparison Set Glass Liner Comparison Set PET Liner Comparison Set
eB UV inhibited clear PET Polypropylene Pop-top eB Foil liner eB Foil liner
eB UV coated clear glass Push & turn vial
eB Normal Clear PET Calyx Polypropylene
eB Normal Clear Glass Calyx Glass
eB Black glass Pearlloc paper tubes
eB Black PET


Figure 1. List of all the package types included in the test.  For all eBottles(eB) jars in the container comparison set an eBottles screw child resistant closure with a foil liner was used.  eBottles black glass and eBottles black PET 53 mm jars were used in the glass and PET liner comparison set respectively.

The testing processes for the compounds being measured (terpenes content) required the flower to be destroyed. Given this, for the 11 sample types studied, three separate samples were made of each of the types for each of the three time intervals (8, 16 and 24 weeks).  Therefore a total of 33 samples were created for each of the three test intervals totalling 99 total samples made for the three time intervals combined.

Additionally three baseline samples of 3.5g of cannabis flower were taken from the same homogenized batch for immediate testing to give a baseline starting value for the terpene compounds being measured.




   Mettler Toledo Scale .01 gram

   Nidec TNP 10 Digital Torque Meter

   Thermo Scientific F1-ClipTip Pipettor

   500 AC Electromagnetic Induction Foil Sealing Machine

   Bud Size Separating Rack

   The Good Lab equipment is identified in their section within the procedure section.


    A single batch of flower was homogenized for bud size using a Bud Size Separating Rack. The buds were then grouped and weighed into 3.5g bundles to a .01g accuracy using the Mettler scale.  Each 3.5g of cannabis flower was then randomly distributed into each of the sample jars. Samples were filled in order of the list in table 2 below.  All three samples to make a sample set for each of the three time trials were filled together, nine total for each test package type.

    Additionally a control sample set of three samples containing 3.5g each were filled into a black glass 53 mm jar and closed with a foil lined child resistant closure and submitted to the lab for immediate testing to establish a baseline value for each of the three compounds tested. 

    The remaining samples were placed into jars and closed in the appropriate manner for the package type.  All jars using a screw closure were applied to the industry standard torque specification using a digital torque meter (glass: 19-32 in-lb, PET: 27-32 in-lb). All jars were then labeled with a unique code to insure blind testing and grouped into storage bins according to the time frame they would be stored (8,16, and 24 weeks).

    The samples were stored in three separate uncovered storage bins (one bin for each time period).  The bins were placed below a skylight in a store room at a grow facility with moderately controlled temperature (range 68° to 72°) and uncontrolled ambient humidity. 

    The bins were transferred to a retail dispensary to get processed through the required regulatory compliance Colorado cannabis sales system.  The samples were then driven to the Good Lab for testing. No samples were opened at any time during this transfer nor were they opened at any other time during the test phase.

    Upon arrival at the lab each sample was tested by The Good Lab for terpene content.

    Upon opening each sample 3.5g of flower, the terpene tests were done immediately .Terpene test results were reported in percentage by weight using Gas Chromatography with Mass Spectrometry (GC/MS). Samples were tested as presented and not adjusted for moisture.

    Moisture content was determined by first weighing each sample as presented, drying in a dehydrator at 145°F for 1 hour 45 minutes (105 minutes), then weighing the sample again. This drying process evaporates the moisture while preserving the cannabinoids, plant material and most terpenes.

    The testing lab was blind as to what the specifics of the package types the samples were stored in order to avoid bias results. Container information was not revealed to the lab until after testing was completed.

    At the completion of each time test period (8, 16 and 24 weeks) the cannabis flower in all container samples included in that time period group was tested. Total terpenes percentages were measured and recorded for each.

    The three sample sets for each of the 11 package types were averaged together for each of the terpene compounds measured.  We will refer to these test set averages as the sample set average. 


    Package type comparisons

    For each three samples comprising a sample test set for the 11 package types a sample set average was calculated for the terpene compounds.

    Findings for each comparison were determined using a comparative analysis where each of the 11 sample sets could be compared relative to one another.  The comparisons were ranked using the deviation from the mean value over all 11 package sample set averages for terpene compounds.

    Terpene Testing

    While individual terpenes were measured as a part of the terpene testing for the purposes of this study all individual terpene percentages measured were combined to give a total terpene percentage value for each sample.  Figure 6 shows the plus and minus deviation from the mean for the 19 package group. Analysis of the mean total terpene content in the cannabis flower in the 19 package types showed a strong correlation in superior terpene preservation among all screw cap packages. (see Figure 6)

  • Non screw cap jars performed poorly compared to the full test group.
  • Polypropylene Pop-tops as well as glass and plastic Calyx jars and the Pearlloc tubes all performed below the mean of the full test group in the 8 ,16 and 24 week test. Combining all of the averages of these five container types for all of three of the test periods they yield a group average mean terpene percentage 0.4 below the mean. The six screw cap eBottles container types combined for a .2 percentage above the mean percentage of the field.  That is a range of .6 percent which is two standard deviations from the mean terpene percentage for the field. 


    % Terpenes Adjusted for Moisture Loss
    Used cap with an eBottles (eB) foil liner on all jars with screw caps


    8 WEEKS

    16 WEEKS

    24 WEEKS

    eB UV inhibited clear PET

    0.20 0.44 0.32

    eB UV coated clear glass

    0.10 0.36 0.29

    eB Normal Clear PET

    0.20 0.36 0.38

    eB Normal Clear Glass

     -0.08  0.06 0.29

    eB Black glass

     -0.01  0.04 0.12

    eB Black PET

    0.16 0.22 0.25


     -0.29   -0.37   -0.32 

    Push & turn vial

     -0.18   -0.40   -0.44 

    Calyx Polypropylene

     -0.38   -0.61   -0.56 

    Calyx Glass

     -0.24   -0.42   -0.49 

    Pearlloc Paper Tubes

     -0.33   -0.53   -0.55 

     Figure 2. Terpene results. The table shows the variation from the mean terpene percent of the 11 package group for each package type.  Values in red were below the mean moisture percent of the entire group for that time period while values in black had a mean moisture above the group mean for that time period.



    Cannabis is a volatile product. Preservation of terpene content over eight, sixteen even twenty four weeks of storage is critical to delivering a viable product to customers. The packaging selected, along with correct application, determines the shelf life of the product. 

    We found that many of the commonly used cannabis packages do not perform satisfactorily and are poor choices for product preservation over time.

    Many customers are confused by the variety of plastic materials available for packaging. Many marketers are not open about the performance capabilities of the packaging they offer.

    Our tests confirmed that opaque PET plastic containers perform as well as Glass in preserving terpenes over time. Polypropylene plastic containers and vials and paper containers by comparison performed poorly.

    Article Categories


    You have added to your cart

    View Cart Close